D50311GC11 sg1

s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
Oracle Database
@ sic de11g: nt RAC
e r de Stu
Administration
v i lav e this
ic i o_ Volume
o usI • Student Guide
r
ab nse t
( f
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
D50311GC11
Edition 1.1
September 2008
D56241
Unauthorized reproduction or distribution prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
Author Copyright © 2008, Oracle. All rights reserved.
James Womack Disclaimer
This document contains proprietary information and is protected by copyright and

Technical Contributors other intellectual property laws. You may copy and print this document solely for your
own use in an Oracle training course. The document may not be modified or altered in
and Reviewers any way. Except where your use constitutes "fair use" under copyright law, you may
Christopher Andrews not use, share, download, upload, copy, print, display, perform, reproduce, publish,
license, post, transmit, or distribute this document in whole or in part without the
Troy Anthony express authorization of Oracle.
Lothar Auert
The information contained in this document is subject to change without notice. If you
Bruce Carter find any problems in the document, please report them in writing to: Oracle University,
500 Oracle Parkway, Redwood Shores, California 94065 USA. This document is not
Michael Cebulla warranted to be error-free.
Carol Colrain
Restricted Rights Notice
Jonathan Creighton
Joel Goodman If this documentation is delivered to the United States Government or anyone using
the documentation on behalf of the United States Government, the following notice is
Arturo Gutierrez applicable:
s a
Lutz Hartmann
)h a
ฺbr
U.S. GOVERNMENT RIGHTS
Pete Jones The U.S. Government’s rights to use, modify, reproduce, release, perform, display, or
m
o deฺ
disclose these training materials are restricted by the terms of the applicable Oracle
c
David Kirby
i ฺ
d Gui
license agreement and/or the applicable U.S. Government contract.
Roland Knapp r e
Miroslav Lorenc
Trademark Notice
@ sic dent
Barb Lundhild
e r de Stu
Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other
lav e this
names may be trademarks of their respective owners.
Roderick Manalac
v i
Sabiha Miri
ic i o_ o us
Philip Newlan
( f a br se t
Roman Niehoff
r d e licen
Erik Peterson
V e ble
Stefan Pommerenk
V a
il fera
r i c i rans
MarshalloPresser
FabSrinivas
Rick Pulliam
o n -t
n Putrevu
Roy Rossebo
Ira Singer
Linda Smalley
Ranbir Singh
Harald van Breederode
Michael Zoll
Editors
Raj Kumar
Richard Wallis
Nita Pavitran
Amitha Narayan
Graphic Designer
Satish Bettegowda
Publishers
Nita Brozowski
Unauthorized
Veenareproduction
Narasimhan or distribution prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
Contents
I Introduction
Overview I-2
Course Objectives I-3
Typical Schedule I-4
A History of Innovation I-5
What Is a Cluster? I-6
Oracle Real Application Clusters I-7
s a
Benefits of Using RAC I-8
)h a
Clusters and Scalability I-9
m ฺbr
Levels of Scalability I-10 i ฺ c o deฺ
Scaleup and Speedup I-11 r e d Gui
Speedup/Scaleup and Workloads I-12 @ sic dent
I/O Throughput Balanced: Example I-13
e r de Stu
i lav e this
Performance of Typical Components I-14
v
i o_ o us
Complete Integrated Clusterware I-15
ic
( f abr nse t
Necessity of Global Resources I-16
rde le lice
Global Resources Coordination I-17
e
l a V ab
Global Cache Coordination: Example I-18
i
V sfer
Write to Disk Coordination: Example I-19
i o
ic -tran
Dynamic Reconfiguration I-20
b r
Fa non Object Affinity and Dynamic Remastering I-21
Global Dynamic Performance Views I-22
Additional Memory Requirement for RAC I-23
Efficient Internode Row-Level Locking I-24
Parallel Execution with RAC I-25
RAC Software Principles I-26
RAC Software Storage Principles I-27
RAC Database Storage Principles I-28
RAC and Shared Storage Technologies I-29
Oracle Cluster File System I-31
Automatic Storage Management I-32
CFS or Raw? I-33
Typical Cluster Stack with RAC I-34
RAC Certification Matrix I-35
iii
RAC and Services I-36
Available Demonstrations I-37
1 Oracle Clusterware Installation and Configuration

Objectives 1-2
Oracle RAC 11g Installation 1-3
Oracle RAC 11g Installation: Outline 1-5
Windows and UNIX Installation Differences 1-6
Preinstallation Tasks 1-7
Hardware Requirements 1-8
Network Requirements 1-9
Virtual IP Addresses and RAC 1-10 a
RAC Network Software Requirements 1-11 as
Package Requirements 1-12 ฺbr )h
Required UNIX Groups and Users 1-13 c m
o deฺ
i ฺ
d Gui
oracle User Environment 1-14 r e
User Shell Limits 1-15
@ sic dent
Configuring for Remote Installation 1-16
e r de Stu
v i lav e this
Required Directories for the Oracle Database Software 1-18
ic i o_ o us
Linux Operating System Parameters 1-20
Cluster Setup Tasks 1-22
( f abr nse t
Verifying Cluster Setup with cluvfy 1-23
e rde le lice
Installing Oracle Clusterware 1-24
l a V ab
i
V sfer
Specifying the Inventory Directory 1-25
i o
ic -tran
Specify Home Details 1-26
b r
Fa nonProduct-Specific Prerequisite Checks 1-27
Cluster Configuration 1-28
Private Interconnect Enforcement 1-29
Oracle Cluster Registry File 1-30
Voting Disk File 1-31
Summary and Install 1-32
Run Configuration Scripts on All Nodes 1-33
End of Installation 1-34
Verifying the Oracle Clusterware Installation 1-35
Summary 1-37
Practice 1: Overview 1-38
2 RAC Software Installation

Objectives 2-2
Installing Automatic Storage Management 2-3
Installation Type 2-4
iv
Install Location 2-5
Hardware Cluster Installation Mode 2-6
Product-Specific Prerequisite Checks 2-7
Select Configuration Option 2-8
Privileged Operating System Groups 2-9
Summary 2-10
Execute Configuration Scripts 2-11
End of Installation 2-12
ASM Configuration 2-13
Creating ASM Disk Groups 2-17
Select a Product to Install 2-19
Select Installation Type 2-20
s a
Install Location 2-21
)h a
Specify Cluster Installation 2-22
m ฺbr
Products Prerequisite Check 2-23
i ฺ c o deฺ
Select Configuration Option 2-24
r e d Gui
Privileged Operating System Groups 2-25
@ sic dent
Check Summary 2-26
e r de Stu
root.sh Script 2-27
v i lav e this
ic i o_ o us
Required Tasks Prior to Database Creation 2-28
Summary 2-31 ( f abr nse t

Checks Before Database Creation 2-29
e rde le lice
l a V ab
i
V Database f er Creation
i
3 o RAC
ic Objectives s
an 3-2
b r - t r
Fa non Management Agent Installation: Specify Installation Type 3-3
Specify Installation Location 3-4
Specify Cluster Installation Mode 3-5
Prerequisite Check and OMS Location 3-6
Agent Registration Password 3-7
Management Agent Installation Finish 3-8
Executing the root.sh Script 3-9
Creating the Cluster Database 3-10
Node Selection 3-11
Select Database Type 3-12
Database Identification 3-13
Cluster Database Management Method 3-14
Passwords for Database Schema Owners 3-15
Storage Options for Database Files 3-16
ASM Disk Groups 3-18
v
Database File Locations 3-19
Recovery Configuration 3-20
Database Content 3-21
Initialization Parameters 3-22
Security Settings and Maintenance Tasks 3-23
Database Storage Options 3-24
Create the Database 3-25
Monitor Progress 3-26
Postinstallation Tasks 3-27
Check Managed Targets 3-28
Single Instance to RAC Conversion 3-29
Single-Instance Conversion Using the DBCA 3-30
s a
Conversion Steps 3-31
)h a
Single-Instance Conversion Using rconfig 3-34
m ฺbr
Single-Instance Conversion Using Grid Control 3-36
i ฺ c o deฺ
Summary 3-38 r e d Gui
@ sic dent
e r de Stu
4 RAC Database Administration
v i lav e this
Objectives 4-2
ic i o_ o us
Cluster Database Home Page b4-3
a r et
Cluster Database Instance e ( f
Home Page ns 4-5
d li c e
r lAdministration
Cluster Database e Instance e Page 4-6
a V a b
Cluster Home
V il Pagefer4-7
The
i o a s
ciConfiguration
n Section 4-8
b r t r
- Viewer 4-10
Fa Topology
non
Enterprise Manager Alerts and RAC 4-11
Enterprise Manager Metrics and RAC 4-12
Enterprise Manager Alert History and RAC 4-14
Enterprise Manager Blackouts and RAC 4-15
Redo Log Files and RAC 4-16
Automatic Undo Management and RAC 4-17
Starting and Stopping RAC Instances 4-18
Starting and Stopping RAC Instances with SQL*Plus 4-19
Starting and Stopping RAC Instances with SRVCTL 4-20
Switch Between the Automatic and Manual Policies 4-21
RAC Initialization Parameter Files 4-22
SPFILE Parameter Values and RAC 4-23
EM and SPFILE Parameter Values 4-24
RAC Initialization Parameters 4-26
Parameters That Require Identical Settings 4-28
vi
Parameters That Require Unique Settings 4-29
Quiescing RAC Databases 4-30
Terminating Sessions on a Specific Instance 4-31
How SQL*Plus Commands Affect Instances 4-32
Transparent Data Encryption and Wallets in RAC 4-33
ASM: General Architecture 4-34
ASM Instance and Crash Recovery in RAC 4-36
ASM Instance Initialization Parameters and RAC 4-37
ASM and SRVCTL with RAC 4-38
ASM and SRVCTL with RAC: Examples 4-39
ASM Disk Groups with EM in RAC 4-40
Disk Group Performance Page and RAC 4-41 a
Summary 4-42 as
)h
ฺbr
c m
o deฺ
i ฺ
d Gui
5 Managing Backup and Recovery in RAC r e
Objectives 5-2
@ sic dent
Protecting Against Media Failure 5-3
e r de Stu
Archived Log File Configurations 5-4
v i lav e this
RAC and the Flash RecoveryiArea
ic o_ 5-5 o us
r t
(fab Settings
RAC Backup and Recovery Using EM 5-6
Configure RAC e Recovery e n se with EM 5-7
e
Archived Redo e ic
rdFilelConventions
l in RAC 5-8
a V b
V il RAC
Configure
f e raBackup Settings with EM 5-9
o
ici Oracle a s
nRecovery Manager 5-10
b r - t r
Fa nonConfigure RMAN Snapshot Control File Location 5-11
Configure Control File and SPFILE Autobackup 5-12
Crosschecking on Multiple RAC Clusters Nodes 5-13
Channel Connections to Cluster Instances 5-14
RMAN Channel Support for the Grid 5-15
RMAN Default Autolocation 5-16
Distribution of Backups 5-17
One Local Drive CFS Backup Scheme 5-18
Multiple Drives CFS Backup Scheme 5-19
Non-CFS Backup Scheme 5-20
Restoring and Recovering 5-21
Summary 5-22
vii
6 RAC Performance Tuning
Objectives 6-2
CPU and Wait Time Tuning Dimensions 6-3
RAC-Specific Tuning 6-4
RAC and Instance or Crash Recovery 6-5
Instance Recovery and Database Availability 6-7
Instance Recovery and RAC 6-8
Analyzing Cache Fusion Impact in RAC 6-10
Typical Latencies for RAC Operations 6-11
Wait Events for RAC 6-12
Wait Event Views 6-13
Global Cache Wait Events: Overview 6-14 a
2-way Block Request: Example 6-16 as
3-way Block Request: Example 6-17 ฺbr )h
2-way Grant: Example 6-18 c m
o deฺ
i ฺ
d Gui
Global Enqueue Waits: Overview 6-19 r e
Session and System Statistics 6-20
@ sic dent
Most Common RAC Tuning Tips 6-21
e r de Stu
v i lav e this
Index Block Contention: Considerations 6-23
ic i o_ o us
Oracle Sequences and Index Contention 6-24
( f abr nse t
Undo Block Considerations 6-25
rde le lice
High-Water Mark Considerations 6-26
e
Concurrent Cross-Instance Calls: Considerations 6-27
V ab
i l a
V sfer
Monitoring RAC Database and Cluster Performance 6-28
i o
ic -tran
Cluster Database Performance Page 6-29
b r
Fa nonDetermining Cluster Host Load Average 6-30
Determining Global Cache Block Access Latency 6-31
Determining Average Active Sessions 6-32
Determining Database Throughput 6-33
Accessing the Cluster Cache Coherency Page 6-35
Viewing Cluster Interconnects Page 6-37
Viewing the Database Locks Page 6-39
AWR Snapshots in RAC 6-40
AWR Reports and RAC: Overview 6-41
Automatic Database Diagnostic Monitor for RAC 6-43
What Does ADDM Diagnose for RAC? 6-45
EM Support for ADDM for RAC 6-46
Summary 6-47
viii
7 Services
Objectives 7-2
Traditional Workload Dispatching 7-3
Grid Workload Dispatching 7-4
Data Warehouse: Example 7-5
RAC and Data Warehouse: An Optimal Solution 7-6
Next Step 7-7
What Is a Service? 7-8
High Availability of Services in RAC 7-9
Possible Service Configuration with RAC 7-10
Service Attributes 7-11
Service Types 7-12 a
Service Goodness 7-13 as
Create Services with Enterprise Manager 7-14 ฺbr )h
Create Services with SRVCTL 7-15 c m
o deฺ
i ฺ
d Gui
Preferred and Available Instances 7-16 r e
sic dent
Modify Services with the DBMS_SERVICE Package 7-17
@
Everything Switches to Services 7-18
e r de Stu
v i lav e this
Use Services with Client Applications 7-19
ic i o_ o us
Use Services with the Resource Manager 7-20
( f abr nse t
Services and Resource Manager with EM 7-21
Services and the Resource Manager: Example 7-22
e rde le lice
Use Services with the Scheduler 7-23
l a V ab
i
V sfer
Services and the Scheduler with EM 7-24
i o
ic -tran
Services and the Scheduler: Example 7-26
b r
Fa non Use Services with Parallel Operations 7-27
Use Services with Metric Thresholds 7-28
Change Service Thresholds by Using EM 7-29
Services and Metric Thresholds: Example 7-30
Service Aggregation and Tracing 7-31
Top Services Performance Page 7-32
Service Aggregation Configuration 7-33
Service Aggregation: Example 7-34
trcsess Utility 7-35
Service Performance Views 7-36
Generalized Trace Enabling 7-37
Manage Services 7-38
Manage Services with Enterprise Manager 7-40
Manage Services with EM 7-41
Manage Services: Example 7-42
Manage Services: Scenario 7-43
ix
Using Distributed Transactions with RAC 7-44
Restricted Session and Services 7-46
Summary 7-47
8 High Availability of Connections

Objectives 8-2
Types of Workload Distribution 8-3
Client-Side Connect-Time Load Balancing 8-4
Client-Side Connect-Time Failover 8-5
Server-Side Connect-Time Load Balancing 8-6
Fast Application Notification: Overview 8-7 a
Fast Application Notification: Benefits 8-8 as
FAN-Supported Event Types 8-9 ฺbr )h
FAN Event Status 8-10 c m
o deฺ
i ฺ
d Gui
FAN Event Reasons 8-11 r e
FAN Event Format 8-12
@ sic dent
Load Balancing Advisory: FAN Event 8-13
e r de Stu
v i lav e this
Server-Side Callouts Implementation 8-14
ic i o_ o us
Server-Side Callout Parse: Example 8-15
( f abr nse t
Server-Side Callout Filter: Example 8-16
Configuring the Server-Side ONS 8-17
e rde le lice
Optionally Configure the Client-Side ONS 8-18
l a V ab
i
V sfer
JDBC Fast Connection Failover: Overview 8-19
i o
ic -tran
Using Oracle Streams Advanced Queuing for FAN 8-20
b r
Fa nonJDBC/ODP.NET FCF Benefits 8-21
Load Balancing Advisory 8-22
JDBC/ODP.NET Runtime Connection Load Balancing: Overview 8-23
Connection Load Balancing in RAC 8-24
Load Balancing Advisory: Summary 8-25
Monitor LBA FAN Events 8-26
FAN Release Map 8-27
Transparent Application Failover: Overview 8-28
TAF Basic Configuration Without FAN: Example 8-29
TAF Basic Configuration with FAN: Example 8-30
TAF Preconnect Configuration: Example 8-31
TAF Verification 8-32
FAN Connection Pools and TAF Considerations 8-33
Summary 8-34
x
9 Oracle Clusterware Administration
Objectives 9-2
Oracle Clusterware: Overview 9-3
Oracle Clusterware Run-Time View 9-4
Manually Control Oracle Clusterware Stack 9-6
CRS Resources 9-7
RAC Resources 9-8
Resource Attributes: Example 9-9
Main Voting Disk Function 9-11
Important CSS Parameters 9-13
Multiplexing Voting Disks 9-14
Change Voting Disk Configuration 9-15 a
Back Up and Recover Your Voting Disks 9-16 as
OCR Architecture 9-17 ฺbr )h
OCR Contents and Organization 9-19 c m
o deฺ
i ฺ
d Gui
Managing OCR Files and Locations: Overview 9-20
r e
Automatic OCR Backups 9-21
@ sic dent
Back Up OCR Manually 9-22
e r de Stu
v i lav e this
Recover OCR Using Physical Backups 9-23
ic i o_ o us
Recover OCR Using Logical Backups 9-24
( f abr nse t
Replace an OCR Mirror: Example 9-25
Repair OCR Configuration: Example 9-26
e rde le lice
OCR Considerations 9-27
l a V ab
i
V sfer
Change VIP Addresses 9-28
i o
ic -tran
Change Public/Interconnect IP Subnet Configuration: Example 9-30
b r
Fa non Third-Party Application Protection: Overview 9-31
Application VIP and RAC VIP Differences 9-32
Use CRS Framework: Overview 9-33
Use CRS Framework: Example 9-35
Summary 9-38
10 Diagnosing Oracle Clusterware and RAC Components

Objectives 10-2
One Golden Rule in RAC Debugging 10-3
Oracle Clusterware Main Log Files 10-4
Diagnostics Collection Script 10-5
Managing Diagnostic Data in RAC 10-6
Cluster Verify: Overview 10-7
Cluster Verify Stages 10-8
Cluster Verify Components 10-9
xi
Cluster Verify Locations 10-10
Cluster Verify Configuration File 10-11
Cluster Verify: Examples 10-13
Cluster Verify Output: Example 10-15
Summary 10-16
11 Node Addition and Removal

Objectives 11-2
Add and Delete Nodes and Instances: Overview 11-3
Main Steps to Add a Node to a RAC Cluster 11-4
Check Prerequisites Before Oracle Clusterware Installation 11-5 a
Add Oracle Clusterware to the New Node 11-6 as
Configure the New ONS 11-9 ฺbr )h
Add ASM Home to the New Node 11-10 c m
o deฺ
i ฺ
d Gui
Add RAC Home to the New Node 11-11 r e
Add a Listener to the New Node 11-12
@ sic dent
Add an ASM Instance to the New Node 11-13
e r de Stu
v i lav e this
Add a Database Instance to the New Node 11-15
ic i o_ o us
Add an Instance to Your RAC Database Using EM 11-18
( f abr nse t
Main Steps to Delete a Node from a RAC Cluster 11-21
Delete the Instance on the Node to Be Deleted 11-22
e rde le lice
Clean Up the ASM Instance 11-24
l a V ab
i
V sfer
Remove the Listener from the Node to Be Deleted 11-25
i o
ic -tran
Remove the Node from the Database 11-26
b r
Fa non
Remove the Node from ASM 11-27
Remove the Node from the Oracle Clusterware 11-28
Node Addition and Deletion and the SYSAUX Tablespace 11-30
Summary 11-31
12 Design for High Availability

Objectives 12-2
Causes of Unplanned Down Time 12-3
Causes of Planned Down Time 12-4
Oracle’s Solution to Down Time 12-5
RAC and Data Guard Complementarity 12-6
Maximum Availability Architecture 12-7
RAC and Data Guard Topologies 12-8
RAC and Data Guard Architecture 12-9
Data Guard Broker (DGB) and Oracle Clusterware (OC) Integration 12-11
xii
Fast-Start Failover: Overview 12-12
Data Guard Broker Configuration Files 12-14
Real-Time Query Physical Standby Database 12-15
Hardware Assisted Resilient Data 12-16
Oracle Clusterware Rolling Upgrade 12-17
Clustered ASM Rolling Upgrade 12-18
Patches and the RAC Environment 12-20
Inventory List Locks 12-21
OPatch Support for RAC: Overview 12-22
Rolling Patch Upgrade Using RAC 12-23
Download and Install Patch Updates 12-24
Rolling Release Upgrade Using SQL Apply 12-26
s a
Database High Availability: Best Practices 12-27
)h a
How Many ASM Disk Groups per Database 12-28
m ฺbr
Database Storage Consolidation 12-29
i ฺ c o deฺ
e d Gui
Which RAID Configuration for Best Availability? 12-30
r
Should You Use RAID 1 or RAID 5? 12-31
@ sic dent
e r de Stu
Should You Use ASM Mirroring Protection? 12-32
v i lav e this
What Type of Striping Works Best? 12-33
ASM Striping Only 12-34
ic i o_ o us
( f abr nse t
Hardware RAID–Striped LUNs 12-35
Hardware RAID–Striped LUNs HA 12-36
e rde le lice
It Is Real Simple 12-37
l a V ab
i
V sfer
Extended RAC: Overview 12-38
i o
ic -tran
Extended RAC Connectivity 12-39
b r
Fa non Extended RAC Disk Mirroring 12-40
Achieving Quorum with Extended RAC 12-41
ASM Preferred Mirror Read: Overview 12-42
ASM Preferred Mirror Read: Setup 12-43
Enterprise Manager ASM Configuration Page 12-44
ASM Preferred Mirror Read: Best Practice 12-45
Additional Data Guard Benefits 12-46
Using a Test Environment 12-47
Summary 12-48
Appendix A: Practices and Solutions
Appendix B: Miscellaneous Topics
Appendix C: Practice Environment Scripts
xiii
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Introduction
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l i ce© 2008, Oracle. All rights reserved.
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Overview
• This course is designed for anyone interested in

implementing a Real Application Clusters (RAC)
database.
• Although coverage is general, most of the examples
and labs in this course are Linux based.
• Knowledge of and experience with Oracle Database 11g
architecture are assumed. s a
h a
• Lecture material is supplemented with hands-on ฺ b r)
c o m eฺ
practices. di ฺ uid
s icre ent G
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
e l ic © 2008, Oracle. All rights reserved.
r lCopyright
l a V ab
i er
Overview io V nsf
b ric in-tthis
Theamaterial
ra course is designed to provide basic information that is needed to plan or
F n
no Database 11g for Real Application Clusters.
manage Oracle
The lessons and practices are designed to build on your knowledge of Oracle used in a
nonclustered environment. The material does not cover basic architecture and database
management: These topics are addressed by the Oracle Database 11g administration courses
offered by Oracle University. If your background does not include working with a current release
of the Oracle database, you should consider taking such training before attempting this course.
The practices provide an opportunity for you to work with the features of the database that are
unique to Real Application Clusters.
Oracle Database 11g: RAC Administration I - 2
Course Objectives
In this course, you:

• Learn the principal concepts of RAC
• Install the RAC components
• Administer database instances in a RAC and ASM
environment
• Manage services a
h as
• Back up and recover RAC databases r)
ฺ b
m eฺ
• Monitor and tune performance of a RAC database
ฺ c o
r e di Guid
• Administer Oracle Clusterware sic t
@ d e n
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Course Objectives
ic -tran
b r
Fa noisndesigned to give you the necessary information to successfully administer Real
This course
Application Clusters and Oracle Clusterware.
You install Oracle Database 11g with the Oracle Universal Installer (OUI) and create your
database with the Database Configuration Assistant (DBCA). This ensures that your RAC
environment has the optimal network configuration, database structure, and parameter settings for
the environment that you selected. As a DBA, after installation, your tasks are to administer your
RAC environment at three levels:
• Instance administration
• Database administration
• Cluster administration
Throughout this course, you use various tools to administer each level of RAC:
• Oracle Enterprise Manager 10g Grid Control to perform administrative tasks whenever
feasible
• Task-specific GUIs such as the Database Configuration Assistant (DBCA)
• Command-line tools such as SQL*Plus, Recovery Manager, Server Control (SRVCTL),
CLUVFY, CRSCTL, and OCRCONFIG
Typical Schedule
Topics Lessons Day
Introduction and installation I, 1, 2, 3 1
RAC administration and tuning 4, 5, 6 2
7, 8 3
Advanced topics 9, 10 4 s a
)h a
11, 12 5
m ฺbr
i ฺ c o deฺ
Workshop: Cloning 5ed
r u i
sic dent G
@
de Stu
e r
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Typical Schedule n
b ric in -this
Thealessons t r aguide are arranged in the order that you will probably study them in class, and
F non
are grouped into the topic areas that are shown in the slide. The individual lessons are ordered so
that they lead from more familiar to less familiar areas. The related practices are designed to let
you explore increasingly powerful features of a Real Application Clusters database.
In some cases, the goals for the lessons and goals for the practices are not completely compatible.
Your instructor may, therefore, choose to teach some material in a different order than found in
this guide. However, if your instructor teaches the class in the order in which the lessons are
printed in this guide, then the class should run approximately as shown in this schedule.
A History of Innovation
Automatic
Workload
Management
Automatic
Storage
Management Enterprise Grid
Control
Grids
RAC
Oracle s a
h a
Data Low-cost br)
Clusterware
m ฺ
Guard commodity
i ฺ c o d e ฺ
clusters red t Gu i
Resource
i
s denc
Nonblocking @
de Stu
manager e r
queries
v i lav e this
OPS
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
A History ofi oInnovation
ic -tran
a b
Oracle
rDatabase
F n o n 11g and the specific new manageability enhancements provided by Oracle RAC
11g enable RAC for everyone—all types of applications and enterprise grids (the basis for fourth-
generation computing). Enterprise grids are built from large configurations of standardized,
commodity-priced components: processors, network, and storage. With Oracle RAC’s cache
fusion technology, the Oracle database adds to this the highest levels of availability and
scalability.
Also, with Oracle RAC 11g, it becomes possible to perform dynamic provisioning of nodes,
storage, CPUs, and memory to maintain service levels more easily and efficiently.
Enterprise grids are the data centers of the future and enable business to be adaptive, proactive,
and agile for the fourth generation.
The next major transition in computing infrastructure is going from the era of big symmetric
multiprocessing (SMP) models to the era of grids.
What Is a Cluster?
• Interconnected nodes
act as a single server.
Private Interconnect
• Cluster software
hides the structure. Node
Public Public Public
• Disks are available network network network
for read and
write by all nodes. s a
)h a
ฺbr
Public
• Operating system network
m
is the same on each Clusterware i ฺ c o deฺ
r e d Gui
machine. on each node
@ sic dent
e r de Stu
v i lav e this Disks
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
What Is a Cluster?
ic -tran
b r
Fa nconsists
A cluster
on of two or more independent, but interconnected, servers. Several hardware
vendors have provided cluster capability over the years to meet a variety of needs. Some clusters
were intended only to provide high availability by allowing work to be transferred to a secondary
node if the active node fails. Others were designed to provide scalability by allowing user
connections or work to be distributed across the nodes.
Another common feature of a cluster is that it should appear to an application as if it were a single
server. Similarly, management of several servers should be as similar to the management of a
single server as possible. The cluster management software provides this transparency.
For the nodes to act as if they were a single server, files must be stored in such a way that they can
be found by the specific node that needs them. There are several different cluster topologies that
address the data access issue, each dependent on the primary goals of the cluster designer.
The interconnect is a physical network used as a means of communication between each node of
the cluster.
In short, a cluster is a group of independent servers that cooperate as a single system.
Note: The clusters you are going to manipulate in this course all have the same operating system.
This is a requirement for RAC clusters.
Oracle Real Application Clusters
• Multiple instances
accessing the same
Interconnect
database
• One instance
per node
Shared
• Physical or
logical access
cache
s a
)h a
to each
m ฺbr
database file i ฺ c o Instances
d e ฺ
red t Guspreadi
• Software-controlled i c
s den across nodes
data access @
de Stu
e r
v i lav e thisDatabase
ic i o_ o us files
( f abr nse t
e rde lCopyright
l a V ab e
i
V sferClusters
Oracle Real i oApplication
ic -tran
b r
Fa non Clusters is a software that enables you to use clustered hardware by running
Real Application
multiple instances against the same database. The database files are stored on disks that are either
physically or logically connected to each node, so that every active instance can read from or
write to them.
The Real Application Clusters software manages data access, so that changes are coordinated
between the instances and each instance sees a consistent image of the database. The cluster
interconnect enables instances to pass coordination information and data images between each
other.
This architecture enables users and applications to benefit from the processing power of multiple
machines. RAC architecture also achieves redundancy in the case of, for example, a system
crashing or becoming unavailable; the application can still access the database on any surviving
instances.
Benefits of Using RAC
• High availability: Surviving node and instance failures

• Scalability: Adding more nodes as you need them in
the future
• Pay as you grow: Paying for only what you need today
• Key grid computing features:
– Growth and shrinkage on demand a
h as
– Single-button addition of servers r)
– Automatic workload management for services om ฺ b
ฺ iฺc uide
e d
s icr ent G
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V RAC f er
Benefits ofioUsing
ic -tran s
b r
Fa noApplication
Oracle Real n Clusters (RAC) enables high utilization of a cluster of standard, low-cost
modular servers such as blades.
RAC offers automatic workload management for services. Services are groups or classifications of
applications that comprise business components corresponding to application workloads. Services
in RAC enable continuous, uninterrupted database operations and provide support for multiple
services on multiple instances. You assign services to run on one or more instances, and alternate
instances can serve as backup instances. If a primary instance fails, the Oracle server moves the
services from the failed instance to a surviving alternate instance. The Oracle server also
automatically load-balances connections across instances hosting a service.
RAC harnesses the power of multiple low-cost computers to serve as a single large computer for
database processing, and provides the only viable alternative to large-scale symmetric
multiprocessing (SMP) for all types of applications.
RAC, which is based on a shared-disk architecture, can grow and shrink on demand without the
need to artificially partition data among the servers of your cluster. RAC also offers a single-
button addition of servers to a cluster. Thus, you can easily provide or remove a server to or from
the database.
Clusters and Scalability
SMP model RAC model
Shared
Memory
storage
s a
Cache Cache )h a
ฺbr
SGA SGA
c m
o deฺ
i ฺ i
CPU CPU CPU CPU BGP BGP red BGPGBGP u
@ sic dent
Cache coherency
e r de Cache S tufusion
v i lav e this BGP (background process)
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Clusters andi o Scalability
ic -tran
b r
Fa non scales transparently on SMP machines, then it is realistic to expect it to scale
If your application
well on RAC, without having to make any changes to the application code.
RAC eliminates the database instance, and the node itself, as a single point of failure, and ensures
database integrity in the case of such failures.
Following are some scalability examples:
• Allow more simultaneous batch processes.
• Allow larger degrees of parallelism and more parallel executions to occur.
• Allow large increases in the number of connected users in online transaction processing
(OLTP) systems.
Note: What is true for SMP is also true for Non-Uniform Memory Architecture (NUMA)
architectures. NUMA architectures are the logical next step in scaling from SMP architectures.
Levels of Scalability
• Hardware: Disk input/output (I/O)

• Internode communication: High bandwidth and low
latency
• Operating system: Number of CPUs
• Database management system: Synchronization
• Application: Design a
as
)h ฺbr
c m
o deฺ
i ฺ
d Gui
r e
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Levels of Scalability
ic -tran
b r
Fa noimplementation
Successful n of cluster databases requires optimal scalability on four levels:
• Hardware scalability: Interconnectivity is the key to hardware scalability, which greatly
depends on high bandwidth and low latency.
• Operating system scalability: Methods of synchronization in the operating system can
determine the scalability of the system. In some cases, potential scalability of the hardware is
lost because of the operating system’s inability to handle multiple resource requests
simultaneously.
• Database management system scalability: A key factor in parallel architectures is whether
the parallelism is affected internally or by external processes. The answer to this question
affects the synchronization mechanism.
• Application scalability: Applications must be specifically designed to be scalable. A
bottleneck occurs in systems in which every session is updating the same data most of the
time. Note that this is not RAC specific and is true on single-instance system too.
It is important to remember that if any of the areas above are not scalable (no matter how scalable
the other areas are), then parallel cluster processing may not be successful. A typical cause for the
lack of scalability is one common shared resource that must be accessed often. This causes the
otherwise parallel operations to serialize on this bottleneck. A high latency in the synchronization
increases the cost of synchronization, thereby counteracting the benefits of parallelization. This is
a general limitation and not a RAC-specific limitation.
Scaleup and Speedup
Original system
Hardware Time 100% of task
Cluster system scaleup Cluster system speedup
s a
Hardware Time Up to
)h a
200% Hardware
m ฺbr
of Up to
i ฺ c o 100%d e ฺ
Hardware task 300%
r i
ed t Guof task
Time of
Hardware i c
s dTime/2 n
task @ e
e r de Stu
Hardware
Time v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Scaleup and i o Speedup
ic -tran
b r
a is othenability to sustain the same performance levels (response time) when both workload
Scaleup
F n increase proportionally:
and resources
Scaleup = (volume parallel) / (volume original)
For example, if 30 users consume close to 100 percent of the CPU during normal processing, then
adding more users would cause the system to slow down due to contention for limited CPU
cycles. However, by adding CPUs, you can support extra users without degrading performance.
Speedup is the effect of applying an increasing number of resources to a fixed amount of work to
achieve a proportional reduction in execution times:
Speedup = (time original) / (time parallel)
Speedup results in resource availability for other tasks. For example, if queries usually take ten
minutes to process and running in parallel reduces the time to five minutes, then additional queries
can run without introducing the contention that might occur were they to run concurrently.
Speedup/Scaleup and Workloads
Workload Speedup Scaleup
OLTP and Internet No Yes
DSS with parallel query Yes Yes

s a
Batch (mixed) Possible Yes
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V and f r
eWorkloads
i o
Speedup/Scaleup s
an determines whether scaleup or speedup capabilities can be achieved using
b ricof workload
Theatype - t r
F non
parallel processing.
Online transaction processing (OLTP) and Internet application environments are characterized by
short transactions that cannot be further broken down and, therefore, no speedup can be achieved.
However, by deploying greater amounts of resources, a larger volume of transactions can be
supported without compromising the response.
Decision support systems (DSS) and parallel query options can attain speedup, as well as scaleup,
because they essentially support large tasks without conflicting demands on resources. The
parallel query capability within the Oracle database can also be leveraged to decrease overall
processing time of long-running queries and to increase the number of such queries that can be run
concurrently.
In an environment with a mixed workload of DSS, OLTP, and reporting applications, scaleup can
be achieved by running different programs on different hardware. Speedup is possible in a batch
environment, but may involve rewriting programs to use the parallel processing capabilities.
I/O Throughput Balanced: Example
Each machine has 2 CPUs:

2 ° 200 MB/s ° 4 = 1600 MB/s
Each machine has 2 HBAs:

HBA1
HBA2
HBA1
HBA2
HBA1
HBA2
HBA2
HBA1
8 ° 200 MB/s = 1600 MB/s
Each switch needs to support 800 MB/s

to guarantee a total system throughput s a
FC-switch
of 1600 MB/s. )h a
m ฺbr
i ฺ c o deฺ
r e d hasGone
Each
u i array
disk
Disk icDisk t 2-Gbit
Disk Disk Disk Disk Disk Disk
s n
e@7 Stude 8 ° 200 MB/s =
array 1 array 2 array 3 array 4 array 5 array 6 array array 8 controller:
r d
i l a ve this 1600 MB/s
v
o_ o us e
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
V Balanced: f er Example
i
I/O Throughputo s
aa nsystem delivers the IO demand that is required, all system components on the
Toa b ricsure that
make - t r
F non
IO path need to be orchestrated to work together.
The weakest link determines the IO throughput.
On the left, you see a high-level picture of a system. This is a system with four nodes, two Host
Bus Adapters (HBAs) per node, two fibre channel switches, which are attached to four disk arrays
each. The components on the IO path are the HBAs, cables, switches, and disk arrays.
Performance depends on the number and speed of the HBAs, switch speed, controller quantity,
and speed of disks. If any one of these components is undersized, the system throughput is
determined by this component. Assuming you have a 2-GBit HBA, the nodes can read about
8 ° 200 MB/s = 1.6 GBytes/s. However, assuming that each disk array has one controller, all 8
arrays can also do 8 ° 200 MB/s = 1.6 GBytes/s. Therefore, each of the fibre channel switches
also need to deliver at least 2 GBit/s per port, to a total of 800 MB/s throughput. The two switches
will then deliver the needed 1.6 GBytes/s.
Note: When sizing a system, also take the system limits into consideration. For instance, the
number of bus slots per node is limited and may need to be shared between HBAs and network
cards. In some cases, dual port cards exist if the number of slots is exhausted. The number of
HBAs per node determines the maximal number of fibre channel switches. And the total number
of ports on a switch limits the number of HBAs and disk controllers.
Performance of Typical Components
Throughput Performance
Component Theory (Bit/s) Maximal Byte/s
HBA ½ Gbit/s 100/200 Mbytes/s
16 Port Switch 8 ° 2 Gbit/s 1600 Mbytes/s
Fibre Channel 2 Gbit/s 200 Mbytes/s
s a
Disk Controller 2 Gbit/s 200 Mbytes/s )h a
m ฺbr
GigE NIC 1 Gbit/s 80 Mbytes/s i ฺ c o deฺ
r e d Gui
Infiniband 10 Gbit/s
@890sicMbytes/s
e n t
r d e Stud
CPU
i l a ve thi200–250 s MB/s
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Typical f r
eComponents
Performancei o of
ic -tran s
b r
Fa non people often confuse bits with bytes. This confusion originates mainly from the
While discussing,
fact that hardware vendors tend to describe component’s performance in bits/s whereas database
vendors and customers describe their performance requirements in bytes/s.
The following is a list of common hardware components with their theoretical performance in
bits/second and typical performance in bytes/second:
• HBAs come in 1 or 2 GBit per second with a typical throughput of 100 or 200 MB/s.
• A 16 Port Switch comes with sixteen 2-GBit ports. However, the total throughput is 8 times 2
Gbit, which results in 1600 Mbytes/s.
• Fibre Channel cables have a 2-GBit/s throughput, which translates into 200 MB/s.
• Disk Controllers come in 2-GBit/s throughput, which translates into about 200 MB/s.
• GigE has a typical performance of about 80 MB/s whereas Infiniband delivers about
160 MB/s.
Complete Integrated Clusterware
9i RAC 11g Oracle Clusterware
Applications
Applications/RAC
Services framework
System Management
Cluster control
Management APIs
Event Services
Event Services
Cluster control/Recovery APIs
Volume Manager
file system Automatic Storage Management
Messaging and locking s a
Messaging and locking
)h a
Membership m ฺ br
Membership o
i ฺ c d e ฺ
red t Gu i
Connectivity i c
Connectivity
s den
@
e Stu
dHardware/OS
Hardware/OS kernel e
v thisr kernel
i l a
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V sClusterware
f
i o
Complete Integrated
ic -tran
b r
Fa non Real Application Clusters in Oracle9i Database. For the first time, you were
Oracle introduced
able to run online transaction processing (OLTP) and decision support system (DSS) applications
against a database cluster without having to make expensive code changes or spend large amounts
of valuable administrator time partitioning and repartitioning the database to achieve good
performance.
Although Oracle9i Real Application Clusters did much to ease the task of allowing applications to
work in clusters, there are still support challenges and limitations. Among these cluster challenges
are complex software environments, support, inconsistent features across platforms, and awkward
management interaction across the software stack. Most clustering solutions today were designed
with failover in mind. Failover clustering has additional systems standing by in case of a failure.
During normal operations, these failover resources may sit idle.
With the release of Oracle Database 11g, Oracle provides you with an integrated software solution
that addresses cluster management, event management, application management, connection
management, storage management, load balancing, and availability. These capabilities are
addressed while hiding the complexity through simple-to-use management tools and automation.
Oracle Clusterware provides an integrated clusterware layer that delivers a complete environment
for applications.
Necessity of Global Resources
SGA1 SGA2 SGA1 SGA2

1008
1008 1008
1 2
s a
)h a
SGA1 SGA2 SGA1 SGA2
m ฺbr
1009 1008 1009
i ฺ c o deฺ
r e d Gui
@ sic dent
Lost
updates! e r de Stu
1008 i lav e th1008 is
v
o_4 o us
r ic i 3
ab nse t
( f
e rde lCopyright
l a V ab e
i
V sResources
f er
Necessity iofo Global
ic -tran
b r
Fa non environments, locking coordinates access to a common resource such as a row
In single-instance
in a table. Locking prevents two processes from changing the same resource (or row) at the same
time.
In RAC environments, internode synchronization is critical because it maintains proper
coordination between processes on different nodes, preventing them from changing the same
resource at the same time. Internode synchronization guarantees that each instance sees the most
recent version of a block in its buffer cache.
Note: The slide shows you what would happen in the absence of cache coordination. RAC
prohibits this problem.
Global Resources Coordination
Cluster
Node1 Noden
Instance1 Instancen
GRD Master Cache GRD Master Cache
… Global …
LMON GES resources GES LMON
LMD0 LMD0
LMSx GCS GCS LMSx
LCK0 Interconnect LCK0
DIAG DIAG
s a
)h a
Global Resource Directory (GRD)
m ฺbr
i ฺ c o deฺ
Global Cache Services (GCS) Global Enqueue Services
r e d (GES) u i
sic dent G
@
de Stu
e r
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V Coordination
f
i o
Global Resources
ic -tran s
b r
Fa operations
Cluster
non
require synchronization among all instances to control shared access to
resources. RAC uses the Global Resource Directory (GRD) to record information about how
resources are used within a cluster database. The Global Cache Services (GCS) and Global
Enqueue Services (GES) manage the information in the GRD.
Each instance maintains a part of the GRD in its System Global Area (SGA). The GCS and GES
nominate one instance to manage all information about a particular resource. This instance is
called the resource master. Also, each instance knows which instance masters which resource.
Maintaining cache coherency is an important part of a RAC activity. Cache coherency is the
technique of keeping multiple copies of a block consistent between different Oracle instances.
GCS implements cache coherency by using what is called the Cache Fusion algorithm.
The GES manages all non–Cache Fusion interinstance resource operations and tracks the status of
all Oracle enqueuing mechanisms. The primary resources of the GES controls are dictionary cache
locks and library cache locks. The GES also performs deadlock detection to all deadlock-sensitive
enqueues and resources.
Global Cache Coordination: Example
Cluster
Node1 Node2
Instance1 Instance2
Cache 1009 1009 Cache
… 3 …
LMON LMON
LMD0 LMD0
LMSx 4 LMSx
LCK0 LCK0
DIAG DIAG
s a
2
Instance 2 has
1
)h a
Block mastered the current version of the block.
ฺ br
Which instance
m
by instance 1 GCS o deฺ
masters the block?
i ฺ c
r e d Gui
@ sic dent
e r de Stu
v i
1008lav e this No disk I/O
o _ s
b r ici to u
e (fa ense
e e l ic © 2008, Oracle. All rights reserved.
rd lCopyright
l a V ab
i
VCoordination:
f er Example
i
Global Cacheo s
an in the slide assumes that the data block has been changed, or dirtied, by
b ric described
Theascenario - t r
F non
the first instance. Furthermore, only one copy of the block exists clusterwide, and the content of
the block is represented by its SCN.
1. The second instance attempting to modify the block submits a request to the GCS.
2. The GCS transmits the request to the holder. In this case, the first instance is the holder.
3. The first instance receives the message and sends the block to the second instance. The first
instance retains the dirty buffer for recovery purposes. This dirty image of the block is also
called a past image of the block. A past image block cannot be modified further.
4. On receipt of the block, the second instance informs the GCS that it holds the block.
Note: The data block is not written to disk before the resource is granted to the second instance.
Write to Disk Coordination: Example
Cluster
Node1 Node2
Instance1 Instance2
Cache 1009 1010 Cache
… 3 …
LMON LMON
LMD0 LMD0
LMSx LMSx
LCK0 5 4 LCK0
DIAG DIAG
s a
1
Block flushed, make room
)h
2 a
Need to make room
in my cache. ฺ br
Instance 2 owns it.
m
Who has the current version GCS
i ฺ c o deฺ
Instance 2, flush the block
of that block?
r e d Gui
to disk.
@ sic dent
e r de StuOnly one
v i
1010lav e this disk I/O
_
io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
V sfer Example
i
Write to Disko Coordination:
an in the slide illustrates how an instance can perform a checkpoint at any
b ric described
Theascenario - t r
F non
time or replace buffers in the cache as a response to free buffer requests. Because multiple
versions of the same data block with different changes can exist in the caches of instances in the
cluster, a write protocol managed by the GCS ensures that only the most current version of the
data is written to disk. It must also ensure that all previous versions are purged from the other
caches. A write request for a data block can originate in any instance that has the current or past
image of the block. In this scenario, assume that the first instance holding a past image buffer
requests that the Oracle server writes the buffer to disk:
1. The first instance sends a write request to the GCS.
2. The GCS forwards the request to the second instance, which is the holder of the current
version of the block.
3. The second instance receives the write request and writes the block to disk.
4. The second instance records the completion of the write operation with the GCS.
5. After receipt of the notification, the GCS orders all past image holders to discard their past
images. These past images are no longer needed for recovery.
Note: In this case, only one I/O is performed to write the most current version of the block to disk.
Dynamic Reconfiguration
Reconfiguration remastering
Node1 Node2 Node3

Instance1 Instance2 Instance3
masters granted masters granted masters granted
R1 1, 2, 3 R3 2, 3 R5 2
R2 1, 3 R4 1, 2 R6 1, 2, 3
s a
)h a
Node1 Node2 Node3
m ฺbr
Instance1 Instance2 Instance3 i ฺ c o deฺ
r e d Gui
masters granted
R1 1, 3
masters granted
R3 2, 3 R5
sic dent
masters granted
@
R2 1, 3 R4 1, 2 rde R6 S1, t3u
R3 3
a e
v this 1 R4
i l
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
i o
Dynamic Reconfiguration
ic -tran
b r
Fa noinstance
When one n departs the cluster, the GRD portion of that instance needs to be redistributed
to the surviving nodes. Similarly, when a new instance enters the cluster, the GRD portions of the
existing instances must be redistributed to create the GRD portion of the new instance.
Instead of remastering all resources across all nodes, RAC uses an algorithm called lazy
remastering to remaster only a minimal number of resources during a reconfiguration. This is
illustrated on the slide. For each instance, a subset of the GRD being mastered is shown along
with the names of the instances to which the resources are currently granted. When the second
instance fails, its resources are remastered on the surviving instances. As the resources are
remastered, they are cleared of any reference to the failed instance.
Object Affinity and Dynamic Remastering
Messages are sent to remote node when reading into cache.
Node1 Node2
Instance1 Before
☺
☺ GCS message to master dynamic
☺ remastering
☺
Instance2
Object
Read from
disk
s a
)h a
Node2
m ฺbr
i ฺ c o deฺ
☺ ☺
r e d Gui
s ic ent After
☺ ☺
d e @ tud dynamic
remastering
Instance1 r
ve this Instance2S
Node1 i
v sel a
i o _remote u when reading into cache.
b r ic
No messages are sent to
t o node
e (fa ense
e rd lCopyright
l a V ab
i er Remastering
Vand Dynamic
f
i o
Object Affinity
ic -tran s
b r
Fa notondynamic resource reconfiguration, the GCS, which is tightly integrated with the
In addition
buffer cache, enables the database to automatically adapt and migrate resources in the GRD. This
is called dynamic remastering. The basic idea is to master a buffer cache resource on the instance
where it is mostly accessed. In order to determine whether dynamic remastering is necessary, the
GCS essentially keeps track of the number of GCS requests on a per-instance and per-object basis.
This means that if an instance, compared to another, is heavily accessing blocks from the same
object, the GCS can take the decision to dynamically migrate all of that object’s resources to the
instance that is accessing the object most.
The upper part of the graphic shows you the situation where the same object has master resources
spread over different instances. In that case, each time an instance needs to read a block from that
object whose master is on the other instance, the reading instance must send a message to the
resource’s master to ask permission to use the block.
The lower part of the graphic shows you the situation after dynamic remastering occurred. In this
case, blocks from the object have affinity to the reading instance which no longer needs to send
GCS messages across the interconnect to ask for access permissions.
Note: The system automatically moves mastership of undo segment objects to the instance that
owns the undo segments.
Global Dynamic Performance Views
• Retrieve information about all started instances

• Have one global view for each local view
• Use one parallel slave on each instance
Cluster s a
)h a
Node1
Instance1
GV$INSTANCE
Noden
Instancen m ฺbr
i ฺ c o deฺ
r e d Gui
V$INSTANCE c
siV$INSTANCE n t
@
e Stud e
r d
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Performance
f er
i c i
Global Dynamic o a n s Views
r r
-tperformance views retrieve information about all started instances accessing one
Fab dynamic
Global
n o n
RAC database. In contrast, standard dynamic performance views retrieve information about the
local instance only.
For each of the V$ views available, there is a corresponding GV$ view except for a few
exceptions. In addition to the V$ information, each GV$ view possesses an additional column
named INST_ID. The INST_ID column displays the instance number from which the associated
V$ view information is obtained. You can query GV$ views from any started instance.
GV$ views use a special form of parallel execution. The parallel execution coordinator runs on the
instance that the client connects to, and one slave is allocated in each instance to query the
underlying V$ view for that instance.
Additional Memory Requirement for RAC
• Heuristics for scalability cases:

– 15% more shared pool
– 10% more buffer cache
• Smaller buffer cache per instance in the case of
single-instance workload distributed across multiple
instances
a
• Current values: has r)
ฺ b
m eฺ
SELECT resource_name,
current_utilization,max_utilization di ฺ c o
r e u id
FROM v$resource_limit s ic ent G
WHERE resource_name like 'g_s%'; e
d @ t ud
r
ve this S
SELECT * FROM v$sgastat i l
v likea
WHERE name like 'g_s%' or i o _name u s e 'KCL%';
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Requirement
f er
i c o
AdditionaliMemory
a n s for RAC
r
ab onmemory
RAC-specific -t r is mostly allocated in the shared pool at SGA creation time. Because
F n
blocks may be cached across instances, you must also account for bigger buffer caches.
Therefore, when migrating your Oracle database from single instance to RAC, keeping the
workload requirements per instance the same as with the single-instance case, about 10% more
buffer cache and 15% more shared pool are needed to run on RAC. These values are heuristics,
based on RAC sizing experience. However, these values are mostly upper bounds.
If you use the recommended automatic memory management feature as a starting point, then you
can reflect these values in your SGA_TARGET initialization parameter.
However, consider that memory requirements per instance are reduced when the same user
population is distributed over multiple nodes.
Actual resource usage can be monitored by querying the CURRENT_UTILIZATION and
MAX_UTILIZATION columns for the Global Cache Services (GCS) and Global Enqueue
Services (GES) entries in the V$RESOURCE_LIMIT view of each instance. You can monitor the
exact RAC memory resource usage of the shared pool by querying V$SGASTAT as shown in the
slide.
Efficient Internode Row-Level Locking
UPDATE UPDATE
1 2
Node1 Node2 Node1 Node2

Instance1 Instance2 Instance1 Instance2
s a
COMMIT
No block-level
UPDATE
)h a
lock
m ฺbr
Node1 Node2 Node1 i ฺ c
Node2
o deฺ
r e d Gui
Instance1ic
Instance1 Instance2
@ nt
s deInstance2
r d e Stu 3
4
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Row-Level f er
i c i o
Efficient Internode
a n s Locking
r r
-tefficient row-level locks. These row-level locks are created when data
Fab supports
Oracle
n o n
manipulation language (DML) operations, such as UPDATE, are executed by an application.
These locks are held until the application commits or rolls back the transaction. Any other
application process will be blocked if it requests a lock on the same row.
Cache Fusion block transfers operate independently of these user-visible row-level locks. The
transfer of data blocks by the GCS is a low-level process that can occur without waiting for row-
level locks to be released. Blocks may be transferred from one instance to another while row-level
locks are held.
GCS provides access to data blocks allowing multiple transactions to proceed in parallel.
Parallel Execution with RAC
Execution slaves have node affinity with the execution

coordinator but will expand if needed.
Node 1 Node 2 Node 3 Node 4
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
Execution
s ic encoordinator
t
Shared disks d @
e Stud Parallel
r
ve this execution
i
v sel a server
o _
b r ici to u
e (fa ense
rd lCopyright
l a V ab
i
V with f r
eRAC
i o
Parallel Execution
ic -tran s
b r
Fa ncost-based
Oracle’s
on optimizer incorporates parallel execution considerations as a fundamental
component in arriving at optimal execution plans.
In a RAC environment, intelligent decisions are made with regard to intranode and internode
parallelism. For example, if a particular query requires six query processes to complete the work
and six parallel execution slaves are idle on the local node (the node that the user connected to),
then the query is processed by using only local resources. This demonstrates efficient intranode
parallelism and eliminates the query coordination overhead across multiple nodes. However, if
there are only two parallel execution servers available on the local node, then those two and four
of another node are used to process the query. In this manner, both internode and intranode
parallelism are used to speed up query operations.
In real-world decision support applications, queries are not perfectly partitioned across the various
query servers. Therefore, some parallel execution servers complete their processing and become
idle sooner than others. The Oracle parallel execution technology dynamically detects idle
processes and assigns work to these idle processes from the queue tables of the overloaded
processes. In this way, the Oracle server efficiently redistributes the query workload across all
processes. Real Application Clusters further extends these efficiencies to clusters by enabling the
redistribution of work across all the parallel execution slaves of a cluster.
RAC Software Principles
Cluster
Node1 Noden
Instance1 Instancen
Cache Cache
… Global …
LMON resources LMON
LMD0 LMD0
LMSx LMSx
LCK0 LCK0
DIAG DIAG
Oracle Clusterware Oracle Clusterware s a

)h a
CRSD & RACGIMON Cluster CRSD & RACGIMON
m ฺ br
EVMD interface EVMD o
i ฺ c d e ฺ
OCSSD & OPROCD OCSSD re&d OPROCD u i
i c
s Applicationsn t G
Applications Global @ d e
ASM, DB, Services, OCR management: e r de ASM,SDB, tu Services, OCR
VIP, ONS, EMD, Listener i av EMthiVIP,
lDBCA, s ONS, EMD, Listener
SRVCTL, v
o_ o us e
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
VPrinciplesf er
i
RAC Software o
ic -tran s
b r
Fa nsee
You may
ona few additional background processes associated with a RAC instance than you
would with a single-instance database. These processes are primarily used to maintain database
coherency among each instance. They manage what is called the global resources:
• LMON: Global Enqueue Service Monitor
• LMD0: Global Enqueue Service Daemon
• LMSx: Global Cache Service Processes, where x can range from 0 to j
• LCK0: Lock process
• DIAG: Diagnosability process
At the cluster level, you find the main processes of Oracle Clusterware. They provide a standard
cluster interface on all platforms and perform high-availability operations. You find these
processes on each node of the cluster:
• CRSD and RACGIMON: Are engines for high-availability operations
• OCSSD: Provides access to node membership and group services
• EVMD: Scans callout directory and invokes callouts in reactions to detected events
• OPROCD: Is a process monitor for the cluster
There are also several tools that are used to manage the various resources available on the cluster
at a global level. These resources are the Automatic Storage Management (ASM) instances, the
RAC databases, the services, and node applications. Some of the tools that you use throughout this
course arereproduction
Unauthorized Server Controlor(SRVCTL), DBCA,
distribution and Enterprise
prohibitedฺ Manager.
Copyright© 2009, Oracle and/or its affiliatesฺ
RAC Software Storage Principles
Node1 Noden Node1 Noden

Instance1 … Instancen Instance1 … Instancen
CRS_HOME CRS_HOME
ORACLE_HOME ORACLE_HOME CRS_HOME CRS_HOME
ASM_HOME ASM_HOME
Local storage Local storage Local storage Local storage
Voting files Voting files

OCR files OCR files
s a
)h a
Shared storage ORACLE_HOME
m ฺ br
ASM_HOME o
i ฺ c d e ฺ
red t Gu
Shared storage i
i
s denc
Permits rolling patch upgrades
@
de Stu
e r
Software is not a single
v i lav e this
point of failure.
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VStorage f r
ePrinciples
i
RAC Software o
ic -tran s
b r
Fa noDatabase
The Oracle n 11g Real Application Clusters installation is a two-phase installation. In the
first phase, you install Oracle Clusterware. In the second phase, you install the Oracle database
software with RAC components and create a cluster database. The Oracle Home that you use for
Oracle Clusterware must be different from the one that is used for the RAC software. Although it
is possible to install the RAC software on your cluster shared storage when using certain cluster
file systems, software is usually installed on a regular file system that is local to each node. This
permits rolling patch upgrades and eliminates the software as a single point of failure. In addition,
at least two files must be stored on your shared storage:
• The monitoring information across the cluster. Its size is set to around 20 MB.
• The Oracle Cluster Registry (OCR) file is also a key component of Oracle Clusterware. It
maintains information about the high-availability components in your cluster, such as the
voting file is essentially used by the Cluster Synchronization Services daemon for node-
cluster node list, cluster database instance to node mapping, and CRS application resource
profiles (such as services, Virtual Interconnect Protocol addresses, and so on). This file is
maintained by administrative tools such as SRVCTL. Its size is around 100 MB.
The voting and OCR files cannot be stored in ASM because they must be accessible before
starting any Oracle instance. OCR and voting files can be on redundant, reliable storage such as
RAID, or mirrored on different disks. The recommended best practice location for those files is
raw devices
Unauthorized on the fastestorI/O
reproduction disks.
distribution prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
RAC Database Storage Principles
Node1 Noden
…
Instance1 Instancen
Archived Archived
log files log files
Local storage Local storage
Data files a
Undo tablespace Undo tablespace s
files for Temp files files for
)h a
instance1 Control files instancen
m ฺbr
Flash recovery area files
i ฺ c o deฺ
Online Change tracking file r e dOnline u i
redo log files SPFILE sic redo n log
G
t files
for instance1 @
e Stufor e
d instancen
TDE Wallet r d
Shared i l a ve this
storage
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
VStorage f r
ePrinciples
i
RAC Databaseo s
an between RAC storage and storage for single-instance Oracle databases is
b ric difference
Theaprimary - t r
F non
that all data files in RAC must reside on shared devices (either raw devices or cluster file systems)
in order to be shared by all the instances that access the same database. You must also create at
least two redo log groups for each instance, and all the redo log groups must also be stored on
shared devices for instance or crash recovery purposes. Each instance’s online redo log groups are
called an instance’s thread of online redo.
In addition, you must create one shared undo tablespace for each instance for using the
recommended automatic undo management feature. Each instance’s undo tablespace must be
shared by all other instances for recovery purposes.
Archive logs cannot be placed on raw devices because their names are automatically generated
and are different for each archive log. That is why they must be stored on a file system. If you use
a cluster file system (CFS), it enables you to access these archive files from any node at any time.
If you do not use a CFS, you are always forced to make the archives available to the other cluster
members at the time of recovery—for example, by using a network file system (NFS) across
nodes. If you are using the recommended flash recovery area feature, then it must be stored in a
shared directory so that all instances can access it.
Note: A shared directory can be an ASM disk group, or a cluster file system.
RAC and Shared Storage Technologies
• Storage is a critical component of grids:

– Sharing storage is fundamental.
– New technology trends
• Supported shared storage for Oracle grids:
– Network Attached Storage
– Storage Area Network a
as
• Supported file storage for Oracle grids: )h ฺbr
– Raw volumes
c m
o deฺ
– Cluster file system i ฺ
d Gui
r e
– ASM @ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Storagef er Technologies
i o
RAC and Shared
ic -tran s
b r
Fa isnaocritical
Storage n component of any grid solution. Traditionally, storage has been directly
attached to each individual server (Directly Attached Storage, or DAS). Over the past few years,
more flexible storage, which is accessible over storage area networks or regular Ethernet
networks, has become popular. These new storage options enable multiple servers to access the
same set of disks, simplifying the provisioning of storage in any distributed environment.
Storage Area Network (SAN) represents the evolution of data storage technology to this point.
Traditionally, on client server systems, data was stored on devices either inside or directly
attached to the server. Next in the evolutionary scale came Network Attached Storage (NAS) that
took the storage devices away from the server and connected them directly to the network. SANs
take the principle a step further by allowing storage devices to exist on their own separate
networks and communicate directly with each other over very fast media. Users can gain access to
these storage devices through server systems that are connected to both the local area network
(LAN) and SAN.
The choice of file system is critical for RAC deployment. Traditional file systems do not support
simultaneous mounting by more than one system. Therefore, you must store files in either raw
volumes without any file system, or on a file system that supports concurrent access by multiple
systems.
RAC and Shared Storage Technologies (continued)
Thus, three major approaches exist for providing the shared storage needed by RAC:
• Raw volumes: These are directly attached raw devices that require storage that operates in
block mode such as fiber channel or iSCSI.
• Cluster file system: One or more cluster file systems can be used to hold all RAC files.
Cluster file systems require block mode storage such as fiber channel or iSCSI.
• Automatic Storage Management (ASM): It is a portable, dedicated, and optimized cluster
file system for Oracle database files.
Note: iSCSI is important to SAN technology because it enables a SAN to be deployed in a local
area network (LAN), wide area network (WAN), or Metropolitan Area Network (MAN).
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Oracle Cluster File System
• Is a shared disk cluster file system for Linux and

Windows
• Improves management of data for RAC by eliminating
the need to manage raw devices
• Provides open solution on the operating system side
• Can be downloaded from OTN: a
as
– http://oss.oracle.com/projects/ocfs2/ (Linux)
)h ฺbr
c m
o deฺ
i ฺ
d Gui
r e
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
VFile System
f
i
Oracle Clustero
ic -tran s
b r
Fa non File System (OCFS) is a shared file system designed specifically for Oracle Real
Oracle Cluster
Application Clusters. OCFS eliminates the requirement that Oracle database files be linked to
logical drives and enables all nodes to share a single Oracle Home (on Windows 2000 and 2003
only), instead of requiring each node to have its own local copy. OCFS volumes can span one
shared disk or multiple shared disks for redundancy and performance enhancements. The
following is a list of files that can be placed on Oracle Cluster File System version 1:
• Oracle software installation: Currently, this configuration is supported only on Windows
2000 and 2003. Oracle Cluster File System 2 1.2.1 provides support for Oracle Home on
Linux as well.
• Oracle files (control files, data files, redo logs, bfiles, and so on)
• Shared configuration files (spfile)
• Files created by Oracle during run time
• Voting and OCR files
Oracle Cluster File System is free for developers and customers. The source code is provided
under the General Public License (GPL) on Linux. It can be downloaded from the Oracle
Technology Network Web site.
Note: From OTN, you can specifically download OCFS for Linux. However, when you download
the database software for Windows, OCFS is already included.
Automatic Storage Management
• Provides the first portable and high-performance

database file system
• Manages Oracle database files
Application
• Contains data spread across
disks to balance load
Database
• Provides integrated mirroring a
across disks File
h as
system
ฺ b
ASM
r)
• Solves many storage Volume o m eฺ
ฺ c
management challenges edi
manager uid
icr ent G
s
r d e@ S t ud system
Operating
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Management
f er
o
AutomaticiStorage
ic -tran s
a b r
Automatic Storage
F n o n Management (ASM) provides a vertical integration of the file system and the
volume manager that is specifically built for Oracle database files. ASM can provide management
for single SMP machines or across multiple nodes of a cluster for Oracle Real Application
Clusters support.
ASM distributes I/O load across all available resources to optimize performance while removing
the need for manual I/O tuning. It helps DBAs manage a dynamic database environment by
allowing them to increase the database size without having to shut down the database to adjust the
storage allocation.
ASM can maintain redundant copies of data to provide fault tolerance, or it can be built on top of
vendor-supplied, reliable storage mechanisms. Data management is done by selecting the desired
reliability and performance characteristics for classes of data rather than with human interaction
on a per-file basis.
The ASM capabilities save DBAs time by automating manual storage and thereby increasing their
ability to manage larger databases (and more of them) with increased efficiency.
Note: ASM is the strategic and stated direction as to where Oracle database files should be stored.
However, OCFS will continue to be developed and supported for those who are using it.
CFS or Raw?
• Using CFS:
– Simpler management
– Use of OMF with RAC
– Single Oracle software installation
– Autoextend
• Using raw:
– Performance ha sa
ฺb r )
– Use when CFS not available m
i ฺ c o deฺ
– Cannot be used for archivelog files
r e d Gui
ic ent
– ASM eases work @s d e Stud
r
ve this
i l a
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
i
CFS or Raw? o an you can either use a cluster file system or place files on raw devices.
Asa b ric explained,
already - t r
F non
Cluster file systems provide the following advantages:
• Greatly simplified installation and administration of RAC
• Use of Oracle Managed Files with RAC
• Single Oracle software installation
• Autoextend enabled on Oracle data files
• Uniform accessibility to archive logs in case of physical node failure
Raw devices implications:
• Raw devices are always used when CFS is not available or not supported by Oracle.
• Raw devices offer best performance without any intermediate layer between Oracle and the
disk.
• Autoextend fails on raw devices if the space is exhausted.
• ASM, Logical Storage Managers, or Logical Volume Managers can ease the work with raw
devices. Also, they can enable you to add space to a raw device online, or you may be able to
create raw device names that make the usage of this device clear to the system administrators.
Typical Cluster Stack with RAC
Servers
Interconnect
High-speed Interconnect: Gigabit Ethernet Proprietary

UDP Proprietary
OS C/W
Oracle Clusterware a
as
RAC RAC RAC RAC )h ฺbr
Linux, UNIX, Linux Linux AIX, HP-UX, Solaris
c m
o deฺ
Windows Windows Windows i ฺ
d Gui
r e
ASM OCFS RAW
@ sic deRAW
ASM nt CFS
e r de StuOS CVM
Database v i lav estorage
shared t his
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er RAC
V Stacksfwith
i o
Typical Cluster
ic -tran
b r
Fa noinna cluster requires a supported interconnect software protocol to support interinstance
Each node
communication, and Transmission Control Protocol/Internet Protocol (TCP/IP) to support Oracle
Clusterware polling.
All UNIX and Linux platforms use User Datagram Protocol (UDP) on Gigabit Ethernet (GbE) as
one of the primary protocols and interconnect for RAC interinstance IPC communication. Other
vendor-specific interconnect protocols include Remote Shared Memory for SCI and SunFire Link
interconnects, and Hyper Messaging Protocol for Hyperfabric interconnects. In any case, your
interconnect must be certified by Oracle for your platform.
Using Oracle Clusterware, you can reduce installation and support complications. However,
vendor clusterware may be needed if customers use non-Ethernet interconnect or if you have
deployed clusterware-dependent applications on the same cluster where you deploy RAC.
Similar to the interconnect, the shared storage solution you choose must be certified by Oracle for
your platform. If a cluster file system (CFS) is available on the target platform, then both the
database area and flash recovery area can be created on either CFS or ASM. If a CFS is
unavailable on the target platform, then the database area can be created either on ASM or on raw
devices (with the required volume manager), and the flash recovery area must be created on the
ASM.
Note: It is strongly recommended that you use UDP over GbE.
RAC Certification Matrix
1. Connect and log in to http://metalink.oracle.com.

2. Click the Certify tab on the menu frame.
3. Click the “View Certifications by Product” link.
4. Select Real Application Clusters. Then click Submit.
5. Select the correct platform and click Submit.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Matrix f er
i o
RAC Certification
ic -tran s
b r
Fa non Clusters Certification Matrix is designed to address any certification
Real Application
inquiries. Use this matrix to answer any certification questions that are related to RAC. To
navigate to Real Application Clusters Certification Matrix, perform the steps shown in the slide.
RAC and Services
Application server
ERP Run-time connection load balancing CRM
Service location transparency
Stop or start service connections.
Service connections
Modify service to instance mapping.

Listeners
Connection load balancing Service availability aware
RAC instances s a
)h a
ERP ERP Backup
m ฺbr ERP ERP
CRM CRM
Priority o
CRM diฺcCRM ide
Alerts
ฺ
i c re t Gu
Tuning
@ s den
e r de Stu
Oracle Clusterware
i v this
lanotification
Up/down/LBA events
o v
_ components s e engine
i
ric e to
Restart failed u
a b
(f ens
d e ic © 2008, Oracle. All rights reserved.
r lCopyright
l
e
V ab e
i l a
V sfer
i o
RAC and Services
ic -tran
b r
Fa nareona logical abstraction for managing workloads. Services divide the universe of work
Services
executing in the Oracle database into mutually disjoint classes. Each service represents a workload
with common attributes, service-level thresholds, and priorities.
Services are built into the Oracle database providing a single-system image for workloads,
prioritization for workloads, performance measures for real transactions, and alerts and actions
when performance goals are violated. These attributes are handled by each instance in the cluster
by using metrics, alerts, scheduler job classes, and the resource manager.
With RAC, services facilitate load balancing, allow for end-to-end lights-out recovery, and
provide full location transparency.
A service can span one or more instances of an Oracle database in a cluster, and a single instance
can support multiple services. The number of instances offering the service is transparent to the
application. Services enable the automatic recovery of work. Following outages, the service is
recovered automatically at the surviving instances. When instances are later repaired, services that
are not running are restored automatically by Oracle Clusterware. Immediately the service
changes state, up, down, or too busy; a notification is available for applications using the service
to trigger immediate recovery and load-balancing actions. Listeners are also aware of services
availability, and are responsible for distributing the workload on surviving instances when new
connections are made. This architecture forms an end-to-end continuous service for applications.
Available Demonstrations
• RAC scalability and transaction throughput

• RAC speedup and parallel queries
• Use TAF with SELECT statements
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
http://www.oracle.com/technology/obe/demos/admin/demos.html
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Available Demonstrations
an concepts that were briefly introduced in this lesson, online demonstrations
Toa b ric the-tmajor
illustrate
r
F nonat http://www.oracle.com/technology/obe/demos/admin/demos.html.
are available
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Oracle Clusterware
Installation and Configuration
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives
After completing this lesson, you should be able to:

• Describe the installation of Oracle RAC 11g
• Perform RAC preinstallation tasks
• Perform cluster setup tasks
• Install Oracle Clusterware
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Oracle Database 11g: RAC Administration 1 - 2
Oracle RAC 11g Installation
Oracle RAC 11g incorporates a two-phase installation

process:
• Phase one installs Oracle Clusterware.
• Phase two installs the Oracle Database 11g software
with RAC.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V Installation
f
Oracle RACio11g n s
b ric -oftraOracle
Theainstallation Database 11g requires that you perform a two-phase process in which you
F non
run the Oracle Universal Installer (OUI) twice. The first phase installs Oracle Clusterware. Oracle
Clusterware provides high-availability components and can also interact with vendor clusterware, if
present, to coordinate cluster membership information.
The second phase installs the Oracle Database 11g software with RAC. The installation also enables
you to configure services for your RAC environment. If you have a previous Oracle cluster database
version, OUI activates the Database Upgrade Assistant (DBUA) to automatically upgrade your
preexisting cluster database. The Oracle Database 11g installation process provides a single-system
image, ease of use, and accuracy for RAC installations and patches.
Installation Utilities: New Features
There are new features and changed screens for OUI, Database Configuration Assistant (DBCA),
Network Configuration Assistant (NetCA), and DBUA. The enhancements include:
DBCA
• Provides a command-line feature (deleteASM) that removes ASM software
• Provides the option to switch from a database configured for Oracle Enterprise Manager
Database Control to Oracle Enterprise Manager Grid Control
Installation Utilities: New Features (continued)
DBUA
• On request for a backup during the upgrade, DBUA creates a restore script to restore the
database.
• Includes an improved pre-upgrade script to provide space estimation, initialization parameters,
statistics gathering, and new warnings. DBUA also provides upgrades from Oracle Database
releases 9.0, 9.2, 10.1, and 10.2.
• Supports in-place patch set upgrades
• Starts any services running prior to upgrades
NetCA
• NetCA deinstallation removes listener CRS resources that are defined in the same Oracle home
and do not have end points in the listener.ora file.
• NetCA converts existing noncluster listeners to cluster listeners.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Oracle RAC 11g Installation: Outline
1. Complete preinstallation tasks:

– Hardware requirements
– Software requirements
– Environment configuration, kernel
parameters, and so on
2. Perform Oracle Clusterware installation.
s a
3. Perform ASM installation. h a
ฺ b r)
4. Perform Oracle Database 11g software o m eฺ
installation. ฺ c
di Guid
r e
ic eGrid t Control.
5. Install EM agent on cluster nodes@ if s
using n
r d e Stud
i l a ve this
6. Perform cluster database creation.
v se
io_ tasks.
7. Complete postinstallation u
ric e to
a b
(f ens
r lCopyright
l
e
V ab e
i l a er Outline
V Installation:
Oracle RACio11g s f
To a b ric -install
successfully t r an Oracle RAC 11g, it is important that you have an understanding of the tasks
F
that must n
on
be completed and the order in which they must occur. Before the installation can begin in
earnest, each node that is going to be part of your RAC installation must meet the hardware and
software requirements that are covered in this lesson. You must perform step-by-step tasks for
hardware and software verification, as well as for the platform-specific preinstallation procedures.
You must install the operating system patches required by the cluster database, and you must verify
that the kernel parameters are correct for your needs.
Oracle Clusterware must be installed using OUI. Make sure that your cluster hardware is functioning
normally before you begin this step. Failure to do so results in an aborted or nonoperative
installation.
After Oracle Clusterware has been successfully installed and tested, again use OUI to install
Automatic Storage Management (ASM) and the Oracle Database 11g software, including software
options required for a RAC configuration. If you intend to use Enterprise Manager Grid Control to
manage your RAC deployments, you must next install the Enterprise Manager (EM) agent on each
cluster node. After the database has been created, there are a few post-installation tasks that must be
completed before your RAC database is fully functional. The remainder of this lesson provides the
necessary knowledge to complete these tasks successfully.
Note: It is not mandatory to install ASM to a separate ORACLE_HOME. This is considered a best
practice.
Windows and UNIX Installation Differences
• Startup and shutdown services

• Environment variables
• DBA account for database administrators
• Account for running OUI
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfeBetween r
Installationio Differences
n Windows and UNIX
r i c r a
t with installing Oracle components in UNIX environments, note that many
Fabarenexperienced
If you
on-
manual setup tasks required on UNIX are not required on Windows. The key differences between
UNIX and Windows installations are discussed below:
• Startup and Shutdown Services
In Windows, OUI creates and sets startup and shutdown services at installation time. In UNIX
systems, administrators are responsible for creating these services.
• Environment Variables
In Windows, OUI sets environment variables such as PATH, ORACLE_BASE, ORACLE_HOME,
and ORACLE_SID in the registry. In UNIX systems, you must manually set these environment
variables.
• DBA Account for Database Administrators
In Windows, OUI creates the ORA_DBA group. In UNIX systems, you must create the DBA
account manually.
• Account for Running OUI
In Windows, you log in with Administrator privileges. You do not need a separate account. In
UNIX systems, you must create this account manually. On Oracle RAC systems, each member
node of the cluster must have user equivalency for the account that installs the database. This
means that the administrative privileges user account and password must be the same on all
nodes.
Preinstallation Tasks
Check system requirements.

Check software requirements.
Check kernel parameters.
Create groups and users.
Perform cluster setup.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VTaskssfer
i o
Preinstallation
ic must abencompleted before the Oracle Clusterware and Oracle Database 11g software
b r - t r
Fa non
Several tasks
can be installed. Some of these tasks are common to all Oracle database installations and should be
familiar to you. Others are specific to Oracle RAC 11g.
Attention to details here simplifies the rest of the installation process. Failure to complete these tasks
can certainly affect your installation and possibly force you to restart the process from the beginning.
Note: It is strongly recommended that Network Time Protocol (NTP) be configured on all cluster
nodes before you install RAC.
Hardware Requirements
• At least 1 GB of physical memory is needed.

# grep MemTotal /proc/meminfo
MemTotal: 1126400 kB
• A minimum of 1 GB of swap space is required.

# grep SwapTotal /proc/meminfo
SwapTotal: 1566328 kB
s a
)h a
• The /tmp directory should be at least 400 MB. ฺbr
c m
o deฺ
# df -k /tmp i ฺ
d Gui
r e
sic de54%
Filesystem 1K-blocks Used Available Use%
/dev/sda6 6198556 3137920
@
2745756 nt
e r de Stu
• The Oracle Database 11g isoftware lav e thirequires s up to 4 GB
of disk space. v
o_ o us
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Hardware Requirements
an the following minimum hardware requirements:
b ric must-trmeet
Theasystem
F non
• At least 1 gigabyte (GB) of physical memory is needed. To determine the amount of physical
memory, enter the following command: grep MemTotal /proc/meminfo
• A minimum of 1 GB of swap space or twice the amount of physical memory is needed. On
systems with 2 GB or more of memory, the swap space can be between one and two times the
amount of physical memory. To determine the size of the configured swap space, enter the
following command: grep SwapTotal /proc/meminfo
• At least 400 megabytes of disk space must be available in the /tmp directory. To determine the
amount of disk space available in the /tmp directory, enter the following command: df -k
/tmp. Alternatively, to list disk space in megabytes or gigabytes, enter: df –h.
• Up to 4 GB of disk space is required for the Oracle Database 11g software, depending on the
installation type. The df command can be used to check for the availability of the required disk
space.
Network Requirements
• Each node must have at least two network adapters.

• Each public network adapter must support TCP/IP.
• The interconnect adapter must support User Datagram
Protocol (UDP).
• The host name and IP address associated with the
public interface must be registered in the domain name a
service (DNS) or the /etc/hosts file. has r)
ฺ b
m eฺ
ฺ c o
r e di Guid
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Network Requirements
anat least two network adapters: one for the public network interface and the
icmust-thave
b r r
Fa non
Each node
other for the private network interface or interconnect. In addition, the interface names associated
with the network adapters for each network must be the same on all nodes.
For the public network, each network adapter must support TCP/IP. For the private network, the
interconnect must support UDP (TCP for Windows) using high-speed network adapters and switches
that support TCP/IP. Gigabit Ethernet or an equivalent is recommended.
Note: For a more complete list of supported protocols, see MetaLink Note: 278132.1.
Before starting the installation, each node requires an IP address and an associated host name
registered in the DNS or the /etc/hosts file for each public network interface. One unused
virtual IP address and an associated VIP name registered in the DNS or the /etc/hosts file that
you configure for the primary public network interface are needed for each node. The virtual IP
address must be in the same subnet as the associated public interface. After installation, you can
configure clients to use the VIP name or IP address. If a node fails, its virtual IP address fails over to
another node. For the private IP address and optional host name for each private interface, Oracle
recommends that you use private network IP addresses for these interfaces, for example, 10.*.*.*
or 192.168.*.*. You can use the /etc/hosts file on each node to associate private host names
with private IP addresses.
Virtual IP Addresses and RAC
clnode-1 clnode-2 clnode-1vip
clnode-2vip
2 Clients 2
ERP=(DESCRIPTION= ERP=(DESCRIPTION=
4 1 ((HOST=clusnode-1)) 5 1 ((HOST=clusnode-1vip))
s a
6 ((HOST=clusnode-2)) 6 ((HOST=clusnode-2vip))
)h a
ฺbr
(SERVICE_NAME=ERP)) (SERVICE_NAME=ERP))
7 co
m e ฺ
Timeout i ฺ i d
5 wait 7
i c red t Gu
@ s den
e r de3 Stu
lav e this
3
_ v i clnode-2vip
clnode-1 clnode-2
ic i o o us 4 clnode-1vip
( f abr nse t
e rde lCopyright
l a V ab e
i
V sand f erRAC
i o
Virtual IP Addresses
an are all about availability of applications when an entire node fails.
ic(VIP)-taddresses
b r r
Fa non
Virtual IP
When a node fails, the VIP address associated with it automatically fails over to some other node in
the cluster. When this occurs:
• The new node indicates the new Media Access Control (MAC) address for the VIP. For directly
connected clients, this usually causes them to see errors on their connections to the old address.
• Subsequent packets sent to the VIP address go to the new node, which will send error reset
(RST) packets back to the clients. This results in the clients getting errors immediately.
This means that when the client issues SQL to the node that is now down (3), or traverses the address
list while connecting (1), rather than waiting on a very long TCP/IP timeout (5), which could be as
long as ten minutes, the client receives a TCP reset. In the case of SQL, this results in an ORA-3113
error. In the case of connect, the next address in tnsnames is used (6). The slide shows you the
connect case with and without VIP. Without using VIPs, clients connected to a node that died will
often wait a 10-minute TCP timeout period before getting an error. As a result, you do not really
have a good High Availability solution without using VIPs.
Note: After you are in the SQL stack and blocked on read/write requests, you need to use Fast
Application Notification (FAN) to receive an interrupt. FAN is discussed in more detail in the lesson
titled “High Availability of Connections.”
RAC Network Software Requirements
• Supported interconnect software protocols are

required:
– TCP/IP
– UDP
– Reliable Data Gram
• Token Ring is not supported on AIX platforms.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i r
V sfeRequirements
RAC Network i o Software
anrequires a supported interconnect software protocol to support Cache Fusion,
icin a -cluster
b r t r
Fa non
Each node
and TCP/IP to support Oracle Clusterware polling. In addition to UDP, other supported vendor-
specific interconnect protocols include Remote Shared Memory, Hyper Messaging protocol, and
Reliable Data Gram. Note that Token Ring is not supported for cluster interconnects on AIX. Your
interconnect must be certified by Oracle for your platform. You should also have a Web browser to
view online documentation.
Oracle Corporation has done extensive testing on the Oracle-provided UDP libraries (and TCP for
Windows). On the basis of this testing and extensive experience with production customer
deployments, Oracle Support strongly recommends the use of UDP (or TCP on Windows) for RAC
environments related to Oracle Database 11g. Best practices for UDP include:
• Use at least a gigabit Ethernet for optimal performance.
• Cross-over cables are not supported (use a high-speed switch).
• Increase the UDP buffer sizes to the OS maximum.
• Turn on UDP checksumming.
Oracle’s clusterware provides the same functionality as vendor-specific clusterware as required by
Oracle RAC. Also, using Oracle Clusterware reduces installation and support complications.
However, vendor clusterware may be needed if customers use non-Ethernet interconnect or if you
have deployed clusterware-dependent applications.
Package Requirements
• Package versions are checked by the cluvfy utility.

• For example, required packages and versions for Red
Hat 4.0 and Oracle Enterprise Linux 4 include:
– glibc-2.3.4-2.25
– glibc-common-2.3.4.2-25
– glibc-devel-2.3.4.2-25
s a
– gcc-3.4.6-3 h a
– gcc-c++-3.4.6-3 ฺ b r)
c o m eฺ
– libaio-0.3.105-2 ฺ
di Guid
r e
– libaio-devel-0.3.105-2
@ sic dent
– libstdc++-3.4.6-3.1 e r de Stu
– make-3.80-6 v i lav e this
– ic i o_ o us
br
sysstat-5.0.5-11 t
(fa ense
e ic © 2008, Oracle. All rights reserved.
rd lCopyright
l
e
V ab e
i
V sferl a
i o
Package Requirements
ic on the an that you intend to install, verify that the packages listed in the slide above
b r - t r
Fa noonn the system. Oracle Universal Installer (OUI) performs checks on your system to
Depending
are installed
products
verify that it meets the Linux package requirements of the cluster database and related services. To
ensure that these checks succeed, verify the requirements before you start OUI.
To determine whether the required packages are installed, enter the following commands:
# rpm -q package_name
# rpm –qa |grep package_name_segment
For example, to check the gcc compatibility packages, run the following command:
# rpm –qa |grep compat
compat-libstdc++-33-3.2.3-47.3
compat-gcc-32-3.2.3-47.3
compat-libgcc-296-2.96-132.7.2
compat-gcc-32-c++-3.2.3-47.3
If a package is not installed, install it from your Linux distribution media as the root user by using
the rpm –i command. For example, to install the compat-db package, use the following command:
# rpm –i compat-db-4.1.25-9.i386.rpm
Required UNIX Groups and Users
• Create an oracle user, a dba, and an oinstall group

on each node:
# groupadd -g 500 oinstall
# groupadd -g 501 dba
# useradd -u 500 -d /home/oracle -g "oinstall" \
–G "dba" -m -s /bin/bash oracle
s a
• Verify the existence of the nobody nonprivileged b user:) h a
ฺ r
c o m e ฺ
# grep nobody /etc/passwd i ฺ
d Gu i d
Nobody:x:99:99:Nobody:/:/sbin/nobody icre
@ s dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VGroups f r
eand
Required UNIX
i c i o a n s Users
r
b create
Youamust r
-tthe oinstall group the first time you install the Oracle database software on the
F o n
n group owns the Oracle inventory, which is a catalog of all the Oracle database software
system. This
installed on the system.
You must create the dba group the first time you install the Oracle database software on the system.
It identifies the UNIX users that have database administrative privileges. If you want to specify a
group name other than the default dba group, you must choose the custom installation type to install
the software, or start OUI as a user that is not a member of this group. In this case, OUI prompts you
to specify the name of this group.
You must create the oracle user the first time you install the Oracle database software on the
system. This user owns all the software installed during the installation. The usual name chosen for
this user is oracle. This user must have the Oracle Inventory group as its primary group. It must
also have the OSDBA (dba) group as the secondary group.
You must verify that the unprivileged user named nobody exists on the system. The nobody user
must own the external jobs (extjob) executable after the installation.
Note: The GIDs (group ID) for the oinstall and dba groups and the UID (user ID)for the
oracle user must be identical on both nodes.
oracle User Environment
• Set umask to 022.

• Set the DISPLAY environment variable.
• Set the ORACLE_BASE environment variable.
• Set the TMP and TMPDIR variables, if needed.
$ cd
s a
$ vi .bash_profile h a
umask 022 ฺ b r)
c o m eฺ
ฺ
ORACLE_BASE=/u01/app/oracle; export ORACLE_BASE
di Guid
TMP=/u01/mytmp; export TMP r e
TMPDIR=$TMP; export TMPDIR sic ent
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
VEnvironmentf er
oracle c
i i
Usero an s
a b r - t r
You
F must
n
run
o n as the oracle user. However, before you start OUI, you must configure the
OUI
environment of the oracle user. To configure the environment, you must:
• Set the default file mode creation mask (umask) to 022 in the shell startup file
• Set the DISPLAY and ORACLE_BASE environment variables
• Secure enough temporary disk space for OUI
If the /tmp directory has less than 400 megabytes of free disk space, identify a file system that is
large enough and set the TMP and TMPDIR environment variables to specify a temporary directory
on this file system. Use the df -k command to identify a suitable file system with sufficient free
space. Make sure that the oracle user and the oinstall group can write to the directory.
# df -k
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/hdb1 3020140 2471980 394744 87% /
/dev/hdb2 3826584 33020 3599180 1% /home
/dev/dha1 386008 200000 186008 0% /dev/shm
/dev/hdb5 11472060 2999244 7890060 28% /u01
# /mkdir /u01/mytmp
# chmod 777 /u01/mytmp
User Shell Limits
• Add the following lines to the

/etc/security/limits.conf file:
* soft nproc 2047
* hard nproc 16384
* soft nofile 1024
* hard nofile 65536
s a
• Add the following line to the /etc/pam.d/login file: ) h a
ฺ b r
c o m e ฺ
session required /lib/security/pam_limits.so i ฺ i d
i c red t Gu
@ s den
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
User Shell Limits
an
To a b ric the-performance
improve t r of the software, you must increase the following shell limits for the
F
oracle n on
user:
• nofile: The maximum number of open file descriptors should be 65536.
• nproc: The maximum number of processes available to a single user must not be less than
16384.
The hard values, or upper limits, for these parameters can be set in the
/etc/security/limits.conf file as shown in the slide above. The entry configures
Pluggable Authentication Modules (PAM) to control session security. PAM is a system of libraries
that handle the authentication tasks of applications (services) on the system. The principal feature of
the PAM approach is that the nature of the authentication is dynamically configurable.
Configuring for Remote Installation
To configure Secure Shell:

1. Create the public and private keys on all nodes:
[vx0044]$ /usr/bin/ssh-keygen -t dsa
2. Concatenate id_dsa.pub from all nodes into the

authorized_keys file on the first node: a
as
[vx0044]$ ssh vx0044 "cat ~/.ssh/id_dsa.pub" >> \ )h ฺbr
~/.ssh/authorized_keys
c m
o deฺ
[vx0044]$ ssh vx0045 "cat ~/.ssh/id_dsa.pub" >> \ i ฺ
d Gui
~/.ssh/authorized_keys r e
@ sic dent
3. Copy the authorized_keys file
e r deto the S tuother nodes:
[vx0044]$ scp ~/.ssh/authorized_keys v i his
lav evx0045:/home/oracle/.ssh/
t
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VRemote f r
eInstallation
Configuring i ofor
ic -tra-i n s
a b r
$ rpm
F n o n
-qa|grep
openssh-3.9p1-8.RHEL4.20
openssh
openssh-askpass-3.9p1-8.RHEL4.20
openssh-clients-3.9p1-8.RHEL4.20
openssh-server-3.9p1-8.RHEL4.20
Assume that your cluster comprises two nodes, vx0044 and vx0045. You can perform the following
steps to configure SSH using DSA on that cluster. Note that SSH using SSA is also supported.
1. As the oracle user, create the public and private keys on both nodes:
Accept the default location for the key file. When prompted for the pass phrase, just press the
Enter key.
2. Concatenate the contents of the id_dsa.pub file from each node into the authorized_keys
file on the first node.
[vx0044]$ ssh vx0044 "cat ~/.ssh/id_dsa.pub" >> \
[vx0044]$ ssh vx0045 "cat ~/.ssh/id_dsa.pub" >> \
Configuring for Remote Installation (continued)
3. Copy the authorized_keys file to the same location on the second node.
[vx0044]$ scp ~/.ssh/authorized_keys vx0045:/home/oracle/.ssh/
4. Test the configuration.
[vx0044]$ ssh vx0045 hostname
$ ssh vx0045 uptime
vx0045.us.oracle.com
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Required Directories for the
Oracle Database Software
You must identify five directories for the Oracle database

software:
• Oracle base directory
• Oracle inventory directory
• Oracle Clusterware home directory
• Oracle home directory for the database a
h as
• Oracle home directory for ASM r)
m ฺb
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfor f r
ethe
i c i o
Required Directories
a n Oracle Database Software
r r
F b nobase
TheaOracle n-t(ORACLE_BASE) directory acts as a top-level directory for the Oracle database
software installations. On UNIX systems, the Optimal Flexible Architecture (OFA) guidelines
recommend that you must use a path similar to the following for the Oracle base directory:
/mount_point/app/oracle_sw_owner
where mount_point is the mount-point directory for the file system that contains the Oracle
database software and oracle_sw_owner is the UNIX username of the Oracle database software
owner, which is usually oracle.
The Oracle inventory directory (oraInventory) stores the inventory of all software installed on
the system. It is required by, and shared by, all the Oracle database software installations on a single
system. The first time you install the Oracle database software on a system, OUI prompts you to
specify the path to this directory. If you are installing the software on a local file system, it is
recommended that you choose the following path: ORACLE_BASE/oraInventory
OUI creates the directory that you specify and sets the correct owner, group, and permissions on it.
Required Directories for the Oracle Database Software (continued)
The Oracle Clusterware home directory is the directory where you choose to install the software for
Oracle Clusterware. You must install Oracle Clusterware in a separate home directory. Because the
clusterware parent directory should be owned by root, it requires a separate base directory from the
one used by the database files. When you run OUI, it prompts you to specify the path to this
directory, as well as a name that identifies it. It is recommended that you specify a path similar to the
following for the Oracle Clusterware home directory:
/u01/crs11g
Note that in the example above, /u01 should be owned by the root user and writable by group
oinstall.
The Oracle home directory is the directory where you choose to install the software for a particular
Oracle product. You must install different Oracle products, or different releases of the same Oracle
product, in separate Oracle home directories. When you run OUI, it prompts you to specify the path
to this directory, as well as a name that identifies it. The directory that you specify must be a a
subdirectory of the Oracle base directory. It is recommended that you specify a path similar h atosthe
following for the Oracle home directory:
m eฺ ฺbr)
ORACLE_BASE/product/11.1.0/db_1
ฺ c o
r e
Consider creating a separate home directory for ASM if you will be using diti to manage
G u id your shared
sic dent
storage. Specify a path similar to the following directory for ASM:
@
ORACLE_BASE/product/11.1.0/asm_1 de tu
e r S
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Linux Operating System Parameters
Parameter Value File

semmsl 250 /proc/sys/kernel/sem
semmns 32000 /proc/sys/kernel/sem
semopm 100 /proc/sys/kernel/sem
semmni 128 /proc/sys/kernel/sem
shmall 2097152 /proc/sys/kernel/shmall
shmmax ½ physical /proc/sys/kernel/shmmax
memory s a
)h a
shmmni 4096 /proc/sys/kernel/shmmni
m ฺ br
file-max 65536 /proc/sys/fs/file-max
i ฺ c o deฺ
rmem_max 4194304 /proc/sys/net/core/rmem_max
r e d Gui
rmem_default 4194304 sic dent
/proc/sys/net/core/rmem_default
@
wmem_max 262144 e r de Stu
/proc/sys/net/core/wmem_max
wmem_default 262144 v i lav e this
/proc/sys/net/core/wmem_default
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V System f erParameters
i
Linux Operatingo s
an shown in the table above are set to values greater than or equal to the
ic the parameters
b r - t r
Fa nonvalue shown. Use the sysctl command to view the default values of the various
Verify that
recommended
parameters. For example, to view the semaphore parameters, run the following command:
# sysctl -a|grep sem
kernel.sem = 250 32000 32 128
The values shown represent semmsl, semmns, semopm, and semmni in that order. Kernel
parameters that can be manually set include:
• SEMMNS: The number of semaphores in the system
• SEMMNI: The number of semaphore set identifiers that control the number of semaphore sets
that can be created at any one time
• SEMMSL: Semaphores are grouped into semaphore sets, and SEMMSL controls the array size, or
the number of semaphores that are contained per semaphore set. It should be about ten more
than the maximum number of the Oracle processes.
• SEMOPM: The maximum number of operations per semaphore operation call
• SHMMAX: The maximum size of a shared-memory segment. This must be slightly larger than the
largest anticipated size of the System Global Area (SGA), if possible.
• SHMMNI: The number of shared memory identifiers
Linux Operating System Parameters (continued)
• RMEM_MAX: The maximum TCP receive window (buffer) size
• RMEM_DEFAULT: The default TCP receive window size
• WMEM_MAX: The maximum TCP send window size
• WMEM_DEFAULT: The default TCP send window size
You can adjust these semaphore parameters manually by writing the contents of the
/proc/sys/kernel/sem file:
# echo SEMMSL_value SEMMNS_value SEMOPM_value \
SEMMNI_value > /proc/sys/kernel/sem
To change these parameter values and make them persistent, edit the /etc/sysctl.conf file as
follows:
# vi /etc/sysctl.conf
...
kernel.sem = 250 32000 100 128
s a
kernel.shmall = 2097152
)h a
ฺbr
kernel.shmmax = 2147483648
kernel.shmmni = 4096
c m
o deฺ
fs.file-max = 65536 i ฺ
d Gui
r e
sic dent
rmem_max = 4194304
rmem_default = 4194304
@
de Stu
wmem_default = 262144
e r
wmem_max = 262144
v i lav e this
ic i o_ o us
net.ipv4.ip_local_port_range = 1024 65000
The kernel parameters shown above b arer recommended
t values only. For production database systems,
( f a s e
r d e licen to optimize the performance of the system.
it is recommended that you tune these values
Note: Because they areV aelot of parameters
l e to check, you can use the Cluster Verification Utility to
a ra b
l verification.
automatically do ithe
o V s f e
b r ici -tran
Fa non
Cluster Setup Tasks
1. View the Certifications by Product section at

http://metalink.oracle.com/.
2. Verify your high-speed interconnects.
3. Determine the shared storage (disk) option for your
system:
– OCFS or other shared file system solution
s a
– Raw devices h a
– ASM ฺ b r)
m eฺ
ฺi cofiles!
ASM cannot be used for the OCR and Voting Disk id d u
re t G
sic patches.
4. Install the necessary operating system n
r d e@ Stude
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
V i f er
i
Cluster Setup o Tasks
ic you-have s
ana certified combination of the operating system and the Oracle database
b r t r
Fa non by referring to the certification information on Oracle MetaLink in the Availability
Ensure that
software version
& Certification section. See the Certifications by Product section at http://metalink.oracle.com.
Verify that your cluster interconnects are functioning properly. If you are using vendor-specific
clusterware, follow the vendor’s instructions to ensure that it is functioning properly.
Determine the storage option for your system, and configure the shared disk. Oracle recommends
that you use Automatic Storage Management (ASM) and Oracle Managed Files (OMF), or a cluster
file system such as Oracle Cluster File System (OCFS). If you use ASM or a cluster file system, you
can also utilize OMF and other Oracle Database 11g storage features.
Oracle Clusterware requires that the OCR files and voting disk files be shared. Note that ASM cannot
be used to store these files as the clusterware components are started before the ASM or RAC
instances. These files could map to shared block or raw devices or exist on an OCFS volume.
Verifying Cluster Setup with cluvfy
• Install the cvuqdisk rpm required for cluvfy:

# su root
# cd /stage/db/rpm
# export CVUQDISK_GRP=dba
# rpm -iv cvuqdisk-1.0.1-1.rpm
• Run the cluvfy utility as oracle as shown below: s a

)h a
# cd /stage/db
m ฺ br
co
./runcluvfy.sh stage -post hwos -n all -verbose eฺ d iฺ uid
e
icr ent G
s
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Setup f r cluvfy
ewith
o
Verifying Cluster
i s
an Utility (cluvfy) enables you to perform many preinstallation and
b ric Verification
TheaCluster - t r
F non
postinstallation checks at various stages of your RAC database installation. The cluvfy utility is
available in Oracle Database 11g Release 1. To check the readiness of your cluster for an Oracle
Clusterware installation, run cluvfy as shown below:
$ runcluvfy.sh stage -post hwos -n all -verbose
Performing post-checks for hardware and operating system setup
Checking node reachability...
...
Result: Node reachability check passed from node "vx0044".
Checking user equivalence...
...
Result: User equivalence check passed for user "oracle".
Checking node connectivity...
...
Result: Node connectivity check passed.
Checking shared storage accessibility...
...
Shared storage check passed on nodes "vx0045,vx0044".
...
Post-check for hardware and operating system setup was successful on all
the nodes.
Installing Oracle Clusterware
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
$ export ORACLE_BASE=/u01/app/oracle
s ic ent
$ /stage/db/runInstaller e
d @ t ud
r
ve this S
i l a
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V Clusterware
f
i o
Installing Oracle s
ic -tran environment variable in accordance with your installation location scheme.
Seta b
the
rORACLE_BASE
F n
Run OUI n byoexecuting the runInstaller command from the installation CD or the staged
software directory. The “Select a Product to Install” screen allows you to install and create a
database, install Oracle client software, or install Oracle Clusterware. Click the Oracle Clusterware
button and then click Next.
If you are performing this installation in an environment in which you have never installed the Oracle
database software (that is, if the environment does not have an OUI inventory), OUI displays the
“Specify Inventory directory and credentials” screen. If you are performing this installation in an
environment where the OUI inventory is already set up, OUI displays the Specify File Locations
screen instead of the “Specify Inventory directory and credentials” screen.
Specifying the Inventory Directory
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VInventoryf erDirectory
o
Specifying ithe s
an directory and credentials” screen, enter the inventory location. If
ic -tInventory
On a b
the
r“Specify r
F non
ORACLE_BASE has been properly set, OUI suggests the proper directory location for the inventory
location as per OFA guidelines. If ORACLE_BASE has not been set, enter the proper inventory
location according to your requirements. Enter the UNIX group name information oinstall in the
“Specify Operating System group name” field, and then click Next.
Specify Home Details
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VDetailssfer
Specify Homei o
ic displays n
athe
b r - t r
Fa non
Next, OUI Specify Home Details screen. The “Specify Home Details” screen contains
predetermined information for the source of the installation files and the target destination
information. OUI provides an Oracle Clusterware Home name in the Name field located in the
Destination section of the screen. You may accept the name or enter a new name at this time. If
ORACLE_BASE has been set, an OFA-compliant directory path appears in the Path field located
below the Destination section. If not, enter the location in the target destination, and click Next to
continue.
If ORACLE_HOME is set in the environment, this appears in the OUI location window.
ORACLE_HOME typically refers to the DB home, and there is no corresponding environment variable
for the Clusterware installation. You should be aware of this, and not just click through because there
is a value. The parent directory of the Clusterware Home should be owned by root and writable by
the install group, oinstall.
Product-Specific Prerequisite Checks
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er Checks
V Prerequisite
f
i o
Product-Specific s
an your environment to ensure that it meets the minimum requirements for an
b ric then
Theainstaller - t rchecks
F non
Oracle Clusterware installation. The installer checks for the existence of critical packages and release
levels, proper kernel parameter settings, network settings, and so on. If discrepancies are found, they
are flagged and you are given an opportunity to correct them. If you are sure that the flagged items
will not cause a problem, it is possible to click the item and change the status to self-checked, and
continue with the installation. Only do this if you are absolutely sure that no problems actually exist,
otherwise correct the condition before proceeding. When all checks complete successfully, click the
Next button to proceed.
Cluster Configuration
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Cluster Configuration n
b ric Cluster
TheaSpecify - t r aConfiguration screen displays predefined node information if OUI detects that
F non
your system has vendor clusterware. Otherwise, OUI displays the Cluster Configuration screen
without the predefined node information. If all your nodes do not appear in the cluster nodes window,
click the Add button. You must supply the public node names, private node names, and virtual host
names for each node that you add. All of these names must be resolvable on every node by using
either DNS or the /etc/hosts file.
In the Cluster Name field, enter a name for your cluster. Ensure that the cluster name is unique in
your network.
Private Interconnect Enforcement
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V sEnforcement
f
i o
Private Interconnect
anInterface Usage screen enables you to select the network interfaces on your
b ric Network
TheaSpecify - t r
F non
cluster nodes to use for internode communication. Ensure that the network interfaces that you choose
for the interconnect have enough bandwidth to support the cluster and RAC-related network traffic.
A gigabit Ethernet interface is highly recommended for the private interconnect. To configure the
interface for private use, click the interface name, and then click Edit. A pop-up window appears and
allows you to indicate the usage for the network interfaces. In the example shown in the slide, there
are three interfaces: eth0, eth1, and eth2. The eth0 interface is the hosts’ primary network
interface and should be marked Public. The eth1 interface is dedicated for the storage network,
which supports this cluster’s shared storage. It should be marked Do Not Use. The eth2 interface is
configured for the private interconnect and should be marked Private. When you finish, click the
Next button to continue.
Oracle Cluster Registry File
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VRegistry f r
eFile
i o
Oracle Cluster n s
b ric Oracle
TheaSpecify - t r aCluster Registry Location screen appears next. Enter a fully qualified file name
F non
for the shared block or raw device or a shared file system file for the OCR file. If you are using an
external disk mirroring scheme, click the External Redundancy option button. You will be
prompted for a single OCR file location. If no mirroring scheme is employed, click the Normal
Redundancy option button. You will be prompted for two file locations. For highest availability,
provide locations that exist on different disks or volumes. Click Next to continue.
Voting Disk File
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Voting DiskioFile
b ric purpose
Theaprimary - t r anof the voting disk is to help in situations where the private network
F non
communication fails. When the private network fails, the clusters are unable to have all nodes remain
available because they cannot synchronize I/O to the shared disk. Therefore, some of the nodes must
go offline. The voting disk is used to communicate the node state information used to determine
which nodes will go offline.
Because the voting disk must be accessible to all nodes to accurately assess membership, the file
must be stored on a shared disk location. The voting disk can reside on a block or raw device or a
cluster file system.
In Oracle Database 11g Release 1, voting disk availability is improved by the configuration of
multiple voting disks. If the voting disk is not mirrored, then there should be at least three voting
disks configured.
Summary and Install
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VInstall sfer
Summary andi o an screen. Note that OUI must install the components shown in the summary
b ric the-trSummary
OUIadisplays
F non
window. Click the Install button. The Install screen is then displayed, informing you about the
progress of the installation.
During the installation, OUI first copies the software to the local node and then copies the software to
the remote nodes.
Run Configuration Scripts on All Nodes
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Scripts f er on All Nodes
i o
Run Configuration
ic displays s
aandialog box indicating that you must run the orainstRoot.sh and
Next,b r
OUI - t r
Fa non
root.sh script on all the nodes that are part of this installation. The root.sh script runs the
following assistants without your intervention:
• Oracle Cluster Registry Configuration Tool (ocrconfig)
• Cluster Configuration Tool (clscfg)
When the root.sh script has been run on all nodes, click the OK button to close the dialog box.
Run the cluvfy utility to verify the post crs installation.
Note: Make sure you run the above mentioned scripts serially on each node in the proposed order.
End of Installation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
End of Installation
an scripts have been run on both nodes, the Configuration Assistants page is
icconfiguration
b r - t r
Fa non
When the
displayed. The ONS Configuration Assistant and Private Interconnect Configuration Assistant are
run and their progress is displayed here. The Cluster Verification Utility is then run to test the
viability of the new installation. When the Next button is clicked, the End of Installation screen
appears. Click Exit to leave OUI.
Verifying the Oracle Clusterware Installation
• Check for Oracle Clusterware processes with the ps

command.
• Check the Oracle Clusterware startup entries in the
/etc/inittab file.
# cat /etc/inittab
# Run xdm in runlevel 5 s a
h a
x:5:respawn:/etc/X11/prefdm -nodaemon
ฺ b r)
h1:35:respawn:/etc/init.d/init.evmd run >/dev/null
c o m eฺ
2>&1 </dev/null i ฺ
d>/dev/null
u id
r e G
sic dent
h2:35:respawn:/etc/init.d/init.cssd fatal
@
de run
2>&1 </dev/null
e r
h3:35:respawn:/etc/init.d/init.crsd S tu>/dev/null
v this
2>&1 </dev/null vila o _ use
i
ric e to
a b
(f ens
r lCopyright
l
e
V ab e
i l a
V Clusterware er
Verifying the i o Oracle n s f Installation
r i c r a
F ab continuing
Before
o n -t with the installation of the Oracle database software, you must verify your Oracle
Clusterwaren installation and startup mechanism. With the introduction of Oracle RAC 11g, cluster
management is controlled by the evmd, ocssd, and crsd processes. Run the ps command on both
nodes to make sure that the processes are running.
$ ps –ef|grep d.bin
oracle 9332 9313 0 Oct30 ? 00:00:38 .../evmd.bin
root 9347 8498 0 Oct30 ? 00:25:20 .../crsd.bin reboot
oracle 9897 9414 0 Oct30 ? 00:09:40 .../ocssd.bin
oracle 18387 1 0 Oct30 ? 00:00:00 .../oclskd.bin
...
Check the startup mechanism for Oracle Clusterware. In Oracle RAC 11g, Oracle Clusterware
processes are started by entries in the /etc/inittab file, which is processed whenever the run
level changes (as it does during system startup and shutdown):
h1:35:respawn:/etc/init.d/init.evmd run >/dev/null 2>&1 </dev/null
h2:35:respawn:/etc/init.d/init.cssd fatal >/dev/null 2>&1 </dev/null
h3:35:respawn:/etc/init.d/init.crsd run >/dev/null 2>&1 </dev/null
Note: The processes are started at run levels 3 and 5 and are started with the respawn flag.
Verifying the Oracle Clusterware Installation (continued)
This means that if the processes abnormally terminate, they are automatically restarted. If you kill the
Oracle Clusterware processes, they automatically restart or, worse, cause the node to reboot. For this
reason, stopping Oracle Clusterware by killing the processes is not recommended. If you want to stop
Oracle Clusterware without resorting to shutting down the node, you should run the crsctl
command:
# /u01/crs11g/bin/crsctl stop
If you encounter difficulty with your Oracle Clusterware installation, it is recommended that you
check the associated log files. To do this, check the directories under the Oracle Clusterware Home:
$ORA_CRS_HOME/log/hostname: This directory contains the alert.log file for the nodes
Clusterware.
$ORA_CRS_HOME/log/hostname/crsd/: This directory contains the log files for the CRSD
process.
s a
$ORA_CRS_HOME/log/hostname/cssd/: This directory contains the log files for the CSSD
)h a
process.
m ฺbr
i ฺcfiles ฺ
o for dtheeEVMD
$ORA_CRS_HOME/log/hostname/evmd/: This directory contains the d log
re t Gu i
process. i c
s den
$ORA_CRS_HOME/log/hostname/client/: Log filesrd for
@
eOCRSaretuwritten here.
e
v i
When you have determined that your Oracle Clusterware his is successful and fully
lav einstallation
t
c i
functional, you may start the Oracle Database
i o_11gosoftware
us installation.
r
ab nse t
( f
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Summary
In this lesson, you should have learned how to:

• Describe the installation of Oracle RAC 11g
• Perform RAC preinstallation tasks
• Perform cluster setup tasks
• Install Oracle Clusterware
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Practice 1: Overview
This practice covers the following topics:

• Performing initial cluster configuration
• Installing Oracle Clusterware
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
RAC Software Installation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Install and configure Automatic Storage Management
(ASM)
• Install the Oracle database software
• Perform required tasks prior to database creation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Installing Automatic Storage Management
s a
)h a
m ฺbr
i ฺ c o deฺ
$ id r e d Gui
oracle
@ sic dent
$ /stage/db/runInstaller
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Storage f er Management
i o
Installing Automatic n
ic -traAutomatic s
b
Forathis
rinstallation, Storage Management (ASM) is used to manage the shared storage for
F non
the cluster database. After Oracle Clusterware is installed, run Oracle Universal Installer (OUI) from
either the installation CD or a staged software location and install ASM as the oracle user.
$ id
oracle
$ cd /stage/db
$./runInstaller
The “Select a Product to Install” screen allows you to install and create a database, install Oracle
client software, or install Oracle Clusterware. Click the Oracle Database 11g button and then click
Next.
Installation Type
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Installationio Type
b r - t r anType screen appears, select your installation type by clicking the Enterprise
icInstallation
Fa non
When the
Edition option button. Click the Next button to proceed.
Install Location
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i
Install Locationo anappears is the “Install Location” screen. Here you specify the location of your
b ricscreen-that
Theanext t r
F non
ASM home directory and installation name. Although it is possible for ASM and the database
installation to reside in the same directory and use the same files, you are installing ASM separately,
into its own ORACLE_HOME to prevent the database ORACLE_HOME from being a point of failure
for the ASM disk groups and to prevent versioning difficulties between the ASM and database file
installations. Be sure to specify a name for your installation that reflects this. Then click the Next
button to continue.
Hardware Cluster Installation Mode
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er Mode
V Installation
f
i o
Hardware Cluster
icSpecify an s
b r - t r
Fa non
When the Hardware Cluster Installation Mode screen appears, click the Cluster Installation
option button. Next, ensure that all nodes in your cluster are selected by clicking the Select All
button. If OUI does not display the nodes properly, perform clusterware diagnostics by executing the
olsnodes -v command from the ORA_CRS_HOME/bin directory, and analyze its output.
Alternatively, you may use the cluvfy utility to troubleshoot your environment. Refer to your
documentation if the detailed output indicates that your clusterware is not running properly.
When this is done, click the Next button to continue.
Product-Specific Prerequisite Checks
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er Checks
V Prerequisite
f
i o
Product-Specific s
b ric -tranPrerequisite Checks screen verifies the operating system requirements that
TheaProduct-Specific
F non
must be met for the installation to be successful. After each successful check, the Succeeded check
box is selected for that test. The test suite results are displayed at the bottom of the screen. Any tests
that fail are also reported here. The example in the slide shows the results of a completely successful
test suite. If you encounter any failures, try opening another terminal window and correct the
deficiency from another terminal window. Then return to OUI, and click the Retry button to rerun the
tests. It is possible to bypass the errors that are flagged by selecting the check box next to the error,
but this is not recommended unless you are absolutely sure that the reported error will not affect the
installation. When all tests have succeeded, click the Next button to continue.
Select Configuration Option
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sOption f er
i o
Select Configuration
an Option screen allows you to choose from the following options:
b ric Configuration
TheaSelect - t r
F non
• Install database software and create a database
• Configure ASM
• Install database software only (no database creation)
This installation is concerned only with installing the ASM Home, so click the Install Software
Only button. When you have done this, click the Next button to proceed.
Privileged Operating System Groups
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sSystemf er Groups
i o
Privileged Operating
an
ic -trOperating
On a b
the
rPrivileged Systems Groups page, choose the operating system groups
F non
corresponding to the Database trusted groups listed to support OS authentication. The default value is
dba for each group: OSDBA, OSOPER, and OSASM. Click Next to continue.
Summary
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i r
Summary io V nsfe
b ric screen
TheaSummary - tra appears next. You may scan the installation tree to verify your choices if you
F no n
like. Then click the Install button to proceed.
You can monitor the progress of the installation on the Install screen. After installing the files and
linking the executables on the first node, the installer copies the installation to the remaining nodes.
When the installation progress reaches 100%, OUI prompts you to execute configuration scripts on
all nodes.
Execute Configuration Scripts
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfScriptser
i o
Execute Configuration
anappears prompts you to run the root.sh script on the specified nodes. Open a
b ricscreen-that
Theanext t r
F non
terminal window for each node listed and run the root.sh script as the root user from the specified
directory.
End of Installation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
End of Installation
icroot.sh anscript has been executed on all nodes in the cluster, return to the Execute
Whenb r
the - t r
Fa non
Configuration scripts window and click the OK button to continue. When the installation is finished,
the End of Installation screen appears. Click the Exit button to quit.
ASM Configuration
s a
)h a
m ฺbr
i ฺ c o deฺ
$ export ORACLE_HOME=/u01/app/oracle/product/11.1.0/asm_1 r e d Gui
$ export PATH=$PATH:ORACLE_HOME/bin
@ sic dent
$ dbca
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ASM Configuration
an has been successfully installed, the ASM instances should be configured and
icASM -software
b r t r
Fa nodisk
After the
started and
n
groups should be created to support the shared storage needs of your cluster
database. DBCA is used to do this quickly and accurately. Execute dbca from the ASM
ORACLE_HOME/bin directory as shown in the slide. Select Oracle Real Application Clusters
database from the Welcome screen and click Next.
ASM Configuration
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V (continued)
f er
i o
ASM Configuration
ic screen, n s
aclick
On a b
the
rnext - t r Configure Automatic Storage Management and click the Next button.
F non
You now choose the nodes on which to manage ASM. Click the Select All button, and then click
Next to continue.
ASM Configuration
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V (continued)
f
i o
ASM Configuration
ic no-tASM s
an instances are running on your cluster nodes. Before starting the instances, you
b r r
Fa non
At this stage,
must provide a password for the ASM SYS user. In addition, choose the type and location of the
parameter file to be used for the ASM instances. If you require specific initialization parameter
values to be set for your ASM instances, you can modify the default values by clicking the ASM
Parameters button. Then click Next to continue.
ASM Configuration
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V (continued)
f er
i o
ASM Configuration
icbox informs s
anyou that listeners are not running on the cluster nodes and asks if you want to
b r - t r
Fa non
A dialog
start them now. When you click Yes to start the listeners, the ASM instances are started on the
cluster nodes.
Creating ASM Disk Groups
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
V i fer
Creating ASM i o Disk n
icinstances s
Groups
a
When b r
the n - traare started, the ASM Disk Groups page is displayed. Click the Create New
F no
button to create a disk group. On the disk group creation page, click the Change Disk Discovery
Path button to refine the search string that is used by DBCA to find candidate disk devices. Click the
OK button after providing the discovery path to your disks.
Creating ASM Disk Groups
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
V i fer (continued)
Creating ASM i o
c tra Disk n
icandidate s
Groups
b r
Fa non- disks and their relative size in megabytes is displayed. In the example in the slide,
A list of
a DATA disk group is created using normal redundancy. When you click OK, the ASM disk group
page is redisplayed showing the status of the newly created disk group. Other disk groups can also be
created at this point.
When you have finished creating all the necessary disk groups, click Finish. A dialog box asks if you
wish to perform other operations. Click No to exit DBCA.
Select a Product to Install
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
$ /stage/db/runInstaller
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V to Installf er
i o
Select a Product n s
b
TheaOUIricis used-ttorainstall the Oracle Database 11g software. You need to run OUI as the oracle
F non
user. Start OUI by executing the runInstaller command from the root directory of the Oracle
Database 11g Release 1 CD-ROM or the software staging location. The “Select a Product to Install”
screen allows you to install and create a database, install Oracle client software, or install Oracle
Clusterware. Click the Oracle Database 11g button and click Next.
Select Installation Type
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Type f er
i o
Select Installation s
an Type screen is displayed next. Your installation options include:
b ric Installation
TheaSelect - t r
F non
• Enterprise Edition
• Standard Edition
• Custom
For most installations, the Enterprise Edition installation is the correct choice (but Standard Edition
is also supported). Selecting the Custom installation type option enables you to install only those
Oracle product components that you deem necessary. For this, you must have a good knowledge of
the installable Oracle components and of any dependencies or interactions that may exist between
them. For this reason, it is recommended that you select the Enterprise Edition installation because it
installs all components that are part of the Oracle Database 11g 11.1.0 distribution.
Install Location
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i
Install Locationo
ic Location an screen, provide a value for ORACLE_BASE if you have not yet already done
On a b
the
rInstall - t r
F non
so. In the Software Location section of the page
The Name field is populated with a default or suggested installation name. Accept the suggested
name or enter your own Oracle Home name. Next, in the Path field, enter the fully qualified path
name for the installation, /u01/app/oracle/product/11.2.0/db_1 in the example in the
slide. After entering the information, review it for accuracy, and click the Next button to continue.
Specify Cluster Installation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VInstallation
f er
i o
Specify Cluster s
an Cluster Installation Mode screen is displayed next. Because OUI is node
b ric Hardware
TheaSpecify - t r
F non
aware, you must indicate whether you want the installation to be copied to the recognized and
selected nodes in your cluster, or whether you want a single, noncluster installation to take place.
Most installation scenarios require the Cluster Installation option.
To do this, click the Cluster Installation option button and make sure that all nodes have been
selected in Node Name list. Note that the local node is always selected for the installation. Additional
nodes that are to be part of this installation must be selected by selecting the check boxes. If you do
not see all your nodes listed here, exit OUI and make sure that Oracle Clusterware is running on all
your nodes. Restart OUI. Click the Next button when you are ready to proceed with the installation.
Products Prerequisite Check
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfChecker
i o
Products Prerequisite
b ric -tranPrerequisite Checks screen verifies the operating system requirements that
TheaProduct-Specific
F non
must be met for the installation to be successful. These requirements include:
• Certified operating system check
• Kernel parameters as required by the database software
• Required operating system packages and correct revisions
• Required glibc and glibc-compat (compatibility) package versions
In addition, OUI checks whether the ORACLE_BASE user environment variable has been set and, if
so, whether the value is acceptable.
After each successful check, the Succeeded check box is selected for that test. The test suite results
are displayed at the bottom of the page. Any tests that fail are also reported here. The example in the
slide shows the results of a completely successful test suite. If you encounter any failures, try opening
another terminal window and correct the deficiency. For example, if your glibc version is too low,
acquire the correct version of the glibc RPM, install it from another terminal window, return to
OUI, and click the Retry button to rerun the tests. When all tests have succeeded, click the Next
button to continue.
Select Configuration Option
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sOption f er
i o
Select Configuration
an Option screen appears. On this screen, you can choose to create a database
b ric Configuration
TheaSelect - t r
F non
as part of the database software installation or install ASM. If you choose to install a database, you
must select one of the preconfigured starter database types:
• General Purpose
• Transaction Processing
• Data Warehouse
• Advanced (user customizable)
If you choose one of these options, you are queried about the specifics of your database (cluster
database name, shared storage options, and so on). After OUI stops, the DBCA is launched to install
your database with the information that you provided.
You may also choose to defer the database creation by clicking the Install Software Only option
button. This option enables you to create the database by manually invoking the DBCA at some point
in time after OUI finishes installing the database software. This choice provides you with more
options than the standard preconfigured database models. Select the Install Software Only option.
Click the Next button to continue.
Privileged Operating System Groups
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sSystemf er Groups
i o
Privileged Operating
an
ic -trOperating
On a b
the
rPrivileged Systems Groups page, choose the operating system groups
F non
corresponding to the Database trusted groups listed to support OS authentication. The default value is
dba for the OSDBA, OSOPER, and OSASM groups respectively. Click Next to continue.
Check Summary
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i
Check Summaryo anis displayed next. Review the information on this page. Node information and
b ric screen
TheaSummary - t r
F non
space requirements can be viewed here, as well as selected software components. If you are satisfied
with the summary, click the Install button to proceed. If you are not, click the Back button to go back
and make the appropriate changes.
On the Install screen, you can monitor the progress of the installation. During installation, OUI
copies the software first to the local node and then to the remote nodes.
root.sh Script
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
root.sh script
i o an
icof the-tinstallation,
b
At the r
end r OUI displays a dialog box indicating that you must run the root.sh
Fa non
script as the root user on all the nodes where the software is being installed. Execute the root.sh
script on one node at a time, and then click the OK button to continue.
Required Tasks Prior to Database Creation
Set the environment variables for the Oracle database:
$ cd
$ vi .bash_profile
export ORACLE_BASE=/u01/app/oracle
export ORACLE_SID=RDB1
a
as
export ORACLE_HOME=/u01/app/oracle/product/11.1.0/db_1;
) h
ฺ b r
export PATH=$PATH:$ORACLE_HOME/bin
c o m e ฺ
i ฺ i d
i c red t Gu
@ s den
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i r
V PriorstofeDatabase
Required Tasks
i c i o a n Creation
r r
b nownset-tthe Oracle database–related environment variables for the oracle user so that they
Youacan
F no by the DBCA during database creation:
are recognized
$ cd
$ vi .bash_profile
export ORACLE_BASE=/u01/app/oracle
export ORACLE_SID=RDB1
export ORACLE_HOME=/u01/app/oracle/product/11.1.0/db_1
export PATH=$PATH:$ORACLE_HOME/bin
Checks Before Database Creation
• Use cluvfy to perform a configuration check before

database creation.
• Use the –pre option with the dbcfg argument when
executing runcluvfy.
• As oracle, run the command as shown below:
$ cd /stage/db
s a
$ ./runcluvfy.sh stage -pre dbcfg -n vx0306,vx0313 -d
)h a
/u01/app/oracle/product/11.1.0
m ฺbr
i ฺ c o deฺ
• Set the Oracle database–related environment r e d variables
u i
sic dent G
for the oracle user. @
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VDatabase f erCreation
i
Checks Beforeo n s
b ric Verify
TheaCluster - t r aUtility can be used for the configuration check prior to database creation. Run the
F nonUtility as the oracle user. Use the –pre option with the dbcfg argument to
Cluster Verify
thoroughly analyze your cluster before creating your RAC database. Your output should be similar to
the following example:
$ ./runcluvfy.sh stage -pre dbcfg -n vx0306,vx0313 -d
Performing pre-checks for database configuration
Checking node reachability...

Node reachability check passed from node "vx0306".
Checking user equivalence...

User equivalence check passed for user "oracle".
Checking administrative privileges...

User existence check passed for "oracle".
Group existence check passed for "dba".
Membership check for user "oracle" in group "dba" [as Primary] passed.
Checks Before Database Creation (continued)
Administrative privileges check passed.
Checking node connectivity...
Node connectivity check passed for subnet "138.2.204.0" with node(s)

vx0313,vx0306.
Node connectivity check passed for subnet "10.0.0.0" with node(s)
vx0313,vx0306.
Node connectivity check passed for subnet "192.168.255.0".
Suitable interfaces for VIP on subnet "138.2.204.0":

vx0313 eth0:138.2.204.33 eth0:138.2.205.171
vx0306 eth0:138.2.204.32 eth0:138.2.205.170
s a
) h a
Suitable interfaces for the private interconnect on subnet "10.0.0.0":
ฺ b r
vx0313 eth1:10.0.0.2
c o m e ฺ
vx0306 eth1:10.0.0.1 i ฺ i d
i c red t Gu
Node connectivity check passed.
@ s den
e r de Stu
Checking CRS integrity...
v i lav e this
Checking daemon liveness...
ic i o_ o us
abr nse t
Liveness check passed for "CRS daemon".
( f
efor "CSS
Liveness check passed
r d l i c e daemon".
Checking daemon V e
liveness...
b le
Liveness check
V i l apassed
e r afor "EVM daemon".
o s f
b r ici CRS-trhealth...
an
Fa health
Checking
CRS noncheck passed.
CRS integrity check passed.
Pre-check for database configuration was successful.
Summary

• Install and configure Automatic Storage Management
(ASM)
• Install the Oracle database software
• Perform required tasks prior to database creation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non

• Installing and configuring ASM
• Installing the Oracle database software
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
RAC Database Creation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Install the Enterprise Manager agent on each cluster node
• Create a cluster database
• Perform required tasks after database creation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Management Agent Installation:
Specify Installation Type
$ cd /stage/EM
$ ./runInstaller
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
VAgent sInstallation:
f
Management
i c i o a n Specify Installation Type
r t r
Fab arentwo
There
on-management tools available for your cluster database, Database Control and Grid
Control. Both tools are based on Enterprise Manager, but Grid Control is the preferred tool for
deploying and managing cluster databases in an enterprise setting. To use Grid Control, the
Management Agent must be installed on each managed node in your cluster. To install the
Management Agent, go to the Enterprise Manager Installation CD or a software staging area and
start the Oracle Universal Installer. The Installation page provides several install types from
which to choose from. Click the Additional Management Agent option button to install an
agent only.
Specify Installation Location
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V sLocation
f
i o
Specify Installation
ic -tran
b r
Fa npage
The next
on is the Specify Installation Location page. Because this cluster database
installation uses an ORACLE_BASE set to /u01/app/oracle, the agent software should be
installed in a subdirectory under ORACLE_BASE, with the other Oracle Homes. After starting the
installation and specifying the installation location as shown in the example in the slide, the
11.1.0 subdirectory looks like this:
$ pwd
$ ls -l
total 20
drwxrwx--- 3 oracle oinstall 4096 Nov 12 07:59 agent10g
drwxr-x--- 54 oracle oinstall 4096 Nov 12 07:00 asm_1
drwxr-x--- 54 oracle oinstall 4096 Nov 12 07:16 db_1
Specify Cluster Installation Mode
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er Mode
V Installation
f
i o
Specify Cluster
ic -tran s
b r
Fa nonCluster Installation Mode screen is displayed next. Because you want to install the
The Specify
agent on all cluster nodes, click the Select All button to choose all the nodes of the cluster. Each
node must be check marked before continuing. If all the nodes do not appear, you must stop the
installation and troubleshoot your environment. If no problems are encountered, click the Next
button to proceed.
Prerequisite Check and OMS Location
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sand f erOMS Location
i
Prerequisite o Check
ic -tran
b r
Fa nonnext checks the minimum requirements for the installation of the agent software.
The installer
Any deficiencies are reported. You should take care of any issues identified by the installer
before continuing with the installation. If the prerequisite check identifies no problems, click
Next to continue.
The next screen requires you to identify the location of the Oracle management service. You
must provide the host name and port number of an existing Grid Control console. In the example
in the slide, a previously installed Grid Control console is specified on the host
ex0043.us.oracle.com. The port number specified is the default port (4889) used by Grid
Control. After you complete this step, click the Next button to continue.
Agent Registration Password
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Passwordf er
i o
Agent Registration
ic -tran s
b
Onathe
rSpecify
non
Agent Registration Password screen, you must provide the agent registration
Fpassword. Do not specify an arbitrary password here, otherwise the agent registration will fail
and the target nodes and their managed components will not be visible from the Grid Control
server. In the example in the slide, you must provide the password for the Grid Control server
located on vx0314.us.oracle.com. When this information is provided, click the Next button
to continue.
The next screen is the Summary screen. Quickly review the information for accuracy and click
the Next button to continue.
Management Agent Installation Finish
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
VAgentsInstallation
f
Management
i c i o a n Finish
r r
-t of the agent installation, you can monitor the progress of the install from the
Fab the
During
n o n
course
Install screen. When the installation is complete, you will be prompted to execute the root.sh
scripts on all nodes where the agent was installed.
Executing the root.sh Script
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i erScript
ExecutingiotheVroot.sh s f
b r ic -tran
Fa no
When the n has been run on all nodes, close the Execute Configuration Scripts screen by
script
clicking the Exit button and return to the installer. When the installer indicates that the agent has
been installed, click the Exit button to quit the installer. You are now ready to create the cluster
database.
Creating the Cluster Database
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
$ cd /u01/app/oracle/product/11.1.0/db_1/bin
@ sic dent
$ ./dbca
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sDatabase
f er
Creating thei o Cluster
ic -tran
a b r
This
F database
n o n creation assumes that the database will use ASM for its shared storage. Change
directory to $ORACLE_HOME/bin on the installing node and execute the DBCA as shown
below:
$ cd /u01/app/oracle/product/11.1.0/db_1/bin
$ ./dbca
The Welcome screen appears first. You must select the type of database that you want to install.
Click the Oracle Real Application Clusters database option button, and then click Next. The
Operations screen appears. For a first-time installation, you have two choices only. The first
option enables you to create a database and the other option enables you to manage database
creation templates. Click the Create a Database option button, and then click Next to continue.
Node Selection
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Node Selection
ic -tran
b r
Fa noSelection
The Node n screen is now displayed. Because you are creating a cluster database, choose
all the nodes. Click the Select All button to choose all the nodes of the cluster. Each node must
be highlighted before continuing. If all nodes do not appear, you must stop the installation and
troubleshoot your environment. If no problems are encountered, click the Next button to
proceed.
Select Database Type
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Type f er
i o
Select Database
ic -tran s
b r
Fa non Templates screen appears. The DBCA tool provides several predefined database
The Database
types to choose from, depending on your needs. The templates include:
• General Purpose or Transaction Processing
• Custom Database
• Data Warehouse
In the example in the slide, the General Purpose or Transaction Processing option is chosen.
Click the Next button to continue.
Database Identification
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Database Identification
ic -tran
In a b rDatabase
non
the Identification screen, you must enter the database name in the Global Database
FName field. The name that you enter in this screen must be unique among all the global database
names used in your environment. The global database name can be up to 30 characters in length
and must begin with an alphabetical character.
A system identifier (SID) prefix is required, and the DBCA suggests a name based on your
global database name. This prefix is used to generate unique SID names for the two instances
that make up the cluster database. For example, if your prefix is RDBB, the DBCA creates two
instances on node 1 and node 2 named RDBB1 and RDBB2, respectively. This example assumes
that you have a two-node cluster. If you do not want to use the system-supplied prefix, enter a
prefix of your choice. The SID prefix must begin with an alphabetical character and contain no
more than 5 characters (on UNIX-based systems) or 61 characters (on Windows-based systems),
or 64 characters (on Linux-based systems). Click the Next button to continue.
Cluster Database Management Method
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Managementf er
i c i o
Cluster Database
a n s Method
r r
-t Options screen is displayed. For small cluster environments, you may choose
FabManagement
The
n o n
to manage your cluster with Enterprise Manager Database Control. To do this, select the
“Configure the Database with Enterprise Manager” check box. If you have Grid Control
installed somewhere on your network, you can select the “Use Grid Control for Database
Management” option. If you select Enterprise Manager with the Grid Control option and the
DBCA discovers agents running on the local node, you can select the preferred agent from a list.
Grid Control can simplify database management in large, enterprise deployments.
You can also configure Database Control to send email notifications when alerts occur. If you
want to configure this, you must supply a Simple Mail Transfer Protocol (SMTP) or outgoing
mail server and an email address. You can also enable daily backups here. You must supply a
backup start time as well as operating system user credentials for this option.
If you want to use Grid Control to manage your database but have not yet installed and
configured a Grid Control server, do not click either of the management methods. After making
your choices, click the Next button to continue.
Passwords for Database Schema Owners
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VDatabase f er Schema Owners
Passwords i ofor
ic -tran s
b r
Fa non Credentials screen appears next. You must supply passwords for the user accounts
The Database
created by the DBCA when configuring your database. You can use the same password for all of
these privileged accounts by clicking the Use the Same Password for All Accounts option
button. Enter your password in the Password field, and then enter it again in the Confirm
Password field.
Alternatively, you may choose to set different passwords for the privileged users. To do this,
click the Use Different Passwords option button, and then enter your password in the Password
field, and then enter it again in the Confirm Password field. Repeat this for each user listed in
the User Name column. Click the Next button to continue.
Storage Options for Database Files
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V forsDatabase
f
i c i
Storage Optionso a n Files
r r
-tOptions screen, you must select the storage medium where your shared database
F b Storage
Onathe
n o n
files are stored. Your three choices are:
• Cluster File System
• Automatic Storage Management (ASM)
• Raw Devices
If you click the Cluster File System option button, you can click the Next button to continue.
If you click the Automatic Storage Management (ASM) option button, you can either use an
existing ASM disk group or specify a new disk group to use. If there is no ASM instance on any
of the cluster nodes, the DBCA displays the Create ASM Instance screen for you. If an ASM
instance exists on the local node, the DBCA displays a dialog box prompting you to enter the
password for the SYS user for ASM. To initiate the creation of the required ASM instance, enter
the password for the SYS user of the ASM instance. After you enter the required information,
click Next to create the ASM instance. After the instance is created, the DBCA proceeds to the
ASM Disk Groups screen. If you have just created a new ASM instance, there is no disk group
from which to select, so you must create a new one by clicking Create New to open the Create
Disk Group screen.
Storage Options for Database Files (continued)
After you are satisfied with the ASM disk groups available to you, select the one that you
want to use for your database files, and click Next to proceed to the Database File Locations
screen.
If you have configured raw devices, click the corresponding button. You must provide a
fully qualified mapping file name if you did not previously set the DBCA_RAW_CONFIG
environment variable to point to it. You can enter your response or click the Browse button
to locate it. The file should follow the format of the example below:
system=/dev/vg_name/rdbname_system_raw_500m
sysaux=/dev/vg_name/rdbname_sysaux_raw_800m
...
redo2_2=/dev/vg_name/rdbname_redo2_2_raw_120m
control1=/dev/vg_name/rdbname_control1_raw_110m
control2=/dev/vg_name/rdbname_control2_raw_110m
s a
spfile=/dev/vg_name/rdbname_spfile_raw_5m
)h a
pwdfile=/dev/vg_name/rdbname_pwdfile_raw_5m
m ฺbr
where VG_NAME is the volume group (if configured) and rdbname is i ฺ c o database
the d e ฺname.
redAutomatic i
u Storage
Because this example uses a preexisting ASM disk group, clickcthe
i
s den t G
Management button, and then the Next button to continue.
@
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
ASM Disk Groups
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
o
ASM Disk iGroups
ic -tran
b
Onathe
rASM
non
Disk Groups screen, you can choose an existing disk group to be used for your
Fshared storage. To select a disk group, choose the Select check box corresponding to the disk
group that you want to use. Alternatively, you can also create a new disk group to be used by
your cluster database. When you are finished, click the Next button to continue.
Database File Locations
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VLocations f er
i o
Database File
ic -tran s
b
Onathe
rDatabase
non
File Locations screen, you must indicate where the database files are created.
FYou can choose to use a standard template for file locations, one common location, or Oracle
Managed Files (OMF). This cluster database uses Oracle-managed files. Therefore, select the
Use Oracle-Managed Files option button, and enter the disk group name preceded with “+” in
the Database Area field. Alternatively, you can click the Browse button to indicate the location
where the database files are to be created. When you have made your choices, click the Next
button to continue.
Recovery Configuration
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
o
Recovery iConfiguration
ic -tran
b
Onathe
rRecovery
non
Configuration screen, you can select redo log archiving by selecting Enable
FArchiving. If you are using ASM or cluster file system storages, you can also select the Flash
Recovery Area size on the Recovery Configuration screen. The size of the area defaults to 2048
megabytes, but you can change this figure if it is not suitable for your requirements. If you are
using ASM and a single disk group, the flash recovery area defaults to the ASM Disk Group. If
more than one disk group has been created, you can specify it here. If you use a cluster file
system, the flash recovery area defaults to $ORACLE_BASE/flash_recovery_area. You
may also define your own variables for the file locations if you plan to use the Database Storage
screen to define individual file locations.
When you have completed your entries, click Next, and the Database Content screen is
displayed.
Database Content
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Database Content
ic -tran
b
Onathe
rDatabase
non
Content screen, you can choose to install the Sample Schemas included with
Fthe database distribution. On the Custom Scripts tabbed page, you can choose to run your own
scripts as part of the database creation process. When finished, click the Next button to continue
to the next page.
Initialization Parameters
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i
Initialization o Parameters
ic -tran
b
Onathe
rInitialization
non
Parameters screen, you can set important database parameters. The
F
parameters are grouped on four tabs:
• Memory
• Sizing
• Character Sets
• Connection Mode
On the Memory tabbed page, you can set parameters that deal with memory allocation,
including shared pool, buffer cache, Java pool, large pool, and PGA size. Automatic Memory
Management is the preferred memory management method and can be selected here. On the
Sizing tab, you can adjust the database block size. Note that the default is 8 KB. In addition, you
can set the number of processes that can connect simultaneously to the database.
By clicking the Character Sets tab, you can change the database character set. You can also
select the default language and the date format. On the Connection Mode tabbed page, you can
choose the connection type that clients use to connect to the database. The default type is
Dedicated Server Mode. If you want to use Oracle Shared Server, click the Shared Server Mode
button. If you want to review the parameters that are not found in the four tabs, click the All
Initialization Parameters button. Click the Use Automatic Memory Management button, and
then click the Next button to continue.
Security Settings and Maintenance Tasks
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V and f r
eMaintenance
i c i o
Security Settings
a n s Tasks
r r
-tSettings page you can choose between enhanced Oracle 11g security and the
F b Security
Onathe
n o n
pre-11g model (which requires slightly less overhead). Make your choice here and click Next to
continue. Click Enable automatic maintenance tasks to schedule metrics collection and
advisor runs using predefined maintenance windows and resource consumption models.
Then click Next to continue.
Database Storage Options
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Optionsf er
i o
Database Storage
ic -tran s
b r
Fa non Storage screen provides full control over all aspects of database storage, including
The Database
table spaces, data files, and log members. Size, location, and all aspects of extent management
are under your control here.
When finished, click the Next button to continue to the next page.
Create the Database
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Create theio
Database
b r ic -tran
Fa nonOptions screen appears next. You can choose to create the database, or save your
The Creation
DBCA session as a database creation script by clicking the corresponding button. Select the
Create Database check box, and then click the Finish button. The DBCA displays the Summary
screen, giving you the last chance to review all options, parameters, and so on that have been
chosen for your database creation.
Review the summary data. When you are ready to proceed, close the Summary screen by
clicking the OK button.
Monitor Progress
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Monitor Progress
ic -tran
b r
Fa nonMonitor screen appears next. In addition to informing you about how fast the
The Progress
database creation is taking place, it also informs you about the specific tasks being performed by
the DBCA in real time. When the database creation progress reaches 100 percent, the DBCA
displays a dialog box announcing the completion of the creation process. It also directs you to
the installation log file location, parameter file location, and Enterprise Manager URL. By
clicking the Password Management button, you can manage the database accounts created by
the DBCA.
Postinstallation Tasks
• Verify the cluster database configuration.

$ srvctl config database -d racdb
vx0306 racdb1 /u01/app/.../db_1
• Back up the root.sh script.

$ cd $ORACLE_HOME s a
$ cp root.sh root.sh.bak )h a
m ฺbr
• Back up the voting disk. i ฺ c o deฺ
r e d Gui
$ dd if=/dev/sdb6 of=/RACdb/OCR/backup/vdisk.bak
@ sic dent
• Download and install the required e r de updates.
patch S tu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Tasks f er
i o
Postinstallation
ic -tran s
b r
a cluster
After
F the
n o n database has been successfully created, run the following command to verify
the Oracle Cluster Registry configuration in your newly installed RAC environment:
$ srvctl config database -d db_name
Server Control (SRVCTL) displays the name of the node and the instance for the node. The
following example shows two nodes, vx0306 and vx0313 running instances named racdb1
and racdb2 respectively. Execute the following command:
$ srvctl config database -d racdb
It is also recommended that you back up the root.sh script after you complete an installation.
If you install other products in the same Oracle Home directory, the OUI updates the contents of
the existing root.sh script during the installation. If you require information contained in the
original root.sh script, you can recover it from the root.sh file copy.
After your Oracle Database 11g RAC installation is complete and after you are sure that your
system is functioning properly, make a backup of the contents of the voting disk by using the dd
utility.
Note: the SRVCTL utility is covered in more detail in the next lesson.
Check Managed Targets
http://vx0314.us.oracle.com:4889/em
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Targetsf er
i o
Check Managed
ic -tran s
b r
Fa npost-installation
Another
on task you should perform if you are using Grid Control is to check that
all the managed nodes and their managed resources are properly registered and available. Open a
browser and enter the address for your Grid Control console. Click the Targets tab to verify that
all the targets appear here.
Single Instance to RAC Conversion
• Single-instance databases can be

converted to RAC using:
– DBCA
– Enterprise Manager Single-instance
database
– RCONFIG utility
• DBCA automates most of the
s a
conversion tasks. a
)h br
• Before conversion, ensure that: mฺ
RAC database
c o deฺ
– Your hardware and operating i ฺ
d Gui
r e
system are supported
@ sic dent
– Your cluster nodes have accessrto
e de Stu
shared storage v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V to RAC f erConversion
i o
Single Instance
ic -tran s
b r
Fa nuse
You can
onthe Database Configuration Assistant (DBCA) to convert from single-instance
Oracle databases to RAC. The DBCA automates the configuration of the control file attributes,
creates the undo tablespaces and the redo logs, and makes the initialization parameter file entries
for cluster-enabled environments. It also configures Oracle Net Services, Oracle Clusterware
resources, and the configuration for RAC database management for use by Oracle Enterprise
Manager or the SRVCTL utility.
Before you use the DBCA to convert a single-instance database to a RAC database, ensure that
your system meets the following conditions:
• It is a supported hardware and operating system software configuration.
• It has shared storage: Either Oracle Cluster File System or ASM is available and accessible
from all nodes.
• Your applications have no design characteristics that preclude their use with cluster
database processing.
If your platform supports a cluster file system, then you can use it for RAC. You can also
convert to RAC and use a nonshared file system. In either case, it is recommended that you use
the Oracle Universal Installer (OUI) to perform an Oracle Database 11g installation that sets up
the Oracle Home and inventory in an identical location on each of the selected nodes in your
cluster.
Single-Instance Conversion Using the DBCA
Conversion steps for a single-instance database on

nonclustered hardware:
1. Back up the original single-instance database.
2. Perform the preinstallation steps.
3. Set up and validate the cluster.
4. Copy the preconfigured database image. a
h as
5. Install the Oracle Database 11g software with Real r)
Application Clusters. m ฺb
co eฺ d iฺ uid
e
icr ent G
s
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Conversionf er Using DBCA
i o
Single-Instance n s
b ric from
Toaconvert - t r asingle-instance
non
a Oracle database that is on a noncluster computer to a RAC
F
database, perform the steps outlined below, and in the order shown:
3. Set up the cluster.
4. Validate the cluster.
5. Copy the preconfigured database image.
6. Install the Oracle Database 11g software with Real Application Clusters.
Conversion Steps
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Conversion i oSteps
an single-instance database.
1. a b ricup the-toriginal
r
non
Back
F
Use the DBCA to create a preconfigured image of your single-instance database by using the
following procedure:
1. Navigate to the bin directory in $ORACLE_HOME, and start the DBCA.
2. On the Welcome screen, click Next.
3. On the Operations screen, select Manage Templates, and click Next.
4. On the Template Management screen, select “Create a database” template and “From an
existing database (structure as well as data),” and click Next. On the Source Database
screen, enter the database name in the Database instance field, and click Next.
5. On the Template Properties screen, enter a template name in the Name field. By default, the
template files are generated in the ORACLE_HOME/assistants/dbca/templates
directory. Enter a description of the file in the Description field, and change the template
file location in the Template data file field if you want. When finished, click Next.
6. On the Location of Database Related Files screen, select “Maintain the file locations,” so
that you can restore the database to the current directory structure, and click Finish. The
DBCA generates two files: a database structure file (template_name.ctl) and a database
preconfigured image file (template_name.dfb).
Conversion Steps

– Tasks include kernel parameter configuration, hardware
setup, network configuration, and shared storage setup.
– Create a cluster with the required number of nodes according
to your hardware vendor’s documentation.
– Validate cluster components before installation. s a
h a
– Install Oracle Clusterware. ฺ b r)
om deฺ
ฺi ccluvfy.
– Validate the completed cluster installation using i r ed t Gu
4. Copy the preconfigured database image. i
s denc
– Copy the preconfigured database
@
e Sincluding:
dimage tu
e r
— The database structurev*.dbc i lav file e t his
The preconfigured ic i o_ o uimage
database s *.dfb file
—
r
ab nse t
( f
e rde lCopyright
l a V ab e
i
V (continued)
f er
Conversion i oSteps
an s
2. a b ric the-tpreinstallation
r
F Perform
n o n steps.
Several tasks must be completed before Oracle Clusterware and Oracle Database 11g software
can be installed. Some of these tasks are common to all Oracle database installations and should
be familiar to you. Others are specific to Oracle RAC 11g. You can review these tasks by
referring to the lesson titled “Oracle Clusterware Installation and Configuration.”
Form a cluster with the required number of nodes according to your hardware vendor’s
documentation. When you have configured all of the nodes in your cluster, validate cluster
components by using the cluvfy utility, and then install Oracle Clusterware. When the
clusterware is installed, validate the completed cluster installation and configuration using the
Cluster Verification Utility, cluvfy.
4. Copy the preconfigured database image.
This includes copying the database structure *.dbc file and the database preconfigured image
*.dfb file that the DBCA created in step one (Back Up the Original Single-Instance Database)
to a temporary location on the node in the cluster from which you plan to run the DBCA.
Conversion Steps
5. Install the Oracle Database 11g software with RAC.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V (continued)
f er
Conversion i oSteps s
an Database 11g software with RAC.
5. a b ricthe Oracle
- t r
non
Install
F1. Run the OUI to perform an Oracle database installation with RAC. Select Cluster
Installation Mode on the Specify Hardware Cluster Installation screen of the OUI, and
select the nodes to include in your RAC database.
2. On the OUI Database Configuration Types screen, select “Advanced install.” After
installing the software, the OUI runs postinstallation tools such as NETCA, DBCA, and so
on.
3. On the DBCA Template Selection screen, use the template that you copied to a temporary
location in the “Copy the Preconfigured Database Image” step. Use the browse option to
select the template location.
4. If you selected raw storage on the OUI Storage Options screen, then on the DBCA File
Locations tab of the Initialization Parameters screen, replace the data files, control files, and
log files, and so on, with the corresponding raw device files if you did not set the
DBCA_RAW_CONFIG environment variable. You must also replace default database files
with raw devices on the Storage page.
5. After creating the RAC database, the DBCA displays the Password Management screen on
which you must change the passwords for database privileged users who have SYSDBA and
SYSOPER roles. When the DBCA exits, the conversion process is complete.
Single-Instance Conversion Using rconfig
1. Edit the ConvertToRAC.xml file located in the

$ORACLE_HOME/assistants/rconfig/sampleXMLs
directory.
2. Modify the parameters in the ConvertToRAC.xml file as
required for your system.
3. Save the file under a different name.
s a
)h a
ฺb r
$ cd $ORACLE_HOME/assistants/rconfig/sampleXMLs
c m
o deฺ
$ vi ConvertToRAC.xml i ฺ
d Gui
r e
$ rconfig my_rac_conversion.xml sic nt
r d e@ Stude
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Conversion
f er Using rconfig
i o
Single-Instance
ic -tran s
b r
a use
You
F can
n o nthe command-line utility rconfig to convert a single-instance database to RAC.
To use this feature, complete the following steps:
1. Go to the $ORACLE_HOME/assistants/rconfig/sampleXMLs directory as the
oracle user, and open the ConvertToRAC.xml file using a text editor, such as vi.
2. Review the ConvertToRAC.xml file, and modify the parameters as required for your
system. The XML sample file contains comment lines that provide instructions for how to
configure the file.
When you have completed making changes, save the file with the syntax filename.xml.
Make a note of the name you select.
3. Assuming that you save your XML file as my_rac_conversion.xml, navigate to the
$ORACLE_HOME/bin directory, and use the following syntax to run the rconfig
command:
$ rconfig my_rac_conversion.xml
Note: The Convert verify option in the ConvertToRAC.xml file has three options:
• Convert verify="YES": rconfig performs checks to ensure that the prerequisites
for single-instance to RAC conversion have been met before it starts conversion.
Single-Instance Conversion Using rconfig (continued)
• Convert verify="NO": rconfig does not perform prerequisite checks; it starts
conversion.
• Convert verify="ONLY": rconfig performs only prerequisite checks; it does not
start conversion after completing prerequisite checks.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Single-Instance Conversion Using Grid Control
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Conversion
f er Using Grid Control
i o
Single-Instance s
anthe DBCA and rconfig for single-instance conversion, you can also use
In a b ric to using
addition - t r
F non
Enterprise Manager Grid Control to convert a single-instance database to RAC. To use this
feature of Grid Control, complete the following steps:
1. Log in to Grid Control. From the Grid Control Home page, click the Targets tab.
2. On the Targets page, click the Databases secondary tab, and click the link in the Names
column of the database that you want to convert to RAC.
3. On the Database Instance Home page, click the Administration secondary tab.
4. On the Administration page, in the Database Administration Change Database section, click
Convert to Cluster Database.
5. Log in as the database user SYS with SYSDBA privileges to the database you want to
convert, and click Next.
6. On the Convert to Cluster Database: Cluster Credentials page, provide a username and
password for the oracle user and password of the target database that you want to convert.
If the target database is using ASM, then provide the ASM SYS user and password and
click Next.
7. On the Hosts page, select the host nodes in the cluster that you want to be cluster members
in the RAC database installed. When you have completed your selection, click Next.
Single-Instance Conversion Using Grid Control (continued)
8. On the Convert to Database: Options page, select whether you want to use the existing
listener and port number or specify a new listener and port number for the cluster. Also,
provide a prefix for the cluster database instances. When you have finished entering
information, click Next.
9. On the Convert to Cluster Database: Shared Storage page, either select the option to use
your existing shared storage area, or select the option to have your database files copied to
a new shared storage location. Also, decide whether you want to use your existing flash
recovery area, or if you want to copy your flash recovery files to a new area using Oracle
Managed Files. When you have finished entering information, click Next.
10. On the Convert to Cluster Database: Review page, review the options you have selected.
Click Submit Job to proceed with the conversion.
11. On the Confirmation page, click View Job to check the status of the conversion.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Summary

• Install the Oracle Enterprise Manager agent
• Create a cluster database
• Perform required tasks after database creation
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non

• Installing the Enterprise Manager agent on each cluster
node
• Confirming that the services needed by the database
creation process are running
• Creating a cluster database
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
RAC Database Administration
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Use Enterprise Manager cluster database pages
• Define redo log files in a RAC environment
• Define undo tablespaces in a RAC environment
• Start and stop RAC databases and instances
• Modify initialization parameters in a RAC environment s a
h a
• Manage ASM instances in a RAC environment ฺbr) m eฺ
ฺ c o
r e di Guid
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Cluster Database Home Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Home f erPage
i o
Cluster Database
ic -tran s
b r
Fa nonDatabase home page serves as a crossroad for managing and monitoring all aspects
The Cluster
of your RAC database. From this page, you can access the other main cluster database tabs:
Performance, Availability, Server, Schema, Data Movement, Software and Support, and
Topology.
On this page, you find General, High Availability, Space Summary, and Diagnostic Summary
sections for information that pertains to your cluster database as a whole. The number of
instances is displayed for the RAC database, in addition to the status. A RAC database is
considered to be up if at least one instance has the database open. You can access the Cluster
home page by clicking the Cluster link in the General section of the page.
Other items of interest include the date of the last RMAN backup, archiving information, space
utilization, and an alert summary. By clicking the link next to the Flashback Database Logging
label, you can go to the Recovery Settings page from where you can change various recovery
parameters.
The Alerts table shows all open recent alerts. Click the alert message in the Message column for
more information about the alert. When an alert is triggered, the name of the metric for which
the alert was triggered is displayed in the Name column.
Cluster Database Home Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Home f erPage (continued)
i o
Cluster Database
ic -tran s
b r
Fa nonAlerts table provides information about alerts for related targets, such as Listeners
The Related
and Hosts, and contains details about the message, the time the alert was triggered, the value,
and the time the alert was last checked.
The Policy Trend Overview page (accessed by clicking the Compliance Score link) provides a
comprehensive view about a group or targets containing other targets with regard to compliance
over a period of time. Using the tables and graphs, you can easily watch for trends in progress
and changes.
The Security At a Glance page shows an overview of the security health of the enterprise for all
the targets or specific groups. This helps you to quickly focus on security issues by showing
statistics about security policy violations and noting the critical security patches that have not
been applied.
The Job Activity table displays a report of the job executions that shows the scheduled, running,
suspended, and problem (stopped/failed) executions for all Enterprise Manager jobs on the
cluster database.
The Instances table lists the instances for the cluster database, their availability, alerts, policy
violations, performance findings, and related ASM Instance. Click an instance name to go to the
home page for that instance. Click the links in the table to get more information about a
particularreproduction
Unauthorized alert, advice, or
or metric.
distribution prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
Cluster Database Instance Home Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Instance
f er Home Page
i o
Cluster Database
ic -tran s
b r
Fa nonDatabase Instance home page enables you to view the current state of the instance
The Cluster
by displaying a series of metrics that portray its overall health. This page provides a launch point
for the performance, administration, and maintenance of the instance environment.
You can access the Cluster Database Instance home page by clicking one of the instance names
from the Instances section of the Cluster Database home page. This page has basically the same
sections as the Cluster Database home page.
The difference is that tasks and monitored activities from these pages apply primarily to a
specific instance. For example, clicking the Shutdown button from this page shuts down only
this one instance. However, clicking the Shutdown button from the Cluster Database home page
gives you the option of shutting down all or specific instances.
By scrolling down on this page, you see the Alerts, Related Alerts, Policy Violations, Jobs
Activity, and Related Links sections. These provide similar information similar to that provided
in the same sections in the Cluster Database home page.
Cluster Database Instance Administration Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Instancef er Administration Page
i o
Cluster Database
ic -tran s
b r
Fa noclassical
This is the n Administration page with an important difference with its corresponding
single-instance version. As you can see it on the screenshot, each database-level related task is
prefixed with a small icon representing a cluster database.
Cluster Home Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VPage sfer
i
Cluster Home o
ic -tran
b r
Fa nshows
The slide
on you the Cluster home page, which can be accessed by clicking the Cluster link
located in the General section of the Cluster Database home page. The cluster is represented as a
composite target composed of nodes and cluster databases. An overall summary of the cluster is
provided here.
The Cluster home page displays several sections, including General, Configuration, Cluster
Databases, Alerts, and Hosts.
The General section provides a quick view of the status of the cluster, providing basic
information such as current Status, Availability, Up nodes, and Clusterware Home and Version.
The Configuration section allows you to view the operating systems (including Hosts and OS
Patches) and hardware (including Hardware configuration and Hosts) for the cluster.
The Cluster Databases table displays the cluster databases (optionally associated with
corresponding services) associated with this cluster, their availability, and any alerts on those
databases. The Alerts table provides information about any alerts that have been issued along
with the severity rating of each.
The Hosts table (not shown on the screenshot) displays the hosts for the cluster, their
availability, corresponding alerts, CPU and memory utilization percentage, and total I/O per
second.
The Configuration Section
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Sectionf er
i o
The Configuration
ic -tran s
b r
Fa nonhome page is invaluable for locating configuration-specific data. Locate the
The Cluster
Configuration section on the Cluster home page. The View drop-down list allows you to inspect
hardware and operating system overview information.
Click the Hosts link, and then click the Hardware Details link of the host that you want. On the
Hardware Details page, you find detailed information regarding your CPU, disk controllers,
network adapters, and so on. This information can be very useful when determining the Linux
patches for your platform.
Click History to access the hardware history information for the host.
Some hardware information is not available, depending on the hardware platform.
Note: The Local Disk Capacity (GB) field shows the disk space that is physically attached
(local) to the host. This value does not include disk space that may be available to the host
through networked file systems.
The Configuration Section
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Sectionf er (continued)
i o
The Configuration
ic -tran s
b r
Fa non System Details General page displays operating system details for a host,
The Operating
including:
• General information, such as the distributor version and the maximum swap space of the
operating system
• Information about operating system properties
The Source column displays where Enterprise Manager obtained the value for each operating
system property.
To see a list of changes to the operating system properties, click History.
The Operating System Details File Systems page displays information about one or more file
systems for the selected hosts:
• Name of the file system on the host
• Type of mounted file system, for example, ufs or nfs
• Directory where the file system is mounted
• The mount options for the file system, for example ro, nosuid, or nobrowse
The Operating System Details Packages page displays information about the operating system
packages that have been installed on a host.
Topology Viewer
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
o
Topology iViewer
ic -tran
b r
Fa noEnterprise
The Oracle n Manager Topology Viewer enables you to visually see the relationships
between target types for each host of your cluster database. You can zoom in or out, pan, and see
selection details. These views can also be used to launch various administration functions.
The Topology Viewer populates icons on the basis of your system configuration. If a listener is
serving an instance, a line connects the listener icon and the instance icon. Possible target types
are:
• Interface
• Listener
• ASM Instance
• Database Instance
If the Show Configuration Details option is not selected, the topology shows the monitoring
view of the environment, which includes general information such as alerts and overall status. If
you select the Show Configuration Details option, additional details are shown in the Selection
Details window, which are valid for any topology view. For instance, the Listener component
would also show the machine name and port number.
You can click an icon and then right-click to display a menu of available actions.
Enterprise Manager Alerts and RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sAlerts f er and RAC
Enterpriseio
Manager
b r ic -tran
Fa nuse
You can
onEnterprise Manager to administer alerts for RAC environments. Enterprise Manager
distinguishes between database- and instance-level alerts in RAC environments.
Enterprise Manager also responds to metrics from across the entire RAC database and publishes
alerts when thresholds are exceeded. Enterprise Manager interprets both predefined and
customized metrics. You can also copy customized metrics from one cluster database instance to
another, or from one RAC database to another. A recent alert summary can be found on the
Database Control home page. Notice that alerts are sorted by relative time and target name.
Enterprise Manager Metrics and RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sMetrics
f er and RAC
EnterpriseioManager
b r ic -tran
Fa non for instance-level alerts, such as archive log alerts, can be set at the instance
Alert thresholds
target level. This enables you to receive alerts for the specific instance if performance exceeds
your threshold. You can also configure alerts at the database level, such as setting alerts for
tablespaces. This enables you to avoid receiving duplicate alerts at each instance.
Enterprise Manager Metrics and RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sMetrics
f er and RAC (continued)
Enterpriseio Manager
b r ic -tran
Fa npossible
It is also
on to view the metric across the cluster in a comparative or overlay fashion. To
view this information, click the Compare Targets link at the bottom of the corresponding metric
page. When the Compare Targets page appears, choose the instance targets that you want to
compare by selecting them and then clicking the Move button. If you want to compare the
metric data from all targets, then click the Move All button. After making your selections, click
the OK button to continue.
The Metric summary page appears next. Depending on your needs, you can accept the default
timeline of 24 hours or select a more suitable value from the View Data drop-down list. If you
want to add a comment regarding the event for future reference, then enter a comment in the
Comment for Most Recent Alert field, and then click the Add Comment button.
Enterprise Manager Alert History and RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sAlert f er History and RAC
Enterpriseio
Manager
ab
In a
ric -tran you can see a summary of the alert history for each participating
non
RAC environment,
Finstance directly from the Cluster Database home page. The drill-down process is shown in the
slide. You click the Alert History link in the Related Links section of the Cluster Database home
page. This takes you to the Alert History page on which you can see the summary for both
instances in the example. You can then click one of the instance’s links to go to the
corresponding Alert History page for that instance. From there, you can access a corresponding
alert page by choosing the alert of your choice.
Enterprise Manager Blackouts and RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sBlackouts
f er
EnterpriseioManager n and RAC
r i c r a
Fabcannuse
You -t
onEnterprise Manager to define blackouts for all managed targets of your RAC
database to prevent alerts from being recorded. Blackouts are useful when performing scheduled
or unscheduled maintenance or other tasks that might trigger extraneous or unwanted events.
You can define blackouts for an entire cluster database or for specific cluster database instances.
To create a blackout event, click the Setup link on top of any Enterprise Manager page. Then,
click the Blackouts link on the left. The Blackouts page appears.
Click the Create button. The Create Blackout: Properties page appears. You must enter a name
or tag in the Name field. If you want, you can also enter a descriptive comment in the Comments
field. This is optional. Enter a reason for the blackout in the Reason field.
In the Targets area of the Properties page, you must choose a target Type from the drop-down
list. In the example in the slide, the entire cluster database RDBB is chosen. Click the cluster
database in the Available Targets list, and then click the Move button to move your choice to the
Selected Targets list. Click the Next button to continue.
The Member Targets page appears next. Expand the Selected Composite Targets tree and ensure
that all targets that must be included appear in the list. Continue and define your schedule as you
normally would.
Redo Log Files and RAC
Node1 Node2
RAC01 RAC02
Shared storage
Group 1 SPFILE
… Group 4
s a
Group 2 RAC01.THREAD=1
)h
Group 5 a
ฺbr
RAC02.THREAD=2
Group 3 …
c m
o Thread e ฺ
i ฺ i d 2
Thread 1
i c red t Gu
@ s den
ALTER DATABASE ADD LOGFILE THREAD 2 GROUP
e r de 4;Stu
ALTER DATABASE ADD LOGFILE THREAD
v i lav2 GROUPe t his5;
ALTER DATABASE ENABLE THREAD
ic i o_2; o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V and sRAC
f er
o
Redo Log iFiles
ic -tran
b r
Fa noApplication
With Real n Clusters (RAC), each instance writes to its own set of online redo log
files, and the redo written by an instance is called a thread of redo, or thread. Thus, each redo log
file group used by an instance is associated with the same thread number determined by the
value of the THREAD initialization parameter. If you set the THREAD parameter to a nonzero value
for a particular instance, the next time the instance is started, it will try to use that thread.
Because an instance can use a thread as long as that thread is enabled, and not in use by another
instance, it is recommended to set the THREAD parameter to a nonzero value with each instance
having different values.
You associate a thread number with a redo log file group by using the ALTER DATABASE ADD
LOGFILE THREAD statement. You enable a thread number by using the ALTER DATABASE ENABLE
THREAD statement. Before you can enable a thread, it must have at least two redo log file groups.
By default, a database is created with one enabled public thread. An enabled public thread is a
thread that has been enabled by using the ALTER DATABASE ENABLE PUBLIC THREAD statement.
Such a thread can be acquired by an instance with its THREAD parameter set to zero. Therefore,
you need to create and enable additional threads when you add instances to your database.
The maximum possible value for the THREAD parameter is the value assigned to the
MAXINSTANCES parameter specified in the CREATE DATABASE statement.
Note: You can use Enterprise Manager to administer redo log groups in a RAC environment.
Automatic Undo Management and RAC
Pending Node1 Node2

offline
RAC01 RAC02
Consistent reads
Transaction recovery
Shared storage
s a
… SPFILE
)h a
undotbs3
RAC01.UNDO_TABLESPACE=undotbs3
m ฺbr
RAC02.UNDO_TABLESPACE=undotbs2
i ฺ c o deฺ
undotbs1
r e dundotbs2 u i
…
sic dent G
@
de Stu
e r
v i lav e this
ALTER SYSTEM SET UNDO_TABLESPACE=undotbs3
ic i o_ o us SID='RAC01';
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Automaticio Undo Management
n in RAC
r i c r a
FabOracle
The
o n -t automatically manages undo segments within a specific undo tablespace
database
n
that is assigned to an instance. Under normal circumstances, only the instance assigned to the
undo tablespace can modify the contents of that tablespace. However, all instances can always
read all undo blocks for consistent-read purposes. Also, any instance can update any undo
tablespace during transaction recovery, as long as that undo tablespace is not currently used by
another instance for undo generation or transaction recovery.
You assign undo tablespaces in your RAC database by specifying a different value for the
UNDO_TABLESPACE parameter for each instance in your SPFILE or individual PFILEs. If you do
not set the UNDO_TABLESPACE parameter, then each instance uses the first available undo
tablespace. If undo tablespaces are not available, the SYSTEM rollback segment is used.
You can dynamically switch undo tablespace assignments by executing the ALTER SYSTEM SET
UNDO_TABLESPACE statement with the SID clause. You can run this command from any instance.
In the example shown in the slide, the previously used undo tablespace assigned to instance
RAC01 remains assigned to it until the RAC01 instance’s last active transaction commits. The
pending offline tablespace may be unavailable for other instances until all transactions against
that tablespace are committed.
Note: You cannot simultaneously use Automatic Undo Management (AUM) and manual undo
management in a RAC database. It is highly recommended that you use the AUM mode.
Starting and Stopping RAC Instances
• Multiple instances can open the same database

simultaneously.
• Shutting down one instance does not interfere with other
running instances.
• SHUTDOWN TRANSACTIONAL LOCAL does not wait for
other instances’ transactions to finish.
s a
• RAC instances can be started and stopped by using:
)h a
– Enterprise Manager m ฺbr
i ฺ c o deฺ
– Server Control (SRVCTL) utility
r e d Gui
– SQL*Plus @ sic dent
• Shutting down a RAC databasevmeans e rde sshutting
S tu down all
instances accessing the database. v i la e thi
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfeRAC r
Starting and
i c i o Stopping
a n Instances
ab
In a
r -t r
F n o n
RAC environment, multiple instances can have the same RAC database open at the same
time. Also, shutting down one instance does not interfere with the operation of other running
instances.
The procedures for starting up and shutting down RAC instances are identical to the procedures
used in single-instance Oracle, with the following exception:
The SHUTDOWN TRANSACTIONAL command with the LOCAL option is useful to shut down an
instance after all active transactions on the instance have either committed or rolled back.
Transactions on other instances do not block this operation. If you omit the LOCAL option, then
this operation waits until transactions on all other instances that started before the shutdown are
issued either a COMMIT or a ROLLBACK.
You can start up and shut down instances by using Enterprise Manager, SQL*Plus, or Server
Control (SRVCTL). Both Enterprise Manager and SRVCTL provide options to start up and shut
down all the instances of a RAC database with a single step.
Shutting down a RAC database mounted or opened by multiple instances means that you need to
shut down every instance accessing that RAC database. However, having only one instance
opening the RAC database is enough to declare the RAC database open.
Starting and Stopping
RAC Instances with SQL*Plus
[stc-raclin01] $ echo $ORACLE_SID
RACDB1
sqlplus / as sysdba
SQL> startup
SQL> shutdown
[stc-raclin02] $ echo $ORACLE_SID

RACDB2
sqlplus / as sysdba
SQL> startup
s a
SQL> shutdown
)h a
OR m ฺbr
i ฺ c o deฺ
[stc-raclin01] $sqlplus / as sysdba
r e d Gui
SQL> startup
@ sic dent
de Stu
SQL> shutdown
SQL> connect sys/oracle@RACDB2 as sysdba e r
SQL> startup
v i lav e this
SQL> shutdown
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfeRAC r
Starting and
i c i o Stopping
a n Instances with SQL*Plus
r r
t up or shut down only one instance, and if you are connected to your local
ab wantoton-start
If you
F
node, you n must first ensure that your current environment includes the SID for the local instance.
To start up or shut down your local instance, initiate a SQL*Plus session connected as SYSDBA
or SYSOPER, and then issue the required command (for example, STARTUP).
You can start multiple instances from a single SQL*Plus session on one node by way of Oracle
Net Services. To achieve this, you must connect to each instance by using a Net Services
connection string (typically an instance-specific alias from your tnsnames.ora file).
For example, you can use a SQL*Plus session on a local node to shut down two instances on
remote nodes by connecting to each using the instance’s individual alias name.
The example in the slide assumes that the alias name for the second instance is RACDB2. In the
example, there is no need to connect to the first instance using its connect descriptor because the
command is issued from the first node with the correct ORACLE_SID.
It is not possible to start up or shut down more than one instance at a time in SQL*Plus, so you
cannot start or stop all the instances for a cluster database with a single SQL*Plus command.
To verify that instances are running, on any node, look at V$ACTIVE_INSTANCES.
Note: SQL*Plus is integrated with Oracle Clusterware to make sure that corresponding
resources are correctly handled during startup and shutdown of instances by using SQL*Plus.
Starting and Stopping
RAC Instances with SRVCTL
• start/stop syntax:
srvctl start|stop instance -d <db_name> -i <inst_name_list>
[-o open|mount|nomount|normal|transactional|immediate|abort>]
[-c <connect_str> | -q]
srvctl start|stop database -d <db_name>

[-o open|mount|nomount|normal|transactional|immediate|abort>]
[-c <connect_str> | -q]
s a
)h a
• Examples:
m ฺbr
i ฺ c o deฺ
$ srvctl start instance -d RACDB -i RACDB1,RACDB2
r e d Gui
$ srvctl stop instance -d RACDB -i RACDB1,RACDB2 @ sic dent
e r de Stu
$ srvctl start database -d RACDBa-o
v i l v open e t his
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfeRAC r
Starting and
i c i o Stopping
a n Instances with SRVCTL
r r
-t database command starts a cluster database, its enabled instances, and its
Fabsrvctl
The
n o nstart
services. The srvctl stop database command stops a database, its instances, and its
services.
The srvctl start instance command starts instances of a cluster database. This command
also starts all enabled and nonrunning services that have the listed instances either as preferred
or as available instances.
The srvctl stop instance command stops instances as well as all enabled and running
services that have these instances as either preferred or available instances.
You must disable an object that you intend to keep stopped after you issue a srvctl stop
command; otherwise, Oracle Clusterware (OC) can restart it as a result of another planned
operation. For commands that use a connect string, if you do not provide a connect string, then
SRVCTL uses / as sysdba to perform the operation. The –q option asks for a connect string
from standard input. SRVCTL does not support concurrent executions of commands on the same
object. Therefore, run only one SRVCTL command at a time for each database, service, or other
object. |
To use the START and STOP options of the SRVCTL command, your service must be an
OC-enabled, nonrunning service.
Note: For more information, see the Oracle Clusterware Administration and Deployment Guide
and the Oracle Real Application Clusters Administration and Deployment Guide.
Switch Between the Automatic
and Manual Policies
$ srvctl config database -d RACB -a

ex0044 RACB1 /u01/app/oracle/product/10.2.0/db_1
ex0045 RACB2 /u01/app/oracle/product/10.2.0/db_1
DB_NAME: RACB
ORACLE_HOME: /u01/app/oracle/product/10.2.0/db_1
SPFILE: +DGDB/RACB/spfileRACB.ora
DOMAIN: null
DB_ROLE: null
s a
START_OPTIONS: null h a
POLICY: AUTOMATIC ฺ b r)
c o m eฺ
ENABLE FLAG: DB ENABLED
i ฺ d
$ red Gui
@ sic dent
e Stu
rdMANUAL;
e
la e this
srvctl modify database -d RACBv-y
v i
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V thesAutomatic
f
i c i
Switch Between o a n and Manual Policies
r r
F b noOracle
Byadefault, n-t Clusterware is configured to start the VIP, listener, instance, ASM, database
services, and other resources during system boot. It is possible to modify some resources to have
their AUTO_START profile parameter set to the value 2. This means that after node reboot, or
when Oracle Clusterware is started, resources with AUTO_START=2 need to be started manually
via srvctl. This is designed to assist in problem troubleshooting and system maintenance.
Starting with Oracle Database 10g Release 2, when changing resource profiles through srvctl,
the command tool automatically modifies the profile attributes of other dependent resources
given the current prebuilt dependencies. The command to accomplish this is:
srvctl modify database -d <dbname> -y AUTOMATIC|MANUAL
To implement Oracle Clusterware and Real Application Clusters, it is best to have Oracle
Clusterware start the defined Oracle resources during system boot, which is the default. The first
example in the slide uses the srvctl config database command to display the current
policy for the RACB database. As you can see, it is currently set to its default: AUTOMATIC. The
second statement uses the srvctl modify database command to change the current policy
to MANUAL for the RACB database. When you add a new database by using the srvctl add
database command, that database is by default placed under the control of Oracle Clusterware
using the AUTOMATIC policy. However, you can use the following statement to directly set the
policy to MANUAL: srvctl add database -d RACZ -y MANUAL.
Note: You can also use this procedure to configure your system to prevent Oracle Clusterware
from autorestarting
Unauthorized failed
reproduction ordatabase instances
distribution more than
prohibitedฺ once.
RAC Initialization Parameter Files
• An SPFILE is created if you use the DBCA.

• The SPFILE must be created on a shared volume or shared
raw device.
• All instances use the same SPFILE.
• If the database is created manually, then create an SPFILE
from a PFILE. a
as
Node1 Node2 )h ฺbr
c m
o deฺ
RAC01 i ฺ
RAC02
d Gui
r e
initRAC01.ora
@ sic deinitRAC02.ora
nt
SPFILE=…
r d e Stu SPFILE=…
l
SPFILE
i a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sferFiles
i
Initialization o Parameter
ic -tran
b r
Fa nocreate
When you n the database, the DBCA creates an SPFILE in the file location that you specify.
This location can be an Automatic Storage Management (ASM) disk group, cluster file system
file, or a shared raw device. If you manually create your database, then it is recommended to
create an SPFILE from a PFILE.
All instances in the cluster database use the same SPFILE at startup. Because the SPFILE is a
binary file, do not edit it. Instead, change the SPFILE parameter settings by using Enterprise
Manager or ALTER SYSTEM SQL statements.
RAC uses a traditional PFILE only if an SPFILE does not exist or if you specify PFILE in your
STARTUP command. Using SPFILE simplifies administration, maintaining parameter settings
consistent, and guarantees parameter settings persistence across database shutdown and startup.
In addition, you can configure RMAN to back up your SPFILE.
In order for each instance to use the same SPFILE at startup, each instance uses its own PFILE
file that contains only one parameter called SPFILE. The SPFILE parameter points to the shared
SPFILE on your shared storage. This is illustrated in the slide. By naming each PFILE using the
init<SID>.ora format, and by putting them in the $ORACLE_HOME/dbs directory of each node,
a STARTUP command uses the shared SPFILE.
SPFILE Parameter Values and RAC
• You can change parameter settings using the ALTER

SYSTEM SET command from any instance:
ALTER SYSTEM SET <dpname> SCOPE=MEMORY sid='<sid|*>';
• SPFILE entries such as:
– *.<pname> apply to all instances
– <sid>.<pname> apply only to <sid> a
– <sid>.<pname> takes precedence over *.<pname> as
ฺb )h
r
• Use current or future *.<dpname> settings for <sid>:
o m ฺ
i ฺ c i d e
ALTER SYSTEM RESET <dpname> SCOPE=MEMORY rsid='<sid>';
i c ed t Gu
@ s den
• Remove an entry from your SPFILE: e r de Stu
v i lav e this
ALTER SYSTEM RESET <dpname>
ic i o_ oSCOPE=SPFILE
u s sid='<sid|*>';
( f abr nse t
e rde lCopyright
l a V ab e
i
V Values f er and RAC
o
SPFILE Parameter
i
ic -tran s
b r
a modify
You
F can
n o n the value of your initialization parameters by using the ALTER SYSTEM SET
command. This is the same as with a single-instance database except that you have the
possibility to specify the SID clause in addition to the SCOPE clause.
By using the SID clause, you can specify the SID of the instance where the value takes effect.
Specify SID='*' if you want to change the value of the parameter for all instances. Specify
SID='sid' if you want to change the value of the parameter only for the instance sid. This
setting takes precedence over previous and subsequent ALTER SYSTEM SET statements that
specify SID='*'. If the instances are started up with an SPFILE, then SID='*' is the default if
you do not specify the SID clause.
If you specify an instance other than the current instance, then a message is sent to that instance
to change the parameter value in its memory if you are not using the SPFILE scope.
The combination of SCOPE=MEMORY and SID='sid' of the ALTER SYSTEM RESET command
allows you to override the precedence of a currently used <sid>.<dparam> entry. This allows
for the current *.<dparam> entry to be used, or for the next created *.<dparam> entry to be
taken into account on that particular sid.
Using the last example, you can remove a line from your SPFILE.
EM and SPFILE Parameter Values
SCOPE=MEMORY
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Parameter
f er Values
EM and SPFILE
i o
ic -tran s
b r
Fa naccess
You can
on the Initialization Parameters page by clicking the Initialization Parameters link
on the Cluster Database Server page.
The Current tabbed page shows you the values currently used by the initialization parameters of
all the instances accessing the RAC database. You can filter the Initialization Parameters page to
show only those parameters that meet the criteria of the filter that you entered in the Name field.
The Instance column shows the instances for which the parameter has the value listed in the
table. An asterisk (*) indicates that the parameter has the same value for all remaining instances
of the cluster database.
Choose a parameter from the Select column and perform one of the following steps:
• Click Add to add the selected parameter to a different instance. Enter a new instance name
and value in the newly created row in the table.
• Click Reset to reset the value of the selected parameter. Note that you may reset only those
parameters that do not have an asterisk in the Instance column. The value of the selected
column is reset to the value of the remaining instances.
Note: For both Add and Reset buttons, the ALTER SYSTEM command uses SCOPE=MEMORY.
EM and SPFILE Parameter Values
SCOPE=BOTH
s a
)h a
SCOPE=SPFILE m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Parameter
f er Values (continued)
EM and SPFILE
i o
ic -tran s
b r
Fa notabbed
The SPFile n page displays the current values stored in your SPFILE.
As on the Current tabbed page, you can add or reset parameters. However, if you select the
“Apply changes in SPFile mode” check box, then the ALTER SYSTEM command uses
SCOPE=BOTH. If this check box is not selected, SCOPE=SPFILE is used.
Click Apply to accept and generate your changes.
RAC Initialization Parameters
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Parameters
f er
i o
RAC Initialization s
ic -tran Enables a database to be started in cluster mode. Set this to TRUE.
b r
Fa non
CLUSTER_DATABASE:
CLUSTER_DATABASE_INSTANCES: Sets the number of instances in your RAC environment.
A proper setting for this parameter can improve memory use.
CLUSTER_INTERCONNECTS: Specifies the cluster interconnect when there is more than one
interconnect. Refer to your Oracle platform-specific documentation for the use of this parameter,
its syntax, and its behavior. You typically do not need to set the CLUSTER_INTERCONNECTS
parameter. For example, do not set this parameter for the following common configurations:
• If you have only one cluster interconnect
• If the default cluster interconnect meets the bandwidth requirements of your RAC database,
which is typically the case
• If NIC bonding is being used for the interconnect
• When OIFCFG’s global configuration can specify the right cluster interconnects. It only
needs to be specified as an override for OIFCFG.
DB_NAME: If you set a value for DB_NAME in instance-specific parameter files, then the setting
must be identical for all instances.
DISPATCHERS: Set this parameter to enable a shared-server configuration, that is a server that
is configured to allow many user processes to share very few server processes.
RAC Initialization Parameters (continued)
With shared-server configurations, many user processes connect to a dispatcher. The
DISPATCHERS parameter may contain many attributes. Oracle recommends that you configure
at least the PROTOCOL and LISTENER attributes.
PROTOCOL specifies the network protocol for which the dispatcher process generates a listening
end point. LISTENER specifies an alias name for the Oracle Net Services listeners. Set the alias
to a name that is resolved through a naming method, such as a tnsnames.ora file.
MAX_COMMIT_PROPAGATION_DELAY: This is a RAC-specific parameter. Starting with
Oracle Database 10g Release 2, the MAX_COMMIT_PROPAGATION_DELAY parameter is
deprecated. By default, commits on one instance are immediately visible on all of the other
instances (broadcast on commit propagation). This parameter is retained for backward
compatibility only. This parameter specifies the maximum amount of time allowed before the
system change number (SCN) held in the System Global Area (SGA) of an instance is refreshed
by the log writer process (LGWR). It determines whether the local SCN should be refreshedsfrom a
) a
h not
the SGA when getting the snapshot SCN for a query. With previous releases, you should
ฺ b r
alter the default setting for this parameter except under a limited set of circumstances.
c o m e ฺ For
example, under unusual circumstances involving rapid updates and queries ฺ
di ofGthe u idsame data
r e
from different instances, the SCN might not be refreshed in a timely
@ sic dmanner.
e nt
SPFILE: When you use an SPFILE, all RAC databasedinstances
r e Stmust u use the SPFILE and the
file must be on shared storage.
i l a ve this
THREAD: If specified, this parameter must o v unique
_have s e values on all instances. The THREAD
i
c thread
riredo u
parameter specifies the number of the
a b e to to be used by an instance. You can specify any
available redo thread number e as(flong asenthat
rd le lic
s thread number is enabled and is not used.
e
V ab
i l a
V sfer
i o
ic -tran
b r
Fa non
Parameters That Require Identical Settings
• ACTIVE_INSTANCE_COUNT
• ARCHIVE_LAG_TARGET
• COMPATIBLE
• CLUSTER_DATABASE/CLUSTER_DATABASE_INSTANCE
• CONTROL_FILES
• DB_BLOCK_SIZE
• DB_DOMAIN
• DB_FILES
• DB_NAME a
• DB_RECOVERY_FILE_DEST/DB_RECOVERY_FILE_DEST_SIZE h as
• DB_UNIQUE_NAME ฺ b r)
• c o m eฺ
INSTANCE_TYPE ฺ
di Guid
• r e
sic dent
PARALLEL_MAX_SERVERS
• REMOTE_LOGIN_PASSWORD_FILE @
• TRACE_ENABLED e r de Stu
• UNDO_MANAGEMENT v i lav e this
_ io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
V sfer Identical Settings
Parameters i oThat Require
ic -tran
b r
Fa initialization
Certain
non
parameters that are critical at database creation or that affect certain
database operations must have the same value for every instance in RAC. Specify these
parameter values in the SPFILE, or in each init_dbname.ora file on each instance. In the list
in the slide, each parameter must have the same value on all instances.
Note: The setting for DML_LOCKS and RESULT_CACHE_MAX_SIZE must be identical on every
instance only if set to zero. Disabling the result cache on some instances may lead to incorrect
results.
Parameters That Require Unique Settings
Instance settings:
• THREAD • INSTANCE_NUMBER
• ROLLBACK_SEGMENTS • UNDO_TABLESPACE
• INSTANCE_NAME
• ASM_PREFERRED_READ_FAILURE_GROUPS
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer Unique Settings
Parameters i oThat Require
ic -tran
b r
Fa nothen THREAD or ROLLBACK_SEGMENTS parameters, it is recommended that you set
If you use
unique values for them by using the SID identifier in the SPFILE. However, you must set a
unique value for INSTANCE_NUMBER for each instance and you cannot use a default value.
The Oracle server uses the INSTANCE_NUMBER parameter to distinguish among instances at
startup. The Oracle server uses the THREAD number to assign redo log groups to specific
instances. To simplify administration, use the same number for both the THREAD and
INSTANCE_NUMBER parameters.
If you specify UNDO_TABLESPACE with Automatic Undo Management enabled, set this
parameter to a unique undo tablespace name for each instance.
Using the ASM_PREFERRED_READ_FAILURE_GROUPS initialization parameter, you can specify
a list of preferred read failure group names. The disks in those failure groups become the
preferred read disks. Thus, every node can read from its local disks. The setting for this
parameter is instance specific, and the values do not need to be the same on all instances.
Quiescing RAC Databases
• Use the ALTER SYSTEM QUIESCE RESTRICTED

statement from a single instance:
SQL> ALTER SYSTEM QUIESCE RESTRICTED;
• The database cannot be opened by other instances after the

ALTER SYSTEM QUIESCE… statement starts. a
• The ALTER SYSTEM QUIESCE RESTRICTED and ALTER h as
ฺ b r)
SYSTEM UNQUIESCE statements affect all instances m in aฺ
o
iฺc uide
RAC environment. e d
• Cold backups cannot be taken when the s icr ent G
database is in a
d e @ t u d
quiesced state. ver s S i la e thi
v
o_ o us
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Quiescingio RAC Databases
n
b ric a RAC
Toaquiesce - t r adatabase, use the ALTER SYSTEM QUIESCE RESTRICTED statement from
F n o n
one instance. It is not possible to open the database from any instance while the database is in
the process of being quiesced from another instance. After all non-DBA sessions become
inactive, the ALTER SYSTEM QUIESCE RESTRICTED statement executes and the database is
considered to be quiesced. In a RAC environment, this statement affects all instances.
The following conditions apply to RAC:
• If you had issued the ALTER SYSTEM QUIESCE RESTRICTED statement, but the Oracle
server has not finished processing it, then you cannot open the database.
• You cannot open the database if it is already in a quiesced state.
• The ALTER SYSTEM QUIESCE RESTRICTED and ALTER SYSTEM UNQUIESCE
statements affect all instances in a RAC environment, not just the instance that issues the
command.
Cold backups cannot be taken when the database is in a quiesced state because the Oracle
background processes may still perform updates for internal purposes even when the database is
in a quiesced state. Also, the file headers of online data files continue to appear as if they are
being accessed. They do not look the same as if a clean shutdown were done.
Terminating Sessions on a Specific Instance
SQL> SELECT SID, SERIAL#, INST_ID

2 FROM GV$SESSION WHERE USERNAME='JFV';
SID SERIAL# INST_ID
---------- ---------- ----------
140 3340 2
SQL> ALTER SYSTEM KILL SESSION '140,3340,@2';
System altered. s a
)h a
SQL>
m ฺbr
i ฺ c o deฺ
ALTER SYSTEM KILL SESSION '140,3340,@2' r e d Gui
* @ sic dent
ERROR at line 1: e r de Stu
v i l av this
ORA-00031: session marked
i o _ for u s ekill
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfeonr a Specific Instance
Terminating i o Sessions
ic -tran
b r
a with
Starting
F n o nOracle RAC 11gR1, you can use the ALTER SYSTEM KILL SESSION statement
to terminate a session on a specific instance.
The slide illustrates this by terminating a session started on a different instance than the one used
to terminate the problematic session.
If the session is performing some activity that must be completed (such as waiting for a reply
from a remote database or rolling back a transaction), Oracle Database waits for this activity to
complete, marks the session as terminated, and then returns control to you. If the waiting lasts a
minute, then Oracle Database marks the session to be terminated and returns control to you with
a message that the session is marked to be terminated. The PMON background process then
marks the session as terminated when the activity is complete.
Note: You can also use the IMMEDIATE clause at the end of the ALTER SYSTEM command to
immediately terminate the session without waiting for outstanding activity to complete.
How SQL*Plus Commands Affect Instances
SQL*Plus Command Associated Instance
ARCHIVE LOG Generally affects the current instance
CONNECT Affects the default instance if no instance is

specified in the CONNECT command
HOST Affects the node running the SQL*Plus session
RECOVER Does not affect any particular instance, but rather

the database s a
)h a
SHOW PARAMETER and Show the current instance parameter and SGA
m ฺbr
SHOW SGA information
i ฺ c o deฺ
STARTUP and Affect the current instance cre
d Gui
SHUTDOWN
@ si dent
r e Stu
dabout
SHOW INSTANCE Displays information
a v e i s the current instance
l
vi se t h
o _
b r ici to u
e (fa ense
rd lCopyright
l a V ab
i
V Commands
f er Affect Instances
i
How SQL*Pluso
ic -tran s
b r
Fa nostatements
Most SQL n affect the current instance. You can use SQL*Plus to start and stop
instances in the RAC database. You do not need to run SQL*Plus commands as root on
UNIX-based systems or as Administrator on Windows-based systems. You need only the proper
database account with the privileges that you normally use for single-instance Oracle database
administration. The following are some examples of how SQL*Plus commands affect instances:
• The ALTER SYSTEM SET CHECKPOINT LOCAL statement affects only the instance to
which you are currently connected, rather than the default instance or all instances.
• ALTER SYSTEM CHECKPOINT LOCAL affects the current instance.
• ALTER SYSTEM CHECKPOINT or ALTER SYSTEM CHECKPOINT GLOBAL affects
all instances in the cluster database.
• ALTER SYSTEM SWITCH LOGFILE affects only the current instance.
• To force a global log switch, use the ALTER SYSTEM ARCHIVE LOG CURRENT
statement.
• The INSTANCE option of ALTER SYSTEM ARCHIVE LOG enables you to archive each
online redo log file for a specific instance.
Transparent Data Encryption and Wallets in RAC
• One wallet shared by all instances on shared storage:

– No additional administration required
• One copy of the wallet on each local storage:
– Local copies need to be synchronized each time master key is
changed
ALTER SYSTEM SET ENCRYPTION KEY 1 a

as
)h ฺbr
Wallet Wallet Walleto
c m e ฺ
i ฺ i d
Manual
i c red t Gu
Master keys
copy
Master key
@ s Masterd e nkey
2 e r de Stu
Node1
v i lav e this
Node2 Noden
i _ ucopy
oManual s
r ic t o
( f ab nse
e rde lCopyright
l a V ab e
i
V sfer
Transparent
i c i o Data Encryption
a n and Wallets in RAC
r r
-t RAC instances for Transparent Database Encryption may be a local copy of a
Fab used
Wallets
n o nby
common wallet shared by multiple nodes, or a shared copy residing on shared storage that all of
the nodes can access.
A deployment with a single wallet on a shared disk requires no additional configuration to use
Transparent Data Encryption.
If you want to use local copies, you must copy the wallet and make it available to all of the other
nodes after initial configuration. For systems using Transparent Data Encryption with encrypted
wallets, you can use any standard file transport protocol. For systems using Transparent Data
Encryption with obfuscated wallets, file transport through a secured channel is recommended.
The wallet must reside in the directory specified by the setting for the WALLET_LOCATION or
ENCRYPTION_WALLET_LOCATION parameter in sqlnet.ora. The local copies of the wallet
need not be synchronized for the duration of Transparent Data Encryption usage until the server
key is rekeyed through the ALTER SYSTEM SET KEY SQL statement. Each time you run the
ALTER SYSTEM SET KEY statement at a database instance, you must again copy the wallet
residing on that node and make it available to all of the other nodes. To avoid unnecessary
administrative overhead, reserve rekeying for exceptional cases where you are certain that the
server master key is compromised and that not rekeying it would cause a serious security
problem. Note that RMAN does not implicitly backup wallets.
ASM: General Architecture
Node1 Group Services Group Services Node2
DB tom=+ASM1 tom=+ASM2 DB
instance bob=+ASM1 bob=+ASM2 instance
SID=sales1 harry=+ASM1 harry=+ASM2 SID=sales2
DBW0 ASMB ASMB DBW0
FG
ASM ASM FG RBAL
RBAL
instance instance
FG SID=+ASM1 SID=+ASM2 FG
ASMB ASMB
GMON RBAL GMON
RBAL
s a
DB DBW0
ARB0 ARB0
DBW0
DB
)h a
ฺbr
instance instance
… …
SID=test1 RBAL
ARB ARB
RBAL SID=test2
c m
o deฺ
A A
i ฺ
d Gui
r e
@ sic dent
ASM disks ASM disks ASM disks
r
ASM disks
e de StASM u disks ASM disks
v i lav e this
ASM disk group Tom
ic i o_diskogroup
ASM usBob ASM disk group Harry
b r t
e (fa ense
rd lCopyright
l a V ab
i
VArchitecture
f er
i
ASM: General o
ic Storage s
anManagement (ASM) is part of the database kernel. One portion of the ASM
b r - t r
Fa non
Automatic
code allows for the startup of a special instance called an ASM instance. ASM instances do not
mount databases but instead manage the metadata needed to make ASM files available to
ordinary database instances. Both ASM instances and database instances have access to a
common set of disks called disk groups. Database instances access the contents of ASM files
directly, communicating with an ASM instance only to obtain information about the layout of
these files.
An ASM instance contains three new types of background processes. The first type is
responsible for coordinating rebalance activity for disk groups, and is called RBAL. The second
type actually performs the data extent movements. There can be many of these at a time, and
they are called ARB0, ARB1, and so on. The third type is responsible for certain disk group–
monitoring operations that maintain ASM metadata inside disk groups. The disk group monitor
process is called GMON.
Each database instance that uses ASM has two new background processes called ASMB and
RBAL. In a database instance, RBAL performs global opens of the disks in the disk groups.
ASMB runs in database instances and connects to foreground processes in ASM instances. Over
those connections, periodic messages are exchanged to update statistics and to verify that both
instances are healthy. During operations that require ASM intervention, such as a file creation by
a database foreground, the database foreground connects directly to the ASM instance to
perform the
Unauthorized operation. or distribution prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
reproduction
ASM: General Architecture (continued)
An ASMB process is started dynamically when an ASM file is first accessed.
When started, the ASM background connects to the desired ASM instance and maintains that
connection until the database instance no longer has any files open in the disk groups served by
that ASM instance. Database instances are allowed to connect to only one ASM instance at a
time, so they have at most one ASMB background process.
Like RAC, the ASM instances themselves may be clustered, using the existing Global Cache
Services (GCS) infrastructure. There is usually one ASM instance per node on a cluster. As with
existing RAC configurations, ASM requires that the operating system make the disks globally
visible to all of the ASM instances, irrespective of node.
Database instances communicate only with ASM instances on the same node. If there are several
database instances for different databases on the same node, they must share the same single
ASM instance on that node. a
A disk group can contain files for many different Oracle databases. Thus, multiple database h a s
instances serving different databases can access the same disk group even on a single ฺ b r ) system
without RAC. Alternatively, one Oracle database may also store its filesฺin c o m
multiple e ฺ groups
disk
i i d
managed by the same ASM instance.
i c red t Gu
Group Services is used to register the connection information @ s dbyenthe database instances
needed
to find ASM instances. When an ASM instance mounts e r daedisk group,
S tu it registers the disk group
and connect string with Group Services. The vdatabase i t his knows the name of the disk
lav einstance
group, and can therefore use it to look c _ usinformation for the correct ASM instance.
upioconnection
b i
r by to Clusterware, which is automatically
Group Services is a functionalityfaprovided
( s e Oracle
nDatabase
installed on every node that d e
r le li c e
runs Oracle 11g.
Note: If an ASM a
e
V fails,
instance b all Oracle database instances dependent on that ASM instance
also fail. Note
i l
Vthat a file r a
fesystem failure usually crashes a node. In a single–ASM instance
i c i o n s
a ASM instance fails while disk groups are open for update, then after the
a b r
configuration, iftrthe
-
F instance
ASM n n
o reinitializes, it reads the disk group’s log and recovers all transient changes. With
multiple ASM instances sharing disk groups, if one ASM instance fails, then another ASM
instance automatically recovers transient ASM metadata changes caused by the failed instance.
The failure of a database instance does not affect ASM instances.
ASM Instance and Crash Recovery in RAC
ASM instance recovery

Both instances mount Disk group repaired
disk group ASM instance failure by surviving instance
Node1 Node2 Node1 Node2 Node1 Node2
+ASM1 +ASM2 +ASM1 +ASM2 +ASM2
Disk group A Disk group A Disk group A
s a
)h a
ASM crash recovery
m ฺbr
Only one instance Disk group
i ฺ c orepaired
d e ฺ
mounts disk group ASM instance failure
rednextt G
when u i
mounted
Node1 Node2 Node1 Node2 i
s denc
Node1 Node2
+ASM1 +ASM2 +ASM1 +ASM2
d @
e Stu +ASM2
r
ve this
i l a
v GroupsAe
Disk Group A
i o _Disk u Disk Group A
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
Vand Crashf er Recovery in RAC
i
ASM Instance o
ic -tran s
b r
Fa nogroup
Each disk n is self-describing, containing its own file directory, disk directory, and other
data such as metadata logging information. ASM automatically protects its metadata by using
mirroring techniques even with external redundancy disk groups.
With multiple ASM instances mounting the same disk groups, if one ASM instance fails,
another ASM instance automatically recovers transient ASM metadata changes caused by the
failed instance. This situation is called ASM instance recovery, and is automatically and
immediately detected by the global cache services.
With multiple ASM instances mounting different disk groups, or in the case of a single ASM
instance configuration, if an ASM instance fails when ASM metadata is open for update, then
the disk groups that are not currently mounted by any other ASM instance are not recovered
until they are mounted again. When an ASM instance mounts a failed disk group, it reads the
disk group log and recovers all transient changes. This situation is called ASM crash recovery.
Therefore, when using ASM clustered instances, it is recommended to have all ASM instances
always mounting the same set of disk groups. However, it is possible to have a disk group on
locally attached disks that are visible to only one node in a cluster, and have that disk group
mounted on only that node where the disks are attached.
Note: The failure of an Oracle database instance is not significant here because only ASM
instances update ASM metadata.
ASM Instance Initialization Parameters and RAC
• CLUSTER_DATABASE: This parameter must be set to TRUE.

• ASM_DISKGROUP:
– Multiple instances can have different values.
– Shared disk groups must be mounted by each ASM instance.
• ASM_DISKSTRING:
– Multiple instances can have different values. a
h as
– With shared disk groups, every instance should be able to) see
r
the common pool of physical disks. mฺb c o deฺ
• ASM_POWER_LIMIT: Multiple instances can ehave i ฺ
d different
u i
r G
values. @ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VInitialization
f er Parameters and RAC
ASM Instance i o n s
In a b ricto enable
- t r aASM
F order
n o n instances to be clustered together in a RAC environment, each ASM
instance initialization parameter file must set its CLUSTER_DATABASE parameter to TRUE.
This enables the global cache services to be started on each ASM instance.
Although it is possible for multiple ASM instances to have different values for their
ASM_DISKGROUPS parameter, it is recommended for each ASM instance to mount the same
set of disk groups. This enables disk groups to be shared among ASM instances for recovery
purposes. In addition, all disk groups used to store one RAC database must be shared by all
ASM instances in the cluster.
Consequently, if you are sharing disk groups among ASM instances, their ASM_DISKSTRING
initialization parameter must point to the same set of physical media. However, this parameter
does not need to have the same setting on each node. For example, assume that the physical
disks of a disk group are mapped by the OS on node A as /dev/rdsk/c1t1d0s2, and on
node B as /dev/rdsk/c2t1d0s2. Although both nodes have different disk string settings,
they locate the same devices via the OS mappings. This situation can occur when the hardware
configurations of node A and node B are different—for example, when nodes are using different
controllers as in the example above. ASM handles this situation because it inspects the contents
of the disk header block to determine the disk group to which it belongs, rather than attempting
to maintain a fixed list of path names.
ASM and SRVCTL with RAC
• SRVCTL enables you to manage ASM from an Oracle

Clusterware (OC) perspective:
– Add an ASM instance to OC.
– Enable an ASM instance for OC automatic restart.
– Start up an ASM instance.
– Shut down an ASM instance.
s a
– Disable an ASM instance from OC automatic restart. h a
– Remove an ASM instance configuration from the OCR. ฺ b r)
c o m eฺ
– Get some status information. ฺ
di Guid
r e
– sic instance.
Set ASM instance dependency to database ent@ d
• The DBCA allows you to create ASM e r de instances
S tu as well as
helps you to add and enable
v i av with
lthem e t hisOC.
_ io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
V with f r
eRAC
ASM and SRVCTL
i o
icuse SRVCTL s
an to perform the following ASM administration tasks:
b r - t r
Fa non
You can
• The ADD option adds Oracle Cluster Registry (OCR) information about an ASM instance to
run under Oracle Clusterware (OC). This option also enables the recourse.
• The ENABLE option enables an ASM instance to run under OC for automatic startup, or
restart.
• The DISABLE option disables an ASM instance to prevent OC inappropriate automatic
restarts. DISABLE also prevents any startup of that ASM instance using SRVCTL.
• The START option starts an OC-enabled ASM instance. SRVCTL uses the SYSDBA
connection to perform the operation.
• The STOP option stops an ASM instance by using the shutdown normal, transactional,
immediate, or abort option.
• The CONFIG option displays the configuration information stored in the OCR for a
particular ASM instance.
• The STATUS option obtains the current status of an ASM instance.
• The REMOVE option removes the configuration of an ASM instance.
• The MODIFY INSTANCE command can be used to establishes a dependency between an
ASM instance and a database instance.
Note: Adding and enabling an ASM instance is automatically performed by the DBCA when
creating the ASM instance.
ASM and SRVCTL with RAC: Examples
• Start an ASM instance on the specified node:

$ srvctl start asm –n clusnode1
• Stop an ASM instance on the specified node:

$ srvctl stop asm –n clusnode1 –o immediate
s a
• Add OCR data about an existing ASM instance:
)h a
mฺ
$ srvctl add asm -n clusnode1 -i +ASM1 -o /ora/ora10
br
i ฺ c o deฺ
$ srvctl modify instance -d crm -i crm1 -s r e d Gui
+asm1
• Disable OC management of an ASMeinstance: @ sic dent
e r d S tu
$ srvctl disable asm –n clusnode1
v i lav e –i t his+ASM1
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V with f r
eRAC:
ASM and SRVCTL
i c i o a n s Examples
r t r
Fabarensome
Here
on-examples:
• The first example starts up the only existing ASM instance on the CLUSNODE1 node. The
–o option allows you to specify in which mode you want to open the instance: open is the
default, but you can also specify mount, or nomount.
• The second example is of an immediate shutdown of the only existing ASM instance on
CLUSNODE1.
• The third example adds to the OCR the OC information for +ASM1 on CLUSNODE1. You
need to specify the ORACLE_HOME of the instance. Although the following should not be
needed in case you use the DBCA, if you manually create ASM instances, you should also
create OC dependency between database instances and ASM instances to ensure that the
ASM instance starts up before starting database instance, and to allow database instances to
be cleanly shut down before ASM instances. To establish the dependency, you have to use a
command similar to the following: srvctl modify instance -d crm -i crm1 -s
+asm1 for each corresponding instance.
• The fourth example prevents OC to automatically restart +ASM1.
Note: For more information, refer to the Oracle Clusterware and Oracle Real Application
Clusters Administration and Deployment Guide.
ASM Disk Groups with EM in RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V with f r
eEM
i c o
ASM Disk iGroups
a n s in RAC
r r
-at new disk group from an ASM instance, this disk group is not automatically
Fab you
When
n n
add
o
mounted by other ASM instances. If you want to mount the newly added disk group on all ASM
instances, for example, by using SQL*Plus, then you need to manually mount the disk group on
each ASM instance.
However, if you are using Enterprise Manager (EM) to add a disk group, then the disk group
definition includes a check box to indicate whether the disk group is automatically mounted to
all the ASM clustered database instances. This is also true when you mount and dismount ASM
disk groups by using Database Control where you can use a check box to indicate which
instances mount or dismount the ASM disk group.
Disk Group Performance Page and RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer Page and RAC
Disk Groupi oPerformance
ic -tran
b
Onathe
rAutomatic
non
Storage Management Performance page, click the Disk Group I/O Cumulative
F
Statistics link in the Additional Monitoring Links section. On the Disk Group I/O Cumulative
Statistics page, click the corresponding disk group name.
A performance page is displayed showing clusterwide performance information for the
corresponding disk group.
By clicking one of the proposed links—for example, I/O Throughput on the slide—you can see
an instance-level performance details graph as shown at the bottom of the slide.
Summary

• Use Enterprise Manager cluster database pages
• Define redo log files in a RAC environment
• Define undo tablespaces in a RAC environment
• Start and stop RAC databases and instances
• Modify initialization parameters in a RAC environment s a
h a
• Manage ASM instances in a RAC environment ฺbr) m eฺ
ฺ c o
r e di Guid
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
This practice covers manipulating redo threads.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Managing Backup and Recovery in RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Configure the RAC database to use ARCHIVELOG mode
and the flash recovery area
• Configure RMAN for the RAC environment
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i r
Objectives io V nsfe
b r ic and-trrecovery
a
RAC
F a backup n is almost identical to other Oracle database backup and recovery
no
operations. This is because you are backing up and recovering a single database. The main difference
is that with RAC you are dealing with multiple threads of redo log files.
Protecting Against Media Failure
Archived Archived
log files log files
Database s a
Mirrored backups )h a
disks
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Media f erFailure
o
Protecting iAgainst s
an you with methods to avoid or to reduce down time due to a failure of one or
icRAC -provides
b r t r
Fa non
Although
more (but not all) of your instances, you must still protect the database itself, which is shared by all
the instances. This means that you need to consider disk backup and recovery strategies for your
cluster database just as you would for a non-clustered database.
To minimize the potential loss of data due to disk failures, you may want to use disk mirroring
technology (available from your server or disk vendor). As in non-clustered databases, you can have
more than one mirror if your vendor allows it, to help reduce the potential for data loss and to
provide you with alternative backup strategies. For example, with your database in ARCHIVELOG
mode and with three copies of your disks, you can remove one mirror copy and perform your backup
from it while the two remaining mirror copies continue to protect ongoing disk activity. To do this
correctly, you must first put the tablespaces into backup mode and then, if required by your cluster or
disk vendor, temporarily halt disk operations by issuing the ALTER SYSTEM SUSPEND command.
After the statement completes, you can break the mirror and then resume normal operations by
executing the ALTER SYSTEM RESUME command and taking the tablespaces out of backup mode.
Archived Log File Configurations
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
Cluster file system scheme: Local archive
@ sicwith d eNFS
n t
Archive logs from each scheme:
r d e Each S t uinstance can
instance are written to the
i l a ve mounted
read
t h is archive
same file location.
i o _v destinations
u s e of all instances.
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V Configurations
f
Archived Logi o File
ic -and n s
arecovery
b r t r
Fa non
During backup operations involving archived log files, the Oracle server determines the
file destinations and names from the control file. If you use RMAN, the archived log file path names
can also be stored in the optional recovery catalog. However, the archived log file path names do not
include the node name, so RMAN expects to find the files it needs on the nodes where the channels
are allocated.
If you use a cluster file system, your instances can all write to the same archive log destination. This
is known as the cluster file system scheme. Backup and recovery of the archive logs are easy because
all logs are located in the same directory.
If a cluster file system is not available, then Oracle recommends that local archive log destinations be
created for each instance with NFS-read mount points to all other instances. This is known as the
local archive with network file system (NFS) scheme. During backup, you can either back up the
archive logs from each host or select one host to perform the backup for all archive logs. During
recovery, one instance may access the logs from any host without having to first copy them to the
local destination.
Using either scheme, you may want to provide a second archive destination to avoid single points of
failure.
RAC and the Flash Recovery Area
Flash
recovery
area s a
)h a
m ฺbr
i ฺ c o deฺ
e d NFSGui
Certified
r
Cluster file system
@ sicdirectory
e n t
r d e Stud
i
ASMl a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Recovery f er Area
RAC and the i oFlash
c tran s
To a b
use
raiflash -
recovery area in RAC, you must place it on an ASM disk group, a cluster file system,
F non
or on a shared directory that is configured through certified NFS for each RAC instance. That is, the
flash recovery area must be shared among all the instances of a RAC database.
RAC Backup and Recovery Using EM
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Recoveryf er Using EM
RAC Backup i o and s
an Database backup and recovery related tasks by clicking the Availability tab
b ric the -Cluster
Youaaccess t r
F non
on the Cluster Database home page. On the Availability tabbed page, you can use RMAN to perform
a range of backup and recovery operations, such as scheduling backups, performing recovery when
necessary, and configuring backup and recovery settings. There are also links related to Oracle
Secure Backup and Service management.
Configure RAC Recovery Settings with EM
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VRecovery f erSettings with EM
i
Configure RACo
c tran s
b riuse
Youacan -
Enterprise Manager to configure important recovery settings for your cluster database.
F non
On the Database home page, click the Availability tab, and then click the Recovery Settings link.
From here, you can ensure that your database is in archivelog mode and configure flash recovery
settings.
With a RAC database, if the Archive Log Destination setting is not the same for all instances, the
field appears blank, with a message indicating that instances have different settings for this field. In
this case, entering a location in this field sets the archive log location for all instances of database.
You can assign instance specific values for an archive log destination by using the Initialization
Parameters page.
Note: You can run the ALTER DATABASE SQL statement to change the archiving mode in RAC as
long as the database is mounted by the local instance but not open in any instances. You do not need
to modify parameter settings to run this statement. Set the initialization parameters
DB_RECOVERY_FILE_DEST and DB_RECOVERY_FILE_DEST_SIZE to the same values on all
instances to configure a flash recovery area in a RAC environment.
Archived Redo File Conventions in RAC
Parameter Description Example
%r Resetlogs identifier log_1_62_23452345
%R Padded resetlogs identifier log_1_62_0023452345
%s Log sequence number, not padded log_251
s a
%S Log sequence number, left-zero-padded log_0000000251
)h a
m ฺbr
%t Thread number, not padded log_1 i ฺ c o deฺ
r e d Gui
%T Thread number, left-zero-padded @ sic dent
log_0001
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
VFile Conventions
f
Archived Redo
i c i o a n s in RAC
r r
-tredo log configuration, uniquely identify the archived redo logs with the
F b archived
Foraany
n o n
LOG_ARCHIVE_FORMAT parameter. The format of this parameter is operating system specific and it
can include text strings, one or more variables, and a file name extension.
All of the thread parameters, in either upper or lower case, are mandatory for RAC. This enables the
Oracle database to create unique names for archive logs across the incarnation. This requirement is in
effect when the COMPATIBLE parameter is set to 10.0 or greater. Use the %R or %r parameter to
include the resetlogs identifier to avoid overwriting the logs from a previous incarnation. If you do
not specify a log format, then the default is operating system specific and includes %t, %s, and %r.
As an example, if the instance associated with redo thread number 1 sets LOG_ARCHIVE_FORMAT to
log_%t_%s_%r.arc, then its archived redo log files are named as:
log_1_1000_23435343.arc
log_1_1001_23452345.arc
log_1_1002_23452345.arc
...
Configure RAC Backup Settings with EM
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VBackup f r
eSettings
Configure RAC
i c i o a n s with EM
r t r
Fab nbackup
Persistent
on- settings can be configured using Enterprise Manager. On the Database Control
home page, click the Availability tab, and then click the Backup Settings link. You can configure
disk settings such as the directory location of your disk backups, and level of parallelism. You can
also choose the default backup type:
• Backup set
• Compressed backup set
• Image copy
You can also specify important tape-related settings such as the number of available tape drives and
vendor-specific media management parameters.
Oracle Recovery Manager
Recovery
RMAN provides the
Manager following benefits for
Recovery Real Application
catalog Clusters:
Archived
log files Oracle • Can read cluster files
Server or raw partitions with
process
Stored no configuration s a
h a
scripts changes ฺ b r)
• Can access c o m eฺ
multiple
ฺ
di Guid
Oracle r
archive elog
database @ sic dent
Backup
rdedestinations
tu S
Snapshot lave
this
storage
control o _vi se
file
b r ici to u
e (fa ense
rd lCopyright
l a V ab
i
V Managerf er
i o
Oracle Recovery s
an (RMAN) can use stored scripts, interactive scripts, or an interactive GUI
ic -tManager
b r r
Fa non
Oracle Recovery
front end. When using RMAN with your RAC database, use stored scripts to initiate the backup and
recovery processes from the most appropriate node.
If you use different Oracle Home locations for your RAC instances on each of your nodes, create a
snapshot control file in a location that exist on all your nodes. The snapshot control file is only
needed on the nodes on which RMAN performs backups. The snapshot control file does not need to
be globally available to all instances in a RAC environment though.
You can use either a cluster file or a shared raw device as well as a local directory that exists on each
node in your cluster. Here is an example:
RMAN> CONFIGURE SNAPSHOT CONTROLFILE TO
'/oracle/db_files/snaps/snap_prod1.cf';
For recovery, you must ensure that each recovery node can access the archive log files from all
instances by using one of the archive schemes discussed earlier, or make the archived logs available
to the recovering instance by copying them from another location.
Configure RMAN Snapshot Control File Location
• The snapshot control file path must be valid on every

node from which you might initiate an RMAN backup.
• Configure the snapshot control file location in RMAN.
– Determine the current location:
RMAN> SHOW SNAPSHOT CONTROLFILE NAME;
/u01/app/oracle/product/11.1.0/dbs/scf/snap_prod.cf
– You can use ASM, a shared file system location or a s a

shared block device if you prefer: h a
r) ฺb
RMAN> CONFIGURE SNAPSHOT CONTROLFILE NAME TO com e ฺ
i ฺ i d
'+FRA/SNAP/snap_prod.cf';
i c red t Gu
RMAN> CONFIGURE SNAPSHOT CONTROLFILE NAME
@ s TO d e n
r
'/ocfs/oradata/dbs/scf/snap_prod.cf';
e de Stu
i lav e thisNAME TO
RMAN> CONFIGURE SNAPSHOT CONTROLFILE
v
'/dev/sdj2'; io_ us ric e to
a b
(f ens
d e
r lCopyright
l ic © 2008, Oracle. All rights reserved.
e
V ab e
i l a
V Snapshot er Control File Location
i
Configure RMANo s f
b ric control
Theasnapshot - t r anfile is a temporary file that RMAN creates to resynchronize from a read-
F on
consistentnversion of the control file. RMAN needs a snapshot control file only when resynchronizing
with the recovery catalog or when making a backup of the current control file.
In a RAC database, the snapshot control file is created on the node that is making the backup. You
need to configure a default path and file name for these snapshot control files that are valid on every
node from which you might initiate an RMAN backup.
Run the following RMAN command to determine the configured location of the snapshot control
file:
SHOW SNAPSHOT CONTROLFILE NAME
You can change the configured location of the snapshot control file. For example, on UNIX-based
systems you can specify the snapshot control file location as snap_prod.cf located in the ASM
disk group +FRA by entering the following at the RMAN prompt:
CONFIGURE SNAPSHOT CONTROLFILE NAME TO '+FRA/SNAP/snap_prod.cf'
This command globally sets the configuration for the location of the snapshot control file throughout
your cluster database.
Note: The CONFIGURE command creates persistent settings across RMAN sessions.
Configure Control File and SPFILE Autobackup
• RMAN automatically creates a control file and SPFILE

backup after BACKUP or COPY:
RMAN> CONFIGURE CONTROLFILE AUTOBACKUP ON;
• Change default location:

RMAN> CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR
DEVICE TYPE DISK TO '+FRA'; s a
)h a
• Location must be available to all nodes in yourm ฺb
RAC
r
database. i ฺ c o deฺ
d ui
s icre ent G
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Filesfander SPFILE Autobackup
i o
Configure Control
icCONFIGURE an CONTROLFILE AUTOBACKUP to ON, RMAN automatically creates a control
b r - t r
Fa
If you set
on backup after you run the BACKUP or COPY command. RMAN can also
file and annSPFILE
automatically restore an SPFILE if required to start an instance to perform recovery. This means that
the default location for the SPFILE must be available to all nodes in your RAC database.
These features are important in disaster recovery because RMAN can restore the control file even
without a recovery catalog. RMAN can restore an autobackup of the control file even after the loss of
both the recovery catalog and the current control file.
You can change the default name that RMAN gives to this file with the CONFIGURE CONTROLFILE
AUTOBACKUP FORMAT command. If you specify an absolute path name in this command, this path
must exist identically on all nodes that participate in backups.
Note: RMAN performs the control file autobackup on the first allocated channel. When you allocate
multiple channels with different parameters (especially if you allocate a channel with the CONNECT
command), you must determine which channel will perform the automatic backup. Always allocate
the channel for the connected node first.
Crosschecking on Multiple RAC Clusters Nodes
When cross-checking on multiple nodes make sure all

backups can be accessed by every node in the cluster.
• This allows you to allocate channels at any node in the
cluster during restore or cross-check operations
• Otherwise you must allocate channels on multiple
nodes by providing the CONNECT option to the
a
CONFIGURE CHANNEL command has r)
• If backups are not accessible because no channel m ฺbwas
i
configured on the node that can access thoseฺ c obackups,
d e ฺ
ed t Gu i
the backups are marked EXPIRED. sicr n
r d e@ Stude
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
Von Multiple
f er RAC Clusters Nodes
i
Crosschecking o s
ic -tranon multiple RAC nodes, configure the cluster so that all backups can be
b r
Fa byoevery
When cross-checking
accessed n
n
node, regardless of which node created the backup. When the cluster is configured
this way, you can allocate channels at any node in the cluster during restore or cross-check
operations.
If you cannot configure the cluster so that each node can access all backups, then, during restore and
cross-check operations, you must allocate channels on multiple nodes by providing the CONNECT
option to the CONFIGURE CHANNEL command so that every backup can be accessed by at least
one node. If some backups are not accessible during cross-check because no channel was configured
on the node that can access those backups, the backups are marked EXPIRED in the RMAN
repository after the cross-check.
For example, you can use CONFIGURE CHANNEL ... CONNECT in an Oracle RAC
configuration in which tape backups are created on various nodes in the cluster and each backup is
accessible only on the node on which it is created.
Channel Connections to Cluster Instances
• During backup, each allocated channel can connect to

a different instance in the cluster.
• Instances to which the channels connect must be
either all mounted or all open.
CONFIGURE DEFAULT DEVICE TYPE TO sbt;
CONFIGURE DEVICE TYPE sbt PARALLELISM 3;
CONFIGURE CHANNEL 1 DEVICE TYPE sbt CONNECT='sys/rac@RACDB1'; s a
)h a
ฺbr
CONFIGURE CHANNEL 2 DEVICE TYPE sbt CONNECT='sys/rac@RACDB2';
CONFIGURE CHANNEL 3 DEVICE TYPE sbt CONNECT='sys/rac@RACDB3';
c m
o deฺ
i ฺ
d Gui
or r e
@ sic dent
CONFIGURE DEFAULT DEVICE TYPE TO sbt; e r de Stu
CONFIGURE DEVICE TYPE sbt PARALLELISM
v i lav e th3;is
CONFIGURE CHANNEL DEVICE TYPE
ic i o_ sbt o u sCONNECT='sys/rac@BR';
( f abr nse t
e rde lCopyright
l a V ab e
i
V sto f erCluster Instances
i o
Channel Connections
ic backups an in parallel, RMAN channels can connect to a different instance in the cluster.
b r - t r
Fa non
When making
The examples in the slide illustrate two possible configurations:
• If you want to dedicate channels to specific instances, you can control at which instance the
channels are allocated by using separate connect strings for each channel configuration as shown
by the first example.
• If you define a special service for your backup and recovery jobs, you can use the second
example shown in the slide. If you configure this service with load balancing turned on, the
channels are allocated at a node as decided by the load-balancing algorithm.
During backup, the instances to which the channels connect must be either all mounted or all open.
For example, if the RACDB1 instance has the database mounted but the RACDB2 and RACDB3
instances have the database open, the backup fails.
Note: In some cluster database configurations, some nodes of the cluster have faster access to certain
data files than to other data files. RMAN automatically detects this, which is known as node affinity
awareness. When deciding which channel to use to back up a particular data file, RMAN gives
preference to the nodes with faster access to the data files that you want to back up. For example, if
you have a three-node cluster, and if node 1 has faster read/write access to data files 7, 8, and 9 than
do the other nodes, then node 1 has greater node affinity to those files than nodes 2 and 3 and RMAN
will take advantage of this automatically.
RMAN Channel Support for the Grid
• RAC allows the use of nondeterministic connect

strings.
• It simplifies the use of parallelism with RMAN in a RAC
environment.
• It uses the load-balancing characteristics of the grid
environment.
s a
– Channels connect to RAC instances that are the least h a
loaded. ฺbr) m eฺ
ฺ c o
r e di Guid
CONFIGURE DEFAULT DEVICE TYPE TO sbt; sic n t
CONFIGURE DEVICE TYPE sbt PARALLELISM @
e 3;Stud e
r d
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
VSupport f r
efor
RMAN Channel
i c i o a n s the Grid
r t r
Fab nDatabase
In Oracle
on- 11g, RAC allows the use of nondeterministic connect strings that can connect to
different instances based on RAC features such as load balancing. Therefore, to support RAC, the
RMAN polling mechanism no longer depends on deterministic connect strings, and makes it possible
to use RMAN with connect strings that are not bound to a specific instance in the grid environment.
Previously, if you wanted to use RMAN parallelism and spread a job between many instances, you
had to manually allocate an RMAN channel for each instance. In Oracle Database 11g, to use
dynamic channel allocation, you do not need separate CONFIGURE CHANNEL CONNECT
statements anymore. You only need to define your degree of parallelism by using a command such as
CONFIGURE DEVICE TYPE disk PARALLELISM, and then run backup or restore commands.
RMAN then automatically connects to different instances and does the job in parallel. The grid
environment selects the instances that RMAN connects to, based on load balancing. As a result of
this, configuring RMAN parallelism in a RAC environment becomes as simple as setting it up in a
non-RAC environment. By configuring parallelism when backing up or recovering a RAC database,
RMAN channels are dynamically allocated across all RAC instances.
Note: RMAN has no control over the selection of the instances. If you require a guaranteed
connection to an instance, you should provide a connect string that can connect only to the required
instance.
RMAN Default Autolocation
• Recovery Manager autolocates the following files:

– Backup pieces
– Archived redo logs during backup
– Data file or control file copies
• If local archiving is used, a node can read only those
archived logs that were generated on that node.
s a
• When restoring, a channel connected to a specific node h a
restores only those files that were backed up to ฺ b
the
r)
c o m eฺ
node. di ฺ uid
s icre ent G
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
VAutolocationf er
RMAN Default i o
icManager n s
aautomatically
a b
Recovery
r - t r discovers which nodes of a RAC configuration can access the files
F n o n
that you want to back up or restore. Recovery Manager autolocates the following files:
• Backup pieces during backup or restore
• Archived redo logs during backup
• Data file or control file copies during backup or restore
If you use a noncluster file system local archiving scheme, then a node can read only those archived
redo logs that were generated by an instance on that node. RMAN never attempts to back up archived
redo logs on a channel that it cannot read.
During a restore operation, RMAN automatically performs the autolocation of backups. A channel
connected to a specific node attempts to restore only those files that were backed up to the node. For
example, assume that log sequence 1001 is backed up to the drive attached to node 1, whereas log
1002 is backed up to the drive attached to node 2. If you then allocate channels that connect to each
node, then the channel connected to node 1 can restore log 1001 (but not 1002), and the channel
connected to node 2 can restore log 1002 (but not 1001).
Distribution of Backups
Three possible backup configurations for RAC:

• A dedicated backup server performs and manages
backups for the cluster and the cluster database.
• One node has access to a local backup appliance and
performs and manages backups for the cluster
database.
s a
• Each node has access to a local backup appliance and h a
can write to its own local backup media. ฺ b r)
m c o deฺ
i ฺ
d Gui
r e
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
DistributionioofV
Backupss f
b r ic -trathenbackup options for RAC, you have three possible configurations:
F• aNetwork
When configuring
nonbackup server: A dedicated backup server performs and manages backups for the
cluster and the cluster database. None of the nodes have local backup appliances.
• One local drive: One node has access to a local backup appliance and performs and manages
backups for the cluster database. All nodes of the cluster should be on a cluster file system to be
able to read all data files, archived redo logs, and SPFILEs. It is recommended that you do not
use the noncluster file system archiving scheme if you have backup media on only one local
drive.
• Multiple drives: Each node has access to a local backup appliance and can write to its own
local backup media.
In the cluster file system scheme, any node can access all the data files, archived redo logs, and
SPFILEs. In the noncluster file system scheme, you must write the backup script so that the backup is
distributed to the correct drive and path for each node. For example, node 1 can back up the archived
redo logs whose path names begin with /arc_dest_1, node 2 can back up the archived redo logs
whose path names begin with /arc_dest_2, and node 3 can back up the archived redo logs whose
path names begin with /arc_dest_3.
One Local Drive CFS Backup Scheme
RMAN> CONFIGURE DEVICE TYPE sbt PARALLELISM 1;

RMAN> CONFIGURE DEFAULT DEVICE TYPE TO sbt;
RMAN> BACKUP DATABASE PLUS ARCHIVELOG DELETE INPUT;
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V CFSsBackup
f er Scheme
i
One Local Driveo an backup scheme, each node in the cluster has read access to all the data files,
b ric file -system
In aacluster t r
F non
archived redo logs, and SPFILEs. This includes Automated Storage Management (ASM), cluster file
systems, and Network Attached Storage (NAS).
When backing up to only one local drive in the cluster file system backup scheme, it is assumed that
only one node in the cluster has a local backup appliance such as a tape drive. In this case, run the
following one-time configuration commands:
Because any node performing the backup has read/write access to the archived redo logs written by
the other nodes, the backup script for any node is simple:
BACKUP DATABASE PLUS ARCHIVELOG DELETE INPUT;
In this case, the tape drive receives all data files, archived redo logs, and SPFILEs.
Multiple Drives CFS Backup Scheme

CONFIGURE CHANNEL 1 DEVICE TYPE sbt CONNECT 'usr1/pwd1@n1';
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VCFS Backupf er Scheme
i o
Multiple Drives
ic up s
anmultiple drives in the cluster file system backup scheme, it is assumed that each
b r - t r
Fa non
When backing to
node in the cluster has its own local tape drive. Perform the following one-time configuration so that
one channel is configured for each node in the cluster. This is a one-time configuration step. For
example, enter the following at the RMAN prompt:
CONFIGURE CHANNEL 1 DEVICE TYPE sbt CONNECT 'user1/passwd1@node1';
CONFIGURE CHANNEL 2 DEVICE TYPE sbt CONNECT 'user2/passwd2@node2';
Similarly, you can perform this configuration for a device type of DISK. The following backup
script, which you can run from any node in the cluster, distributes the data files, archived redo logs,
and SPFILE backups among the backup drives:
Non-CFS Backup Scheme

s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Scheme f er
i
Non-CFS Backupo n s
b ric -file
In aanoncluster t r asystem environment, each node can back up only its own local archived redo logs.
F non
For example, node 1 cannot access the archived redo logs on node 2 or node 3 unless you configure
the network file system for remote access. To configure NFS, distribute the backup to multiple
drives. However, if you configure NFS for backups, then you can back up to only one drive.
When backing up to multiple drives in a noncluster file system backup scheme, it is assumed that
each node in the cluster has its own local tape drive. You can perform a similar one-time
configuration as the one shown in the slide to configure one channel for each node in the cluster.
Similarly, you can perform this configuration for a device type of DISK. Develop a production
backup script for whole database backups that you can run from any node. With the BACKUP
example, the data file backups, archived redo logs, and SPFILE backups are distributed among the
different tape drives. However, channel 1 can read only the logs archived locally on /arc_dest_1.
This is because the autolocation feature restricts channel 1 to back up only the archived redo logs in
the /arc_dest_1 directory. Because node 2 can read files only in the /arc_dest_2 directory,
channel 2 can back up only the archived redo logs in the /arc_dest_2 directory, and so on. The
important point is that all logs are backed up, but they are distributed among the different drives.
Restoring and Recovering
• Media recovery may require one or more archived log

files from each thread.
• The RMAN RECOVER command automatically restores
and applies the required archived logs.
• Archive logs may be restored to any node performing
the restore and recover operation.
a
• Logs must be readable from the node performing thehas
restore and recovery activity. ฺ b r)
c o m eฺ
• Recovery processes request additional threads ฺ
di Guid
r e
enabled during the recovery period. sic ent
d e @ tud
• Recovery processes notify you v r
e of threads
s S no longer
l a h i
needed because they were et
_vi disabled.
io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
VRecovering
f er
i
Restoring ando s
ic -oftraa ndatabase that is accessed by RAC may require at least one archived log file for
b r
Fa non
Media recovery
each thread. However, if a thread’s online redo log contains enough recovery information, restoring
archived log files for any thread is unnecessary.
If you use RMAN for media recovery and you share archive log directories, you can change the
destination of the automatic restoration of archive logs with the SET clause to restore the files to a
local directory of the node where you begin recovery. If you backed up the archive logs from each
node without using a central media management system, you must first restore all the log files from
the remote nodes and move them to the host from which you will start recovery with RMAN.
However, if you backed up each node’s log files using a central media management system, you can
use RMAN’s AUTOLOCATE feature. This enables you to recover a database using the local tape
drive on the remote node.
If recovery reaches a time when an additional thread was enabled, the recovery process requests the
archived log file for that thread. If you are using a backup control file, when all archive log files are
exhausted, you may need to redirect the recovery process to the online redo log files to complete
recovery. If recovery reaches a time when a thread was disabled, the process informs you that the log
file for that thread is no longer needed.
Summary

• Configure RAC recovery settings with EM
• Configure RAC backup settings with EM
• Initiate archiving
• Configure RMAN
• Perform RAC backup and recovery using EM s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Note i o
ic recovery anprocedures for OCR and voting disk are described in the Oracle Clusterware
b r - t r
Fa non
Backup and
Administration lesson covered later in this course.

• Configuring the RAC database to use ARCHIVELOG
mode and the flash recovery area
• Configuring RMAN for the RAC environment
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
RAC Performance Tuning
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Determine RAC-specific tuning components
• Tune instance recovery in RAC
• Determine RAC-specific wait events, global enqueues,
and system statistics
• Implement the most common RAC tuning tips a
h a s
• Use the Cluster Database Performance pages r)
ฺ b
min RAC
• Use the Automatic Workload Repository (AWR)
i ฺ c o d e ฺ
d Gu i
icre ent(ADDM)
• Use Automatic Database Diagnostic sMonitor in
RAC de@ tud e r sS
i v
la e thi
v
o_ o us
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
CPU and Wait Time Tuning Dimensions
CPU
time
Possibly Scalable
needs SQL application
tuning
s a
)h a
Scalable Needs
instance/RAC
No gain achieved
by adding m ฺbr
application
CPUs/nodes i ฺ c o deฺ
tuning
r e d Gui
@ sic dent
e r de S tu
Wait
v i lav e this time
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VTime Tuning f er Dimensions
i
CPU and Wait o
ic your n s
asystem,
b r - t r
Fa non
When tuning it is important that you compare the CPU time with the wait time of your
system. Comparing CPU time with wait time helps to determine how much of the response time is
spent on useful work and how much on waiting for resources potentially held by other processes.
As a general rule, the systems where CPU time is dominant usually need less tuning than the ones
where wait time is dominant. On the other hand, heavy CPU usage can be caused by badly written
SQL statements.
Although the proportion of CPU time to wait time always tends to decrease as load on the system
increases, steep increases in wait time are a sign of contention and must be addressed for good
scalability.
Adding more CPUs to a node, or nodes to a cluster, would provide very limited benefit under
contention. Conversely, a system where the proportion of CPU time does not decrease significantly
as load increases can scale better, and would most likely benefit from adding CPUs or Real
Application Clusters (RAC) instances if needed.
Note: Automatic Workload Repository (AWR) reports display CPU time together with wait time in
the Top 5 Timed Events section, if the CPU time portion is among the top five events.
RAC-Specific Tuning
• Tune for a single instance first.

• Tune for RAC:
– Instance recovery
– Interconnect traffic
– Point of serialization can be exacerbated
• RAC-reactive tuning tools: a
Certain combinations
as
– Specific wait events are characteristic of
ฺbr )h
– System and enqueue statistics well-known tuning cases.
m
i ฺ c o deฺ
– Enterprise Manager performance pages ed
r u i
sic ent G
– Statspack and AWR reports
• RAC-proactive tuning tools: r d e@ Stud
i l a ve this
– AWR snapshots _v use
– ADDM reports ric i o
a b e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
RAC-Specific i o Tuning
icthere -are n
aspecific
a b r
Although t r tuning areas for RAC, such as instance recovery and interconnect traffic,
F n o n
you get most benefits by tuning your system like a single-instance system. At least, this must be your
starting point.
Obviously, if you have serialization issues in a single-instance environment, these may be
exacerbated with RAC.
As shown in the slide, you have basically the same tuning tools with RAC as with a single-instance
system. However, certain combinations of specific wait events and statistics are well-known RAC
tuning cases.
In this lesson, you see some of those specific combinations, as well as the RAC-specific information
that you can get from the Enterprise Manager performance pages, and Statspack and AWR reports.
Finally, you see the RAC-specific information that you can get from the Automatic Database
Diagnostic Monitor (ADDM).
RAC and Instance or Crash Recovery
Use information
Remaster for other caches
enqueue LMS
resources Remaster recovers
1 cache GRD
resources
2
s a
Build )h a
SMON recovery set
m ฺ br
recovers Resource
i ฺ c o deฺ
the 3 claim
r e d Gui
database Merge failed 4
@ sicRolldforward
e nt
redo threads
r d e Srecoveryt u set
i l a ve this 5
v
_ us time
oRecovery e
ic i
r e to
a b
(f ens
r lCopyright
l
e
V ab e
i l a
V orsCrash er Recovery
i o
RAC and Instance f
b r icinstance
- t r an and the failure is detected by another instance, the second instance performs
Fa non
When an fails
the following recovery steps:
1. During the first phase of recovery, Global Enqueue Services (GES) remasters the enqueues.
2. The Global Cache Services (GCS) remasters its resources. The GCS processes remaster only
those resources that lose their masters. During this time, all GCS resource requests and write
requests are temporarily suspended. However, transactions can continue to modify data blocks
as long as these transactions have already acquired the necessary resources.
3. After enqueues are reconfigured, one of the surviving instances can grab the Instance Recovery
enqueue. Therefore, at the same time as GCS resources are remastered, SMON determines the
set of blocks that need recovery. This set is called the recovery set. Because, with Cache Fusion,
an instance ships the contents of its blocks to the requesting instance without writing the blocks
to the disk, the on-disk version of the blocks may not contain the changes that are made by either
instance. This implies that SMON needs to merge the content of all the online redo logs of each
failed instance to determine the recovery set. This is because one failed thread might contain a
hole in the redo that needs to be applied to a particular block. So, redo threads of failed instances
cannot be applied serially. Also, redo threads of surviving instances are not needed for recovery
because SMON could use past or current images of their corresponding buffer caches.
RAC and Instance or Crash Recovery (continued)
4. Buffer space for recovery is allocated and the resources that were identified in the previous
reading of the redo logs are claimed as recovery resources. This is done to avoid other instances
to access those resources.
5. All resources required for subsequent processing have been acquired and the Global Resource
Directory (GRD) is now unfrozen. Any data blocks that are not in recovery can now be
accessed. Note that the system is already partially available.
Then, assuming that there are past images or current images of blocks to be recovered in other
caches in the cluster database, the most recent image is the starting point of recovery for these
particular blocks. If neither the past image buffers nor the current buffer for a data block is in
any of the surviving instances’ caches, then SMON performs a log merge of the failed instances.
SMON recovers and writes each block identified in step 3, releasing the recovery resources
immediately after block recovery so that more blocks become available as recovery proceeds.
Refer to the section “Write to Disk Coordination: Example” in this lesson for more information
about past images. s a
) h a
After all blocks have been recovered and the recovery resources have been released,bthe
ฺ r system is
again fully available.
c o m e ฺ
i ฺ i d
i c red tbecomes
In summary, the recovered database or the recovered portions of the database
G u available
earlier, and before the completion of the entire recovery sequence.
@ s Thisdmakes
e n the system available
sooner and it makes recovery more scalable.
e r de Stu
Note: The performance overhead of a log merge iis v thisto the number of failed instances and
laproportional
v
to the size of the amount of redo written inio
ric e to
the usefor each instance.
_redo logs
a b
(f ens
e
r le lic
d
e
V ab
i l a
V sfer
i o
ic -tran
b r
Fa non
Instance Recovery and Database Availability
Full A G H
Database availability
Partial B F
2
s a
4 )h a
m ฺbr
1 2 3
i ฺ c o deฺ
None C D E
r e d Gui
Elapsed time @s
ic ent
r d e Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V and f r
eDatabase
i c i
Instance Recoveryo a n s Availability
r r
n-t the degree of database availability during each step of Oracle instance
F b noillustrates
Theagraphic
recovery:
A. Real Application Clusters is running on multiple nodes.
B. Node failure is detected.
C. The enqueue part of the GRD is reconfigured; resource management is redistributed to the
surviving nodes. This operation occurs relatively quickly.
D. The cache part of the GRD is reconfigured and SMON reads the redo log of the failed instance
to identify the database blocks that it needs to recover.
E. SMON issues the GRD requests to obtain all the database blocks it needs for recovery. After the
requests are complete, all other blocks are accessible.
F. The Oracle server performs roll forward recovery. Redo logs of the failed threads are applied to
the database, and blocks are available right after their recovery is completed.
G. The Oracle server performs rollback recovery. Undo blocks are applied to the database for all
uncommitted transactions.
H. Instance recovery is complete and all data is accessible.
Note: The dashed line represents the blocks identified in step 2 in the previous slide. Also, the dotted
steps represent the ones identified in the previous slide.
Instance Recovery and RAC
Instance startup
Instance + Instance
crashes crash recovery opens
FAST_START_MTTR_TARGET
Instance
starts
Rolling
Instance forward a
ends as
Instance recovery )h ฺbr
…
crashes first pass + lock claim
c m
o deฺ
FAST_START_MTTR_TARGET ed
i ฺ u i
r G
Instance @ sic dent
recovery r de Stu
V$INSTANCE_RECOVERY.ESTD_CLUSTER_AVAILABLE_TIME
e
starts
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V and f r
eRAC
i o
Instance Recovery n s
b ric -traenvironment,
In aasingle-instance the instance startup combined with the crash recovery time is
F n o n
controlled by the setting of the FAST_START_MTTR_TARGET initialization parameter. You can set its
value if you want incremental checkpointing to be more aggressive than autotune checkpointing.
However, this is at the expense of a much higher I/O overhead.
In a RAC environment, including the startup time of the instance in this calculation is useless
because one of the surviving instances is doing the recovery.
In a RAC environment, it is possible to monitor the estimated target (in seconds) for the duration
from the start of instance recovery to the time when GCD is open for lock requests for blocks that are
not needed for recovery. This estimation is published in the V$INSTANCE_RECOVERY view through the
ESTD_CLUSTER_AVAILABLE_TIME column. Basically, you can monitor the time your cluster is frozen
during instance recovery situations.
In a RAC environment, the FAST_START_MTTR_TARGET initialization parameter is used to bound the
entire instance recovery time, assuming that it is instance recovery for single-instance death.
Note: If you really want to have small instance recovery time by setting FAST_START_MTTR_TARGET,
you can safely ignore the alert log messages about raising its value.
Instance Recovery and RAC
• Use parallel instance recovery.

• Increase PARALLEL_EXECUTION_MESSAGE_SIZE.
• Set PARALLEL_MIN_SERVERS.
• Use Async I/O.
• Increase the size of the default buffer cache.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V and f r
eRAC
i c i o
Instance Recovery
a n s (continued)
r -t r
Fabare some
Here
is faster:n o nguidelines you can use to make sure that instance recovery in your RAC environment
• Use parallel instance recovery by setting RECOVERY_PARALLISM.

• Increase PARALLEL_EXECUTION_MESSAGE_SIZE from its default of 2,148 bytes to
4 KB or 8 KB. This should provide better recovery slave performance.
• Set PARALLEL_MIN_SERVERS to CPU_COUNT-1. This will prespawn recovery slaves at
startup time.
• Using asynchronous I/O is one of the most crucial factors in recovery time. The first-pass log
read uses asynchronous I/O.
• Instance recovery uses 50 percent of the default buffer cache for recovery buffers. If this is not
enough, some of the steps of instance recovery will be done in several passes. You should be
able to identify such situations by looking at your alert.log file. In that case, you should
increase the size of your default buffer cache.
Analyzing Cache Fusion Impact in RAC
• The cost of block access and cache coherency is

represented by:
– Global Cache Services statistics
– Global Cache Services wait events
• The response time for Cache Fusion transfers is
determined by:
s a
– Overhead of the physical interconnect components h a
– IPC protocol ฺ b r)
c o m eฺ
– GCS protocol di ฺ uid re G
• The response time is not generally@ sic denby
affected t disk I/O
e Stu
factors. erd
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Fusion f erImpact in RAC
i o
Analyzing Cache s
an blocks in the global cache and maintaining cache coherency is represented
b ricof accessing
Theaeffect - t r
F
by: non
• The Global Cache Services statistics for current and cr blocks; for example, gc current blocks
received, gc cr blocks received, and so on.
• The Global Cache Services wait events for gc current block 3-way, gc cr grant 2-way, and so on.
The response time for Cache Fusion transfers is determined by the messaging time and processing
time imposed by the physical interconnect components, the IPC protocol, and the GCS protocol. It is
not affected by disk input/output (I/O) factors other than occasional log writes. The Cache Fusion
protocol does not require I/O to data files in order to guarantee cache coherency, and RAC inherently
does not cause any more I/O to disk than a nonclustered instance.
Typical Latencies for RAC Operations
AWR Report Latency Name Lower Typical Upper

Bound Bound
Average time to process cr block request 0.1 1 10
Avg global cache cr block receive time (ms) 0.3 4 12
Average time to process current block request 0.1 3 23
Avg global cache current block receive time(ms) 0.3 8 30
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V forsRACf er Operations
i o
Typical Latencies
an there is a table in the RAC Statistics section containing average times
b ricAWR-treport,
In aaRAC r
F non
(latencies) for some Global Cache Services and Global Enqueue Services operations. This table is
shown in the slide and is called “Global Cache and Enqueue Services: Workload Characteristics.”
Those latencies should be monitored over time, and significant increases in their values should be
investigated. The table presents some typical values, based on empirical observations. Factors that
may cause variations to those latencies include:
• Utilization of the IPC protocol. User-mode IPC protocols are faster, but only Tru64’s RDG is
recommended for use.
• Scheduling delays, when the system is under high CPU utilization
• Log flushes for current blocks served
Other RAC latencies in AWR reports are mostly derived from V$GES_STATISTICS and may be
useful for debugging purposes, but do not require frequent monitoring.
Note: The time to process consistent read (CR) block request in the cache corresponds to (build
time + flush time + send time), and the time to process current block request in the
cache corresponds to (pin time + flush time + send time).
Wait Events for RAC
• Wait events help to analyze what sessions are waiting

for.
• Wait times are attributed to events that reflect the
outcome of a request:
– Placeholders while waiting
– Precise events after waiting
s a
• Global cache waits are summarized in a broader h a
category called Cluster Wait Class. ฺ b r)
c o m eฺ
• These wait events are used in ADDM to enable ฺ
di GCacheu id
r e
Fusion diagnostics. sic ent
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
VRAC sfer
Wait Events i ofor
an are waiting for is an important method to determine where time is spent. In
icwhat-tsessions
b r r
Fa
Analyzing
on
RAC, thenwait time is attributed to an event that reflects the exact outcome of a request. For example,
when a session on an instance is looking for a block in the global cache, it does not know whether it
will receive the data cached by another instance or whether it will receive a message to read from
disk. The wait events for the global cache convey precise information and wait for global cache
blocks or messages. They are mainly categorized by the following:
• Summarized in a broader category called Cluster Wait Class
• Temporarily represented by a placeholder event that is active while waiting for a block
• Attributed to precise events when the outcome of the request is known
The wait events for RAC convey information valuable for performance analysis. They are used in
ADDM to enable precise diagnostics of the impact of Cache Fusion.
Wait Event Views
Total waits for an event V$SYSTEM_EVENT
Waits for a wait event class V$SESSION_WAIT_CLASS

by a session
Waits for an event by a session V$SESSION_EVENT
Activity of recent active sessions V$ACTIVE_SESSION_HISTORY a

as
Last 10 wait events ฺ )h
br
for each active session om ฺ
V$SESSION_WAIT_HISTORY
iฺc uide
e d
Events for which
s icr ent G
e@ Stud
active sessions are waiting V$SESSION_WAIT
r d
Identify SQL statements impacted i l a ve this
by interconnect latencies i o _v use V$SQLSTATS
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
o
Wait Event iViews
c tran
itakes
a
When b r
it - time to acquire resources because of the total path length and latency for
F n o nsome
requests, processes sleep to avoid spinning for indeterminate periods of time. When the process
decides to wait, it wakes up either after a specified timer value expires (timeout) or when the event it
is waiting for occurs and the process is posted. The wait events are recorded and aggregated in the
views shown in the slide. The first three are aggregations of wait times, timeouts, and the number of
times waited for a particular event, whereas the rest enable the monitoring of waiting sessions in real
time, including a history of recent events waited for.
The individual events distinguish themselves by their names and the parameters that they assume.
For most of the global cache wait events, the parameters include file number, block number, the
block class, and access mode dispositions, such as mode held and requested. The wait times for
events presented and aggregated in these views are very useful when debugging response time
performance issues. Note that the time waited is cumulative, and that the event with the highest score
is not necessarily a problem. However, if the available CPU power cannot be maximized, or response
times for an application are too high, the top wait events provide valuable performance diagnostics.
Note: Use the CLUSTER_WAIT_TIME column in V$SQLSTATS to identify SQL statements impacted
by interconnect latencies, or run an ADDM report on the corresponding AWR snapshot.
Global Cache Wait Events: Overview
Just requested gc [current/cr] [multiblock] request

(placeholder)
gc [current/cr] block [2/3]-way gc [current/cr] block busy

Received after two or three network Received but not sent immediately
hops, immediately after request
gc [current/cr] grant 2-way gc current grant busy

Not received and not mastered Not received and not mastered
s a
locally. Grant received immediately.
)h
locally. Grant received with delay.a
m ฺbr
gc [current/cr] [block/grant] congested
i ฺ c o deฺ
gc [current/cr] [failure/retry]
r e d Gui
Block or grant received with delay
because of CPU or memory lack sic dbecause
Not received
@ e nt of failure
e r de Stu
gc buffer
i v this
busy
laarrival
o v
_than buffer
Block
s etime
r ic i
less
o u pin time
b t
e (fa ense
e rd lCopyright
l a V ab
V i fer Overview
i
Global Cache o Wait n s
Events:
b ricglobal-tcache
Theamain ra wait events for Oracle Database 11g are described briefly in the slide:
F no n
• gc current/cr request: These wait events are relevant only while a gc request for a cr
or current buffer is in progress. They act as placeholders until the request completes.
• gc [current/cr] [2/3]-way: A current or cr block is requested and received after two
or three network hops. The request is processed immediately: the block is not busy or congested.
• gc [current/cr] block busy: A current or cr block is requested and received, but is
not sent immediately by LMS because some special condition that delayed the sending was
found.
• gc [current/cr] grant 2-way: A current or cr block is requested and a grant message
received. The grant is given without any significant delays. If the block is not in its local cache,
a current or cr grant is followed by a disk read on the requesting instance.
• gc current grant busy: A current block is requested and a grant message received. The
busy hint implies that the request is blocked because others are ahead of it or it cannot be
handled immediately.
Note: For dynamic remastering, two events are of the most practical importance: gc remaster
and gc quiesce. They can be symptoms of the effect of remastering on the running processes.
Global Cache Wait Events: Overview (continued)
• gc [current/cr] [block/grant] congested: A current or cr block is requested
and a block or grant message received. The congested hint implies that the request spent more
than 1 ms in internal queues.
• gc [current/cr] [failure/retry]: A block is requested and a failure status
received or some other exceptional event has occurred.
• gc buffer busy: If the time between buffer accesses becomes less than the time the buffer
is pinned in memory, the buffer containing a block is said to become busy and as a result
interested users may have to wait for it to be unpinned.
Note: For more information, refer to the Oracle Database Reference guide.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
2-way Block Request: Example
Wait:
gc current block request
1
FGP
Direct send
SGA2 2
LGWR:
LGWR
Log sync
s a
)h a
ฺbr
SGA1
c m
o deฺ
i ฺ
d Gui
r e
@ sic dent
Block transfer
e r de Stu
Wait complete: 3 vil
av this LMS
gc current block 2-way io_ u s e

a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V sfExample
2-way Block i o Request:
anwhat typically happens when the master instance requests a block that is not
b ricshows-tyou
Thisaslide r
F non
cached locally. Here it is supposed that the master instance is called SGA1, and SGA2 contains the
requested block. The scenario is as follows:
1. SGA1 sends a direct request to SGA2. So SGA1 waits on the gc current block
request event.
2. When SGA2 receives the request, its local LGWR process may need to flush some recovery
information to its local redo log files. For example, if the cached block is frequently changed,
and the changes have not been logged yet, LMS would have to ask LGWR to flush the log
before it can ship the block. This may add a delay to the serving of the block and may show up
in the requesting node as a busy wait.
3. Then, SGA2 sends the requested block to SGA1. When the block arrives in SGA1, the wait event
is complete, and is reflected as gc current block 2-way.
Note: Using the notation R = time at requestor, W = wire time and transfer delay, and S = time at
server, the total time for a round-trip would be:
R(send) + W(small msg) + S(process msg,process block,send) + W(block) + R(receive block)
3-way Block Request: Example
Wait:
1 LMS 2
FGP Resource
Direct
Master
message
SGA2 3
LGWR
s a
)h a
ฺbr
SGA1
c m
o deฺ
i ฺ
d Gui
r e
@ sic dent
Block transfer rde t uLMS
ve this S
Wait complete: i
v se
4 l a
gc current block 3-way io_ u
b r ic t o
e (fa ense
rd lCopyright
l a V ab
i er
V sfExample
3-way Block i o Request:
c tran
imodified
b r
Fa non- scenario for a cluster with more than two nodes. It is very similar to the previous
This is a
one. However, the master for this block is on a node that is different from that of the requestor, and
where the block is cached. Thus, the request must be forwarded. There is an additional delay for one
message and the processing at the master node:
R(send) + W(small msg) + S(process msg,send) + W(small msg) + S(process msg,process block,send)
+ W(block) + R(receive block)
While a remote read is pending, any process on the requesting instance that is trying to write or read
the data cached in the buffer has to wait for a gc buffer busy. The buffer remains globally busy
until the block arrives.
2-way Grant: Example
Wait:
1 LMS
FGP Resource
Direct
Master
message
SGA2
s a
2
)h a
ฺbr
SGA1
3
c m
o deฺ
i ฺ
d Gui
r e
@ sic dent
Grant message
e r de Stu
Wait complete:
v i lav e this
gc current grant 2-way io_ us
r ic t o
( f ab nse
e rde lCopyright
l a V ab e
i
V sfer
2-way Grant: i o Example
an message is sent by the master because the requested block is not cached in
ic -targrant
b r
Fa non
In this scenario,
any instance.
If the local instance is the resource master, the grant happens immediately. If not, the grant is always
2-way, regardless of the number of instances in the cluster.
The grant messages are small. For every block read from the disk, a grant has to be received before
the I/O is initiated, which adds the latency of the grant round-trip to the disk latency:
R(send) + W(small msg) + S(process msg,send) + W(small msg) + R(receive block)
The round-trip looks similar to a 2-way block round-trip, with the difference that the wire time is
determined by a small message, and the processing does not involve the buffer cache.
Global Enqueue Waits: Overview
• Enqueues are synchronous.

• Enqueues are global resources in RAC.
• The most frequent waits are for:
TX TM
s a
)h a
US HW
m ฺbr
i ฺ c o deฺ
r e d Gui
TA SQ @s
ic ent
r d e Stud
• The waits may constitute serious
i l a ve thserialization
is points.
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Waits: f r
eOverview
i o
Global Enqueue n s
An a b ric wait-trisanot
enqueue RAC specific, but involves a global lock operation when RAC is enabled.
F of the
Most
n requests
noglobal for enqueues are synchronous, and foreground processes wait for them.
Therefore, contention on enqueues in RAC is more visible than in single-instance environments.
Most waits for enqueues occur for enqueues of the following types:
• TX: Transaction enqueue; used for transaction demarcation and tracking
• TM: Table or partition enqueue; used to protect table definitions during DML operations
• HW: High-water mark enqueue; acquired to synchronize a new block operation
• SQ: Sequence enqueue; used to serialize incrementing of an Oracle sequence number
• US: Undo segment enqueue; mainly used by the Automatic Undo Management (AUM) feature
• TA: Enqueue used mainly for transaction recovery as part of instance recovery
In all of the cases above, the waits are synchronous and may constitute serious serialization points
that can be exacerbated in a RAC environment.
Note: In Oracle Database 11g, the enqueue wait events specify the resource name and a reason for
the wait—for example, TX Enqueue index block split. This makes diagnostics of enqueue waits
easier.
Session and System Statistics
• Use V$SYSSTAT to characterize the workload.

• Use V$SESSTAT to monitor important sessions.
• V$SEGMENT_STATISTICS includes RAC statistics.
• RAC-relevant statistic groups are:
– Global Cache Service statistics
– Global Enqueue Service statistics a
as
– Statistics for messages sent )h ฺ br
• V$ENQUEUE_STATISTICS determines the enqueue c m
o dwithe ฺ
i ฺ i
the highest impact.
i c red t Gu
• V$INSTANCE_CACHE_TRANSFER breaks @ s down d e n GCS
statistics into block classes. e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V sStatistics
f
Session and i oSystem
an based on V$SYSSTAT enables characterization of the database activity based
ic statistics
b r - t r
Fa noItnis the basis for many metrics and ratios used in various tools and methods, such as
Using system
on averages.
AWR, Statspack, and Database Control.
In order to drill down to individual sessions or groups of sessions, V$SESSTAT is useful when the
important session identifiers to monitor are known. Its usefulness is enhanced if an application fills in
the MODULE and ACTION columns in V$SESSION.
V$SEGMENT_STATISTICS is useful for RAC because it also tracks the number of CR and current
blocks received by the object.
The RAC-relevant statistics can be grouped into:
• Global Cache Service statistics: gc cr blocks received, gc cr block receive time, and so on
• Global Enqueue Service statistics: global enqueue gets and so on
• Statistics for messages sent: gcs messages sent and ges messages sent
V$ENQUEUE_STATISTICS can be queried to determine which enqueue has the highest impact on
database service times and eventually response times.
V$INSTANCE_CACHE_TRANSFER indicates how many current and CR blocks per block class are
received from each instance, including how many transfers incurred a delay.
Note: For more information about statistics, refer to the Oracle Database Reference guide.
Most Common RAC Tuning Tips
• Application tuning is often the most beneficial.

• Reducing long full-table scans in OLTP systems
• Using Automatic Segment Space Management
• Increasing sequence caches
• Using partitioning to reduce interinstance traffic
• Avoiding unnecessary parsing s a
h a
• Minimizing locking usage ฺ b r)
• Removing unselective indexes c o m eฺ
ฺ
di Guid
Configuring interconnect properly sic r e
• nt
r d e@ Stude
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
VRAC Tuning
f er Tips
Most Common i o
ic -system, s
an RAC or single instance, the most significant performance gains are usually
b r t r
Fa nontraditional application-tuning techniques. The benefits of those techniques are even
In any database
obtained from
more remarkable in a RAC database. In addition to traditional application tuning, some of the
techniques that are particularly important for RAC include the following:
• Try to avoid long full-table scans to minimize GCS requests. The overhead caused by the global
CR requests in this scenario is because of the fact that when queries result in local cache misses,
an attempt is first made to find the data in another cache, based on the assumption that the
chance that another instance has cached the block is high.
• Automatic Segment Space Management can provide instance affinity to table blocks.
• Increasing sequence caches improves instance affinity to index keys deriving their values from
sequences. That technique may result in significant performance gains for multiinstance insert-
intensive applications.
• Range or list partitioning may be very effective in conjunction with data-dependent routing, if
the workload can be directed to modify a particular range of values from a particular instance.
• Hash partitioning may help to reduce buffer busy contention by making buffer access
distribution patterns sparser, enabling more buffers to be available for concurrent access.
Most Common RAC Tuning Tips (continued)
• In RAC, library cache and row cache operations are globally coordinated. So, excessive parsing
means additional interconnect traffic. Library cache locks are heavily used, in particular by
applications using PL/SQL or Advanced Queuing. Library cache locks are acquired in exclusive
mode whenever a package or procedure has to be recompiled.
• Because transaction locks are globally coordinated, they also deserve special attention in
RAC. For example, using tables instead of Oracle sequences to generate unique numbers is not
recommended because it may cause severe contention even for a single instance system.
• Indexes that are not selective do not improve query performance, but can degrade DML
performance. In RAC, unselective index blocks may be subject to interinstance contention,
increasing the frequency of cache transfers for indexes belonging to INSERT-intensive tables.
• Always verify that you use a private network for your interconnect, and that your private
network is configured properly. Ensure that a network link is operating in full duplex mode.
Ensure that your network interface and Ethernet switches support MTU size of 9 KB. Note that a
single GBE can scale up to ten thousand 8-KB blocks per second before saturation. as a
r ) h
m ฺ b
i ฺ c o d e ฺ
red t Gu i
i c
s den
@
de Stu
e r
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Index Block Contention: Considerations
Wait events
enq: TX - index
Index
contention block
gc buffer busy
Split in
progress
gc current block
busy
gc current split
s a
System statistics
)h a
Leaf node splits m ฺbr
i ฺ c o deฺ
Branch node splits
r e d Gui
Exchange deadlocks
@ sic dent
gcs refuse xid
e r de Stu
gcs ast xid
v i lav e this
Service ITL waits
ic i o_ o us RAC01 RAC02
( f abr nse t
e rde lCopyright
l a V ab e
i
V sferConsiderations
Index BlockioContention:
b r ic systems
- t r an where the loading or batch processing of data is a dominant business function,
Fa non
In application
there may be performance issues affecting response times because of the high volume of data
inserted into indexes. Depending on the access frequency and the number of processes concurrently
inserting data, indexes can become hot spots and contention can be exacerbated by:
• Ordered, monotonically increasing key values in the index (right-growing trees)
• Frequent leaf block splits
• Low tree depth: All leaf block access go through the root block.
A leaf or branch block split can become an important serialization point if the particular leaf block or
branch of the tree is concurrently accessed.
The tables in the slide sum up the most common symptoms associated with the splitting of index
blocks, listing wait events and statistics that are commonly elevated when index block splits are
prevalent. As a general recommendation, to alleviate the performance impact of globally hot index
blocks and leaf block splits, a more uniform, less skewed distribution of the concurrency in the index
tree should be the primary objective. This can be achieved by:
• Global index hash partitioning
• Increasing the sequence cache, if the key value is derived from a sequence
• Use natural keys as opposed to surrogate keys
• Use reverse key indexes
Oracle Sequences and Index Contention
Can contain 500 rows
1…50000 50001…100000
s a
)h a
m ฺbr
CACHE 50000 NOORDER i ฺ c o deฺ
r e d Gui
@ sic dent
RAC01
r d e Stu RAC02
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V and f r
eIndex
i c i o
Oracle Sequences
a n s Contention
r r
-t values generated by sequences tend to be subject to leaf block contention when the
Fab with
Indexes
n o key
n
insert rate is high. That is because the index leaf block holding the highest key value is changed for
every row inserted, as the values are monotonically ascending. In RAC, this may lead to a high rate
of current and CR blocks transferred between nodes.
One of the simplest techniques that can be used to limit this overhead is to increase the sequence
cache, if you are using Oracle sequences. As the difference between sequence values generated by
different instances increases, successive index block splits tend to create instance affinity to index
leaf blocks. For example, suppose that an index key value is generated by a CACHE NOORDER
sequence and each index leaf block can hold 500 rows. If the sequence cache is set to 50000, while
instance 1 inserts values 1, 2, 3, and so on, instance 2 concurrently inserts 50001, 50002, and so on.
After some block splits, each instance writes to a different part of the index tree.
So, what is the ideal value for a sequence cache to avoid interinstance leaf index block contention,
yet minimizing possible gaps? One of the main variables to consider is the insert rate: the higher it is,
the higher must be the sequence cache. However, creating a simulation to evaluate the gains for a
specific configuration is recommended.
Note: By default, the cache value is 20. Typically, 20 is too small for the example above.
Undo Block Considerations
Index Changes
Reads
SGA1 SGA2 s a
)h a
Undo Undo m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
Additional
ic i o_ o us traffic
interconnect
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
Undo BlockioConsiderations
b r t r anshipment and contention for undo buffers usually happens when index blocks
icundo-block
Fa non
Excessive
containing active transactions from multiple instances are read frequently.
When a SELECT statement needs to read a block with active transactions, it has to undo the changes
to create a CR version. If the active transactions in the block belong to more than one instance, there
is a need to combine local and remote undo information for the consistent read. Depending on the
amount of index blocks changed by multiple instances and the duration of the transactions, undo
block shipment may become a bottleneck.
Usually this happens in applications that read recently inserted data very frequently, but commit
infrequently. Techniques that can be used to reduce such situations include the following:
• Shorter transactions reduce the likelihood that any given index block in the cache contains
uncommitted data, thereby reducing the need to access undo information for consistent read.
• As explained earlier, increasing sequence cache sizes can reduce interinstance concurrent access
to index leaf blocks. CR versions of index blocks modified by only one instance can be
fabricated without the need of remote undo information.
Note: In RAC, the problem is exacerbated by the fact that a subset of the undo information has to be
obtained from remote instances.
High-Water Mark Considerations
Wait events
enq: HW -
contention Heavy
gc current grant inserts
HWM
Heavy
s a
inserts
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
New extent
e r de Stu
v i lav e this
RAC01
ic i o_ o us RAC02
( f abr nse t
e rde lCopyright
l a V ab e
i er
V Considerations
f
High-Waterio Mark s
an of wait events and statistics presents itself in applications where the insertion
iccombination
b r - t r
Fa non
A certain
of data is a dominant business function and new blocks have to be allocated frequently to a segment.
If data is inserted at a high rate, new blocks may have to be made available after unfruitful searches
for free space. This has to happen while holding the high-water mark (HWM) enqueue.
Therefore, the most common symptoms for this scenario include:
• A high percentage of wait time for enq: HW – contention
• A high percentage of wait time for gc current grant events
The former is a consequence of the serialization on the HWM enqueue, and the latter is because of
the fact that current access to the new data blocks that need formatting is required for the new block
operation. In a RAC environment, the length of this space management operation is proportional to
the time it takes to acquire the HWM enqueue and the time it takes to acquire global locks for all the
new blocks that need formatting. This time is small under normal circumstances because there is
never any access conflict for the new blocks. Therefore, this scenario may be observed in
applications with business functions requiring a lot of data loading, and the main recommendation to
alleviate the symptoms is to define uniform and large extent sizes for the locally managed and
automatic space managed segments that are subject to high-volume inserts.
Concurrent Cross-Instance Calls: Considerations
Dirty
block
SGA1 SGA2
Table1 Table1
CKPT CKPT
Table2 Table2
s a
)h a
1 m ฺbr
i ฺ c o deฺ
r e d G2 ui
3 4
@ sic dent
Truncate Table1
r d e Truncate
S t u Table2
Cross-instance
i l a ve call t his
v
o_ o us e
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer Calls: Considerations
Concurrentio Cross-Instance
b r ic -traandn data mart environments, it is not uncommon to see a lot of TRUNCATE
Fa non
In data warehouse
operations. These essentially happen on tables containing temporary data.
In a RAC environment, truncating tables concurrently from different instances does not scale well,
especially if, in conjunction, you are also using direct read operations such as parallel queries.
As shown in the slide, a truncate operation requires a cross-instance call to flush dirty blocks of the
table that may be spread across instances. This constitutes a point of serialization. So, while the first
TRUNCATE command is processing, the second has to wait until the first one completes.
There are different types of cross-instance calls. However, all use the same serialization mechanism.
For example, the cache flush for a partitioned table with many partitions may add latency to a
corresponding parallel query. This is because each cross-instance call is serialized at the cluster level,
and one cross-instance call is needed for each partition at the start of the parallel query for direct read
purposes.
Monitoring RAC Database
and Cluster Performance Database
Directly from Database Control and Grid Control: Instances

• Status of each node in the cluster
• Aggregated alert messages across all the instances
• Review issues that are affecting the entire cluster or
each instance
• Monitor cluster cache coherency statistics a
h a s
• Determine if any of the services for the cluster r)
ฺ b
database are having availability problems com eฺ
e d iฺ uid
• Review any outstanding Clusterware interconnect
icr ent G
alerts s
e@ ud r d S t
e
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Database f er and Cluster Performance
o
MonitoringiRAC s
an Manager Database Control and Grid Control are cluster-aware and provide a
ic Enterprise
b r - t r
Fa non to manage your cluster database. From the Cluster Database Home page, you can do
Both Oracle
central console
the following:
• View the overall system status, such as the number of nodes in the cluster and their current
status so that you do not have to access each individual database instance for details
• View alert messages aggregated across all the instances with lists for the source of each alert
message
• Review issues that are affecting the entire cluster as well as those that are affecting individual
instances
• Monitor cluster cache coherency statistics to help you identify processing trends and optimize
performance for your Oracle RAC environment. Cache coherency statistics measure how well
the data in caches on multiple instances is synchronized.
• Determine if any of the services for the cluster database are having availability problems. A
service is determined to be a problem service if it is not running on all preferred instances, if its
response time thresholds are not met, and so on.
• Review any outstanding Clusterware interconnect alerts
Cluster Database Performance Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Performance
f er
i c i
Cluster Databaseo a n s Page
r r
F b noDatabase
TheaCluster n-t Performance page provides a quick glimpse of the performance statistics for a
database. Enterprise Manager accumulates data from each instance over specified periods of time
(known as collection-based data). Enterprise Manager also provides current data from each instance
(known as real-time data).
Statistics are rolled up across all the instances in the cluster database. Using the links next to the
charts, you can get more specific information and perform any of the following tasks:
• Identify the causes of performance issues.
• Decide whether resources need to be added or redistributed.
• Tune your SQL plan and schema for better optimization.
• Resolve performance issues.
The screenshot shows a partial view of the Cluster Database Performance page. You access this page
by clicking the Performance tab on the Cluster Database Home page.
Determining Cluster Host Load Average
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Host f er Load Average
Determiningi oCluster s
an Average chart on the Cluster Database Performance page shows potential
b ric Host-trLoad
TheaCluster
F non
problems that are outside the database. The chart shows maximum, average, and minimum load
values for available nodes in the cluster for the previous hour.
If the load average is higher than the average of the total number of CPUs across all the hosts in the
cluster, then too many processes are waiting for CPU resources. SQL statements that are not tuned
often cause high CPU usage. Compare the load average values with the values displayed for CPU
Used in the Average Active Sessions chart. If the sessions value is low and the load average value is
high, something else on the host—other than your database—is consuming the CPU.
You can click any of the load value labels for the Cluster Host Load Average chart to view more
detailed information about that load value. For example, if you click the label Average, the Hosts:
Average Load page appears, displaying charts that depict the average host load for up to four nodes
in the cluster.
You can select whether the data is displayed in a summary chart (combining the data for each node in
one display) or in tile charts (where the data for each node is displayed in its own chart). You can
click Customize to change the number of tile charts displayed in each row or the method of ordering
the tile charts.
Determining Global Cache Block Access Latency
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Cache f er Block Access Latency
Determiningi oGlobal n s
b ric Cache
TheaGlobal - t r aBlock Access Latency chart shows the latency for each different type of data
F non
block requests: current and consistent-read (CR) blocks. That is the elapsed time it takes to locate
and transfer consistent-read and current blocks between the buffer caches.
You can click either metric for the Global Cache Block Access Latency chart to view more detailed
information about that type of cached block.
If the Global Cache Block Access Latency chart shows high latencies (high elapsed times), the cause
can be any of the following:
• A high number of requests caused by SQL statements that are not tuned
• A large number of processes in the queue waiting for the CPU, or scheduling delays
• Slow, busy, or faulty interconnects. In these cases, check your network connection for dropped
packets, retransmittals, or cyclic redundancy check (CRC) errors.
Concurrent read and write activity on shared data in a cluster is a frequently occurring activity.
Depending on the service requirements, this activity does not usually cause performance problems.
However, when global cache requests cause a performance problem, optimizing SQL plans and the
schema to improve the rate at which data blocks are located in the local buffer cache, and minimizing
I/O is a successful strategy for performance tuning. If the latency for consistent-read and current
block requests reaches 10 milliseconds, the first step in resolving the problem should be to go to the
Cluster Cache Coherency page for more detailed information.
Determining Average Active Sessions
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sActive f er Sessions
i
Determining oAverage
anSessions chart in the Cluster Database Performance page shows potential
b ric Active
TheaAverage - t r
F non
problems inside the database. Categories, called wait classes, show how much of the database is
using a resource, such as CPU or disk I/O. Comparing CPU time to wait time helps determine how
much of the response time is consumed with useful work rather than waiting for resources that are
potentially held by other processes.
At the cluster database level, this chart shows the aggregate wait class statistics across all the
instances. For a more detailed analysis, click the clipboard icon at the bottom of the chart to view the
ADDM analysis for the database for that time period.
If you click the wait class legends beside the Average Active Sessions chart, you can view instance-
level information stored in “Active Sessions by Instance” pages. Use the Wait Class action list on the
“Active Sessions by Instance” page to view the different wait classes. The “Active Sessions by
Instance” pages show the service times for up to four instances. Using the Customize button, you can
select the instances that are displayed. You can view the data for the instances separately using tile
charts, or you can combine the data into a single summary chart.
Determining Database Throughput
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V sfThroughput
Determining i oDatabase
c tran
b richart
Thealast - the Performance page monitors the usage of various database resources. By clicking
on
F non
the Throughput tab at the top of this chart, you can view the Database Throughput chart. Compare
the peaks on the Average Active Sessions chart with those on the Database Throughput charts. If
internal contention is high and throughput is low, consider tuning the database.
The Database Throughput charts summarize any resource contention that appears in the Average
Active Sessions chart, and also show how much work the database is performing on behalf of the
users or applications. The Per Second view shows the number of transactions compared to the
number of logons, and (not shown here) the amount of physical reads compared to the redo size per
second. The Per Transaction view shows the amount of physical reads compared to the redo size per
transaction. Logons is the number of users that are logged on to the database.
You can also obtain information at the instance level by clicking one of the legends to the right of the
charts to access the “Database Throughput by Instance” page. This page shows the breakdown of the
aggregated Database Throughput chart for up to four instances. You can select the instances that are
displayed. You can drill down further on the “Database Throughput by Instance” page to see the
sessions of an instance that is consuming the greatest resources. Click an instance name legend just
under the chart to go to the Top Sessions subpage of the Top Consumers page for that instance.
Determining Database Throughput
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V sfThroughput
Determining
i c i oDatabase
a n (continued)
r r
-tthe Performance page monitors the usage of various database resources. By clicking
F b chart
Thealast
n o non
the Instances tab at the top of this chart, you can view the “Active Sessions by Instance” chart, which
summarizes any resource contention that appears in the Average Active Sessions chart. You can thus
quickly determine how much of the database work is being performed on each instance.
You can also obtain information at the instance level by clicking one of the legends to the right of the
chart to access the Top Sessions page, where you can view real-time data showing the sessions that
consume the greatest system resources. In the graph in the slide, the orac2 instance after 8:20 PM
consistently shows more active sessions than the orac1 instance.
Accessing the Cluster Cache Coherency Page
Block
Class
Segment s a
name )h a
m ฺbr
i ฺ c o deฺ
r e d Gui
Segment
@ sic dent
name
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er Coherency Page
VClustersfCache
o
Accessing ithe
anCache Coherency page, click the Performance tab on the Cluster Database
To a b ricthe Cluster
access - t r
F non
Home page, and click Cluster Cache Coherency in the Additional Monitoring Links section at the
bottom of the page. Alternatively, click either of the legends to the right of the Global Cache Block
Access Latency chart.
The Cluster Cache Coherency page contains summary charts for cache coherency metrics for the
cluster:
• Global Cache Block Access Latency: Shows the total elapsed time, or latency, for a block
request. Click one of the legends to the right of the chart to view the average time it takes to
receive data blocks for each block type (current or CR) by instance. On the “Average Block
Receive Time by Instance” page, you can click an instance legend under the chart to go to the
“Block Transfer for Local Instance” page, where you can identify which block classes (such as
undo blocks, data blocks, and so on) are subject to intense global cache activity. This page
displays the block classes that are being transferred and the instances that are transferring most
of the blocks. Cache transfer indicates how many current and CR blocks for each block class
were received from remote instances, including how many transfers incurred a delay (busy) or
an unexpected longer delay (congested).
Accessing the Cluster Cache Coherency Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er Coherency Page (continued)
VClustersfCache
o
Accessing ithe n
b ric Cache
• aGlobal - t r aBlock Transfer Rate: Shows the total aggregated number of blocks received by
F non
all instances in the cluster by way of an interconnect. Click one of the legends to the right of the
chart to go to the “Global Cache Blocks Received by Instance” page for that type of block. From
there, you can click an instance legend under the chart to go to the “Segment Statistics by
Instance” page, where you can see which segments are causing cache contention.
• Global Cache Block Transfers and Physical Reads: Shows the percentage of logical read
operations that retrieved data from the buffer cache of other instances by way of Direct Memory
Access and from disk. It is essentially a profile of how much work is performed in the local
buffer cache rather than the portion of remote references and physical reads (which both have
higher latencies). Click one of the legends to the right of the chart to go to the “Global Cache
Block Transfers vs. Logical Reads by Instance” and “Physical Reads vs. Logical Reads by
Instance” pages. From there, you can click an instance legend under the chart to go to the
“Segment Statistics by Instance” page, where you can see which segments are causing cache
contention.
Viewing Cluster Interconnects Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er Page
V Interconnects
f
i o
Viewing Cluster s
an page is useful for monitoring the interconnect interfaces, determining
b ric Interconnects
TheaCluster - t r
F non
configuration issues, and identifying transfer rate-related issues, including excess traffic. This page
helps determine the load added by individual instances and databases on the interconnect. Sometimes
you can immediately identify interconnect delays that are due to applications outside Oracle.
You can use this page to perform the following tasks:
• Viewing all interfaces that are configured across the cluster
• Viewing statistics for the interfaces, such as absolute transfer rates and errors
• Determining the type of interfaces, such as private or public
• Determining whether the instance is using a public or private network
• Determining which database instance is currently using which interface
• Determining how much the instance is contributing to the transfer rate on the interface
The Private Interconnect Transfer Rate value shows a global view of the private interconnect traffic,
which is the estimated traffic on all the private networks in the cluster. The traffic is calculated as the
summary of the input rate of all private interfaces that are known to the cluster.
Viewing Cluster Interconnects Page (continued)
On the Cluster Interconnects page, you can access the Hardware Details page, where you obtain more
information about all the network interfaces defined on each node of your cluster.
Similarly, you can access the Transfer Rate metric page, which collects the internode communication
traffic of a cluster database instance. The critical and warning thresholds of this metric are not set by
default. You can set them according to the speed of your cluster interconnects.
Note: You can query the V$CLUSTER_INTERCONNECTS view to see information about the private
interconnect.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Viewing the Database Locks Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sLocks f er Page
Viewing theioDatabase
b
Useathe
ic -trLocks
rDatabase an page to determine if multiple instances are holding locks for the same
F non
object. The page shows user locks, all database locks, or locks that are blocking other users or
applications. You can use this information to stop a session that is unnecessarily locking an object.
To access the Database Locks page, select Performance on the Cluster Database Home page, and
click Database Locks in the Additional Monitoring Links section at the bottom of the Performance
subpage.
AWR Snapshots in RAC
MMON Coordinator
In-memory
statistics
SYSAUX
SGA (Inst1)
AWR tables
… 6:00 a.m.
9:00 a.m. s a
7:00 a.m.
)h a
8:00 a.m.
m ฺ br
In-memory
MMON 9:00 a.m.
i ฺ c o deฺ
statistics r e d Gui
@ sic dent
SGA (Instn) e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V in RAC f er
AWR Snapshotsi o
icDatabase n
a11g,s
b r - t r
Fa non
In Oracle AWR has been enhanced to provide more comprehensive support of RAC.
In RAC environments, each AWR snapshot captures data from all active instances within the cluster.
The data for each snapshot set that is captured for all active instances is from roughly the same point
in time. In addition, the data for each instance is stored separately and is identified with an instance
identifier. For example, the buffer_busy_wait statistic shows the number of buffer waits on
each instance. The AWR does not store data that is aggregated from across the entire cluster. That is,
the data is stored for each individual instance.
The statistics snapshots generated by the AWR can be evaluated by producing reports displaying
summary data such as load and cluster profiles based on regular statistics and wait events gathered on
each instance.
The AWR functions in a similar way as Statspack. The difference is that the AWR automatically
collects and maintains performance statistics for problem detection and self-tuning purposes. Unlike
in Statspack, in the AWR there is only one snapshot_id per snapshot across instances.
AWR Reports and RAC: Overview
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
V i f r
eOverview
i
AWR Reports o and RAC:s
an in an AWR report are organized in different sections. A RAC statistics
b ric -trstatistics
TheaRAC-related
F non
section appears after the Top 5 Timed Events. This section contains:
• The number of instances open at the time of the begin snapshot and the end snapshot to indicate
whether instances joined or left between the two snapshots
• Global Cache Load Profile: Essentially lists the number of blocks and messages that are sent
and received, as well as the number of fusion writes
• Global Cache Efficiency Percentages: Indicate the percentage of buffer is divided into buffers
received from the disk, local cache, and remote caches. Ideally, the percentage of disk buffer
access should be close to zero.
• GCS and GES Workload Characteristics: Gives you an overview of the more important
numbers first. Because the global enqueue convert statistics have been consolidated with the
global enqueue get statistics, the report prints only the average global enqueue get time. The
round-trip times for CR and current block transfers follow, as well as the individual sender-side
statistics for CR and current blocks. The average log flush times are computed by dividing the
total log flush time by the number of actual log flushes. Also, the report prints the percentage of
blocks served that actually incurred a log flush.
AWR Reports and RAC: Overview (continued)
• GCS and GES Messaging Statistics: The most important statistic here is the average message
sent queue time on ksxp, which gives a good indicator of how well the IPC works. Average
numbers should be less than 1 ms.
Additional RAC statistics are then organized in the following sections:
• The Global Enqueue Statistics section contains data extracted from V$GES_STATISTICS.
• The Global CR Served Stats section contains data from V$CR_BLOCK_SERVER.
• The Global CURRENT Served Stats section contains data from
V$CURRENT_BLOCK_SERVER.
• The Global Cache Transfer Stats section contains data from
V$INSTANCE_CACHE_TRANSFER.
The Segment Statistics section also includes the GC Buffer Busy Waits, CR Blocks Received, and
CUR Blocks Received information for relevant segments.
h a sa
Note: For more information about wait events and statistics, see the Oracle Database Reference.
ฺ b r )
c o m e ฺ
i ฺ i d
i c red t Gu
@ s den
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Automatic Database Diagnostic Monitor for RAC
Database ADDM
Self-diagnostic engine
s a
Instance ADDM )h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
AWR
… verde is Stu
_ v ila e th
Inst1
r ic io o us Instn
b t
e (fa ense
rd lCopyright
l a V ab
i er
V sDiagnostic
f
i c o
Automatic iDatabase
a n Monitor for RAC
r r
-t11g offers an extension to the set of functionality that increases the database’s
FabDatabase
Oracle
n o n
manageability by offering clusterwide analysis of performance. A special mode of Automatic
Database Diagnostic Monitor (ADDM) analyzes a RAC database cluster and reports on issues that
affect the entire cluster as well as on those that affect individual instances. This mode is called
database ADDM (as opposed to instance ADDM, which already existed with Oracle Database 10g).
Database ADDM for RAC is not simply a report of reports. Rather, it has independent analysis that is
appropriate for RAC.
Note: The Database ADDM report is generated on the AWR snapshot coordinator.
Automatic Database Diagnostic Monitor for RAC
• Identifies the most critical performance problems for

the entire RAC cluster database
• Runs automatically when taking AWR snapshots
• Performs database-wide analysis of:
– Global resources (for example I/O and global locks)
– High-load SQL and hot blocks a
– Global cache interconnect traffic h as
ฺ b r)
– Network latency issues m eฺ
ฺ c o
– Skew in instance response times edi uid
icr tG
s den
• Is used by DBAs to analyze cluster@performance
• Eliminates need to investigate e r
n
dereports
S tu
to spot
i l a v this
common problems io_v se
r ic t o u
b
(fa ense
e ic © 2008, Oracle. All rights reserved.
rd lCopyright
l
e
V ab e
i l a er
V sDiagnostic
o
Automatic iDatabase n f Monitor for RAC (continued)
r i c t r a
Fab nDatabase
In Oracle
on- 11g, you can create a period analysis mode for ADDM that analyzes the
throughput performance for an entire cluster. When the advisor runs in this mode, it is called
database ADDM. You can run the advisor for a single instance, which is equivalent to Oracle
Database 10g ADDM and is now called instance ADDM.
Database ADDM has access to AWR data generated by all instances, thereby making the analysis of
global resources more accurate. Both database and instance ADDM run on continuous time periods
that can contain instance startup and shutdown. In the case of database ADDM, there may be several
instances that are shut down or started during the analysis period. You must, however, maintain the
same database version throughout the entire time period.
Database ADDM runs automatically after each snapshot is taken. The automatic instance ADDM
runs are the same as in Oracle Database 10g. You can also perform analysis on a subset of instances
in the cluster. This is called partial analysis ADDM.
I/O capacity finding (the I/O system is overused) is a global finding because it concerns a global
resource affecting multiple instances. A local finding concerns a local resource or issue that affects a
single instance. For example, a CPU-bound instance results in a local finding about the CPU.
Although ADDM can be used during application development to test changes to either the
application, the database system, or the hosting machines, database ADDM is targeted at DBAs.
What Does ADDM Diagnose for RAC?
• Latency problems in interconnect

• Congestion (identifying top instances affecting the
entire cluster)
• Contention (buffer busy, top objects etc.)
• Top consumers of multiblock requests
• Lost blocks a
h as
• Information about interconnect devices (warns about r)
ฺ b
m eฺ
using PUBLIC interfaces) c o
i ฺ uid
edused
• Throughput of devices: How much of iitcris Gby
Oracle and for what purpose (GC,e@ s dPQ)
locks, e nt
u r d S t
e
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Diagnose f er for RAC?
o
What DoesiADDM
ic are:-tran s
b r
F• aWaitnevents
Data sources
on (especially Cluster class and buffer busy)
• ASH
• Instance cache transfer data
• Interconnect statistics (throughput, usage by component, pings)
ADDM analyzes the effects of RAC for both the entire database (DATABASE analysis mode) and
for each instance (INSTANCE analysis mode).
EM Support for ADDM for RAC
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VADDMsffor er RAC
EM Supportiofor
b r ic -t11g
r anEnterprise Manager displays the ADDM analysis on the Cluster Database home
Fa non
Oracle
page.
Database
On the Automatic Database Diagnostic Monitor (ADDM) page, the Database Activity chart (not
shown here) plots the database activity during the ADDM analysis period. Database activity types are
defined in the legend based on its corresponding color in the chart. Each icon below the chart
represents a different ADDM task, which in turn corresponds to a pair of individual Oracle Database
snapshots saved in the Workload Repository.
In the ADDM Performance Analysis section, the ADDM findings are listed in descending order,
from highest impact to least impact. For each finding, the Affected Instances column displays the
number (m of n) of instances affected. Drilling down further on the findings takes you to the
Performance Findings Detail page. The Informational Findings section lists the areas that do not have
a performance impact and are for informational purpose only.
The Affected Instances chart shows how much each instance is impacted by these findings. The
display indicates the percentage impact for each instance.
Summary

• Determine RAC-specific tuning components
• Tune instance recovery in RAC
• Determine RAC-specific wait events, global enqueues,
and system statistics
• Implement the most common RAC tuning tips a
h as
• Use the Cluster Database Performance pages r)
ฺ b
m eฺ
• Use the Automatic Workload Repository in RAC ฺ c o
r e di Guid
• Use Automatic Database Diagnostic sMonitor
ic tin RAC
@ d e n
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
This practice covers studying a scalability case by using

the ADDM.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Services
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Configure and manage services in a RAC environment
• Use services with client applications
• Use services with the Database Resource Manager
• Use services with the Scheduler
• Set performance-metric thresholds on services s a
h a
• Configure services aggregation and tracing ฺbr) m eฺ
ฺ c o
r e di Guid
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Traditional Workload Dispatching
Daytime
HR DW CRM Batch
s a
)h a
Payday Holiday season
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
i lav HR t is
hDW
HR DW CRM Batch
o v
_ us e CRM Batch
i
ric e to
a b
(f ens
r lCopyright
l
e
V ab e
i l a er
V sDispatching
o
Traditional iWorkload f
b
In aastandard - t an isolated computing units of different sizes are permanently dedicated to
ric environment,
r
F non
specific applications such as Human Resources, Data Warehouses, Customer Relationship
Management, and Retail Batches.
These computing units need to be sized for their peak workload. As the peak workload occurs for
some hours only, a considerable amount of resources is idle for a long time.
Grid Workload Dispatching
Daytime
Idle
DW HR Batch
CRM
Payday Holiday season s a

)h a
Idle Idle m ฺbr
i ฺ c o deฺ
DW Batch r e d Gui
Batch sic e n t
d @
e Stud
HR
r
ve this DW
CRM_v i l a HR CRM
i o u s e
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
VDispatchingf er
Grid Workloadi o
iccomputing, s
ana global pool of computing units can be provided, and the computing units can
b r - t r
Fa non
With grid
be temporarily assigned to specific applications. Computing units can then be dynamically
exchanged between applications. During business hours, more units can be used for CRM
applications, and after business hours, some of them can be transferred to Retail Batches.
Grid computing minimizes unused resources. This means that overall a grid-enabled environment
needs less computing power than an environment that is not grid enabled.
In the example, 25 percent of the computing resource units are idle. This unused extra capacity is
there so that service levels can still be met in case of failure of components, such as nodes or
instances, and also to deal with unexpected workloads. This is much better than the industry average
of 70 to 90 percent idle rates when each machine is sized for its individual maximum.
Data Warehouse: Example
ETL DB EUT ETL DB EUT
11:00 p.m. 3:00 a.m.

s a
ETL DB EUT
)h a
m ฺbr
i ฺ c o deฺ
r
9:00e d Gui
a.m.
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Example f er
i o
Data Warehouse:
ic building s
ana business intelligence system required the integration of multiple server
b r - t r
Fa non
Previously,
products. The result was that such systems were unnecessarily complex. The integration of multiple
servers was costly. After the system was implemented, there was an ongoing administration cost in
maintaining different servers and keeping the data synchronized across all servers.
• At 11:00 p.m.: ETL server is busy using ETL outside the database; moderate load on database;
no load on end user server.
• At 3:00 a.m.: No load on ETL server; moderate load on database (canned reporting,
aggregation, and potential data mart maintenance); no load on end user server.
• At 9:00 a.m.: No load on ETL server; moderate load on database; end user server is busy using
analysis outside the database.
In addition, each system has to be sized according to the expected workload peaks.
RAC and Data Warehouse: An Optimal Solution
ETL
DB
EUT
s a
)h a
m ฺbr
• Maximum total workload used for system sizing: i ฺ c o deฺ
r e d Gui
Size(Workload max total) < Σ Size(workload max
@ sic dent
components)
• The entire workload is evenly spread
e r de Sacross tu all nodes
at any point in time.
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Next Step
What works for a single data warehouse …
ETL
DB
EUT
s a
)h a
ETL
m ฺbr
i
DBฺ c o deฺ
r e d Gui EUT
@ sic dent
… works in a larger environment as well. rde S tu
e
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
What Is a Service?
• Is a means of grouping sessions that are doing the

same kind of work
• Provides a single-system image instead of a multiple-
instances image
• Is a part of the regular administration tasks that provide
dynamic service-to-instance allocation
s a
• Is the base for High Availability of connections h a
ฺ b r)
• Provides a new performance-tuning dimension om ฺ
e d iฺc uide
s icr ent G
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
i o
What Is a Service?
an was first introduced in Oracle8i as a means for the listener to perform
b ric of -atservice
Theaconcept r
F non
connection load balancing between nodes and instances of a cluster. However, the concept,
definition, and implementation of services have been dramatically expanded. Services are a feature
for workload management that organizes the universe of work execution within the database to make
that work more manageable, measurable, tunable, and recoverable. A service is a grouping of related
tasks within the database with common functionality, quality expectations, and priority relative to
other services. A service provides a single-system image for managing competing applications
running within a single instance and across multiple instances and databases.
Using standard interfaces, such as the DBCA, Enterprise Manager, and SRVCTL, services can be
configured, administered, enabled, disabled, and measured as a single entity.
Services provide availability. Following outages, a service is recovered quickly and automatically at
surviving instances.
Services provide a new dimension to performance tuning. With services, workloads are visible and
measurable. Tuning by “service and SQL” replaces tuning by “session and SQL” in the majority of
systems where sessions are anonymous and shared.
Services are dynamic in that the number of instances a service runs on can be augmented when load
increases, and reduced when load declines. This dynamic resource allocation enables a cost-effective
solution for meeting demands as they occur.
High Availability of Services in RAC
• Services are available continuously with load shared

across one or more instances.
• Additional instances are made available in response to
failures.
• Preferred instances:
– Set the initial cardinality for the service
s a
– Are the first to start the service
)h a
ฺb r
• Available instances are used in response to preferred-
m
o deฺ
i ฺ c
instance failures. ed ui
s icr ent G
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V of Services
f er in RAC
i o
High Availability
ic the -focus s
anof High Availability (HA) is on protecting the logically defined application
b r t r
Fa nonfocus is more flexible than focusing on high availability of instances.
With RAC,
services. This
Services must be location independent and the RAC HA framework is used to implement this.
Services are made available continuously with load shared across one or more instances in the
cluster. Any instance can offer services in response to run-time demands, failures, and planned
maintenance.
To implement the workload balancing and continuous availability features of services, Oracle
Clusterware stores the HA configuration for each service in the Oracle Cluster Registry (OCR). The
HA configuration defines a set of preferred and available instances that support the service.
A preferred instance set defines the number of instances (cardinality) that support the corresponding
service. It also identifies every instance in the cluster that the service will run on when the system
first starts up.
An available instance does not initially support a service. However, it begins accepting connections
for the service when a preferred instance cannot support the service. If a preferred instance fails, then
the service is transparently restored to an available instance defined for the service.
Note: An available instance can become a preferred instance and vice versa.
Possible Service Configuration with RAC
Active/spare
RAC01 RAC02 RAC03
AP AP
GL GL
s a
)h a
Active/symmetric br
Active/asymmetric
m ฺ
o RAC03 ฺ
iฺc uide
RAC01 RAC02 RAC03 RAC01 RAC02
e d
AP AP AP AP
s icr eAP n t G AP
r d e@ Stud
GL GL GL
i l a ve thisGL GL GL
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V Configuration
f
i c i
Possible Service o a n s with RAC
b r r
-t With this service configuration, the simplest redundancy known as
F• aActive/spare:
n o n
primary/secondary, or 1+1 redundancy is extended to the general case of N+M redundancy,
where N is the number of primary RAC instances providing service, and M is the number of
spare RAC instances available to provide the service. An example of this solution is a three-node
configuration in which one instance provides the AP service, the second instance provides the
GL service, and the third instance provides service failover capability for both services. The
spare node can still be available for other applications during normal operation.
• Active/symmetric: With this service configuration, the same set of services is active on every
instance. An example of this is illustrated in the slide, with both AP and GL services being
offered on all three instances. Each instance provides service load-sharing and service failover
capabilities for the other.
• Active/asymmetric: With this service configuration, services with lower capacity needs can be
defined with single cardinality and configured as having all other instances capable of providing
the service in the event of failure. The slide shows the AP service running on only one instance,
and the GL service running on two instances. The first instance supports the AP services and
offers failover for the GL service. Likewise, the second and third instances support the GL
service and offer failover for AP. If either the first or third instance dies, then GL and AP are
still offered through the second instance.
Service Attributes
• Global unique name

• Network name
• Load Balancing Advisory goal*
• Distributed transactions flag*
• Advance queuing notification characteristics for OCI
and ODP.NET clients* a
h as
• Failover characteristics* r)
ฺ b
m eฺ
• Connection load-balancing algorithm* ฺco di Gu
id
• Threshold s icre ent
• Priority r d e@ Stud
l a
• High-availability configuration*
i ve this
v o _ use
i
ric e to
a b
(f ens
d e
r lCopyright
e
V ab e
i
V sferl a
i o
Service Attributes
anservices for your database, you should define each service’s workload
iccreate-trnew
b r
Fa noncharacteristics. The characteristics of a service include:
When you
management
• A unique global name to identify the service
• A Net Service name that a client uses to connect to the service
• A service goal that determines whether work requests are made to the service based on best
service quality (service response time), or best throughput (how much work is completed in a
unit of time), as determined by the Load Balancing Advisory
• An indicator that determines whether the service will be used for distributed transactions
• An indicator that determines whether RAC HA events are sent to OCI and ODP.NET clients that
have registered to receive them through Advanced Queuing
• The characteristics of session failovers when using transparent application failover
• The method for load balancing (which you can define) of connections for each service:
- SHORT: Use Load Balancing Advisory.
- LONG: Use session count by service.
- NONE: Run queue.
• Services metric thresholds (which you can define) for response time and CPU consumption
• Services to consumer groups (which you can map) instead of usernames
• How the service is distributed across instances when the system first starts
Note: Attributes highlighted with an * in the slide cannot be defined for single-instance
environments.
Service Types
• Application services:
– Limit of 100 services per database
• Internal services:
– SYS$BACKGROUND
– SYS$USERS
– Cannot be deleted or changed a
as
)h ฺbr
c m
o deฺ
i ฺ
d Gui
r e
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i
Service Types o
ic -t11g ansupports two broad types of services: application services and internal services.
b r r
Fa non
Oracle Database
Application services are mainly functional maps to workloads. Sessions doing work for a common
business function are grouped together. For Oracle E-Business suite, AP, AR, GL, MFG, WIP, BOM,
and so on create a functional division of work within the database and can thus be categorized as
services.
In addition to application services, the RDBMS also supports two internal services.
SYS$BACKGROUND is used by the background processes only. SYS$USERS is the default service
for user sessions that are not associated with any application service. Both internal services support
all the workload management features and neither one can be stopped or disabled.
There is a limitation of 100 application services per database that you can create. Also, a service
name is restricted to 64 characters.
Note: Shadow services are also included in the application service category. For more information
about shadow services, see the lesson titled “High Availability of Connections.” In addition, a service
is also created for each Advanced Queue created. However, these types of services are not managed
by Oracle Clusterware. Using service names to access a queue provides location transparency for the
queue within a RAC database.
Service Goodness
• Value that reflects the ability of a node and instance to

deliver work for a service
• Appropriate metrics used to compute goodness
depending on the service goal:
– Service time
– Service throughput
• Automatically computed at each instance by MMNL s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
Service Goodness
ic -trisaanmeasure of the attractiveness of an instance to provide resources for a service.
b r
Fa non
Service goodness
MMNL calculates moving average for service time and service throughput. These values are exposed
in GV$SERVICEMETRIC and GV$SERVICEMETRIC_HISTORY.
Note: MMNL is the Manageability MoNitor Light process.
Create Services with Enterprise Manager
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V with sEnterprise
f
i c i
Create Serviceso a n Manager
On a
r
b Cluster r
-tDatabase home page, click the Availability tab. On the Availability tabbed page,
F your n o n
click Cluster Managed Database Services. On the Cluster Managed Database Services page, click
Create Service.
Use the Create Service page to configure a new service for which you do the following:
• Select the desired service policy for each instance configured for the cluster database.
• Select the desired service properties. Refer to the “Service Attributes” topic in this lesson for
more information about the properties you can specify on this page. The Transparent
Application Failover (TAF) policy attribute on this page does not configure server-side TAF.
Note: Although Enterprise Manager configures Oracle Clusterware resources for your newly created
services, it does not generate the corresponding entries in your tnsnames.ora files. You have to
manually edit them. For that, you can use the srvctl config database command with the –
t option, which displays the TNS entries that you should use for the services created with srvctl.
Here is an example:
$ srvctl config database -d xwkE –t
Example client-side TNS entry for service Standard:
Standard = (DESCRIPTION=(ADDRESS=(PROTOCOL=TCP)(HOST=db_vip)
(PORT=dedicated_port))(CONNECT_DATA=(SERVICE_NAME=Standard))) …
Create Services with SRVCTL
$ srvctl add service –d PROD –s GL -r RAC02 -a RAC01

$ srvctl add service –d PROD –s AP –r RAC01 -a RAC02
s a
RAC02
)h a
m ฺ br
AP GL AP GL
i ฺ c o deฺ
r e d Gui
RAC01
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V with sSRVCTL
f
i o
Create Services n
b ric in-the
Theaexample t r aslide
non
shows a two-node cluster with an instance named RAC01 on one node and
F
an instance called RAC02 on the other. The cluster database name is PROD.
Two services—AP and GL—are created and stored in the cluster repository to be managed by Oracle
Clusterware. The AP service is defined with a preferred instance of RAC01 and an available instance
of RAC02.
If RAC01 dies, the AP service member on RAC01 is restored automatically on RAC02. A similar
scenario holds true for the GL service.
Note that it is possible to assign more than one instance with both the -r and -a options. However,
-r is mandatory whereas -a is optional.
Services enable you to move beyond the simple two-node primary/secondary configuration of RAC
Guard in Oracle9i.
With Oracle Database 11g, multiple primary nodes can support a service with RAC. Possible
configurations for service placement are active/spare, active/symmetric, and active/asymmetric.
Once a service is created, you can use the srvctl add service command with the –u –r or –
u –a option to add a preferred or available instance to a service.
Note: You can also set up a service for Transparent Application Failover by using the –P option of
SRVCTL. Possible values are NONE, BASIC, and PRECONNECT.
Preferred and Available Instances
$ srvctl add service –d PROD –s ERP \

–r RAC01,RAC02 -a RAC03,RAC04
1 2
RAC01 RAC02 RAC03 RAC04 RAC01 RAC02 RAC03 RAC04
ERP ERP ERP ERP ERP ERP ERP ERP

s a
)h a
m ฺbr
3 iฺc
o deฺ
4
r e d Gui
RAC01 RAC02 RAC03 RAC04 RAC01
@ sic dRAC03
RAC02
e nt RAC04
ERP ERP ERP ERP eERP r de ERP S tu ERP ERP
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
V i f r
eInstances
Preferred andi o Availables
ic -titrisanassumed that you have a four-node cluster.
b r
Fa non
In this example,
You define a service called ERP. The preferred instances for ERP are RAC01 and RAC02. The
available instances for ERP are RAC03 and RAC04.
1. Initially, ERP connections are directed only to RAC01 and RAC02.
2. RAC02 is failing and goes down.
3. Oracle Clusterware detects the failure of RAC02, and because the cardinality of ERP is 2,
Oracle Clusterware restores the service on one of the available instances, in this case RAC03.
4. ERP connection requests are now directed to RAC01 and RAC03, which are the instances that
currently offer the service. Although Oracle Clusterware is able to restart RAC02, the ERP
service does not fall back to RAC02. RAC02 and RAC04 are now the instances that are
accessible if subsequent failures occur.
Note: If you want to fall back to RAC02, you can use SRVCTL to relocate the service. This operation
can be done manually by the DBA, or by coding the SRVCTL relocation command using a callback
mechanism to automate the fallback. However, relocating a service is a disruptive operation.
Modify Services with the DBMS_SERVICE Package
Modify a service in RAC with the following:

• SRVCTL
• Enterprise Manager
• DBMS_SERVICE.MODIFY_SERVICE
exec DBMS_SERVICE.MODIFY_SERVICE (
'SELF-SERVICE', 'SELF-SERVICE.us.oracle.com',
s a
h a
ฺbr)
goal => DBMS_SERVICE.GOAL_SERVICE_TIME,
clb_goal => DBMS_SERVICE.CLB_GOAL_SHORT);
c o m eฺ
ฺ
di Guid
r e
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V withsthe
f erDBMS_SERVICE Package
o
Modify Services
i
b ric -tran package supports the management of services in the database for the purposes
TheaDBMS_SERVICE
F n
nomeasurement,
of workload management, prioritization, and distributed transaction management. This
package allows the creation, deletion, starting, and stopping of services in both RAC and a single
instance. Additionally, it provides the ability to disconnect all sessions that connect to the instance
with a service name when RAC removes that service name from the instance. Although the preferred
method to create a service in a RAC environment is to use SRVCTL or Enterprise Manager, you can
use the DBMS_SERVICE.CREATE_SERVICE procedure to create a service in a single-instance
environment. This is because the DBMS_SERVICE package is not integrated with Oracle
Clusterware to define preferred and available instances for the service.
However, you can use the DBMS_SERVICE.MODIFY_SERVICE procedure to modify some of the
service’s attributes in a RAC environment that cannot be modified using Enterprise Manager (for
example, the FAILOVER_RETRIES parameter).
The example in the slide shows you how to use DBMS_SERVICE.MODIFY_SERVICE to set the
Load Balancing Advisory goal for SELF-SERVICE. Refer to the “Service Attributes” topic in this
lesson for more information about these attributes.
Note: For more information about the DBMS_SERVICE package, see the PL/SQL Packages and
Types Reference.
Everything Switches to Services
• Data dictionary maintains services.

• The AWR measures the performance of services.
• The Database Resource Manager uses services in
place of users for priorities.
• Job scheduler, Parallel Execution, and Streams queues
run under services. a
h as
• RAC keeps services available within a site. r)
ฺ b
m eฺ
• Data Guard Broker with RAC keeps primary c services
o
i ฺ d
available across sites. red Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i r
V stofeServices
o
EverythingiSwitches
an support services. A session is tracked by the service with which it connects.
ic -tfeatures
b r r
Fa non
Several database
In addition, performance-related statistics and wait events are also tracked by services.
The Automatic Workload Repository (AWR) manages the performance of services. It records the
service performance, including SQL execution times, wait classes, and resources consumed by
service. The AWR alerts the DBA when service response time thresholds are exceeded. Specific
dynamic performance views report current service status with one hour of history.
In Oracle Database 11g, the Database Resource Manager is capable of managing services for
prioritizing application workloads within an instance. In addition, jobs can now run under a service,
as opposed to a specific instance. Parallel slave processes inherit the service of their coordinator.
The RAC HA framework keeps services available within a site. Data Guard Broker, in conjunction
with RAC, migrates the primary service across Data Guard sites for disaster tolerance.
Use Services with Client Applications
ERP=(DESCRIPTION=
(LOAD_BALANCE=on)
(ADDRESS=(PROTOCOL=TCP)(HOST=node-1vip)(PORT=1521))
(CONNECT_DATA=(SERVICE_NAME=ERP)))
url="jdbc:oracle:oci:@ERP"
s a
)h a
url="jdbc:oracle:thin:@(DESCRIPTION=
m ฺbr
(LOAD_BALANCE=on) i ฺ c o deฺ
(ADDRESS=(PROTOCOL=TCP)(HOST=node-1vip)(PORT=1521)) r e d Gui
@ sic dent
e r de Stu
v i lav e this
o_ o us
(CONNECT_DATA=(SERVICE_NAME=ERP)))"
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
V Client f erApplications
i
Use Services o with s
an connection pools select a service by using the TNS connection descriptor.
ic and-trmid-tier
b r
Fa non
Applications
The selected service must match the service that has been created using SRVCTL or the DBCA.
The address lists in each example in the slide use virtual IP addresses. Using the virtual IP addresses
for client communication ensures that connections and SQL statements issued against a node that is
down do not result in a TCP/IP timeout.
The first example in the slide shows the TNS connect descriptor that can be used to access the ERP
service.
The second example shows the thick JDBC connection description using the previously defined TNS
connect descriptor.
The third example shows the thin JDBC connection description using the same TNS connect
descriptor.
Note: The LOAD_BALANCE=ON clause is used by Oracle Net to randomize its progress through the
protocol addresses of the connect descriptor. This feature is called client connection load balancing.
Use Services with the Resource Manager
• Consumer groups are automatically assigned to

sessions based on session services.
• Work is prioritized by service inside one instance.
Instance resources s a
AP
)h a
m ฺbr
AP i
75%ฺ c o deฺ
Connections
r e d Gui
@ sic dent
BATCH e r de Stu 25%
BATCH
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V thesResource
f
Use Services
i c i o with
a n Manager
r -t r
F b nonResource
TheaDatabase Manager (also called Resource Manager) enables you to identify work by
using services. It manages the relative priority of services within an instance by binding services
directly to consumer groups. When a client connects by using a service, the consumer group is
assigned transparently at connect time. This enables the Resource Manager to manage the work
requests by service in the order of their importance.
For example, you define the AP and BATCH services to run on the same instance, and assign AP to a
high-priority consumer group and BATCH to a low-priority consumer group. Sessions that connect
to the database with the AP service specified in their TNS connect descriptor get priority over those
that connect to the BATCH service.
This offers benefits in managing workloads because priority is given to business functions rather than
the sessions that support those business functions.
Note: The Database Resource Manager applies only when the system resources are under heavy
utilization. If there are free CPU cycles, BATCH (in the slide example) could get more than 25
percent.
Services and Resource Manager with EM
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfeManagerr
Services and
i c i o Resource
a n with EM
r t r
Fab nManager
Enterprise
on- (EM) presents a GUI through the Consumer Group Mapping page to
automatically map sessions to consumer groups. You can access this page by clicking the Consumer
Group Mappings link on the Server page.
Using the General tabbed page of the Consumer Group Mapping page, you can set up a mapping of
sessions connecting with a service name to consumer groups (as shown in the right part of the slide).
With the ability to map sessions to consumer groups by service, module, and action, you have greater
flexibility when it comes to managing the performance of different application workloads.
Using the Priorities tabbed page of the Consumer Group Mapping page, you can change priorities for
the mappings that you set up on the General tabbed page. The mapping options correspond to
columns in V$SESSION. When multiple mapping columns have values, the priorities you set
determine the precedence for assigning sessions to consumer groups.
Note: You can also map a service to a consumer group directly on the Create Service page (as shown
in the left part of the slide).
Services and the Resource Manager: Example
exec DBMS_RESOURCE_MANAGER.CREATE_PENDING_AREA;
exec DBMS_RESOURCE_MANAGER.CREATE_CONSUMER_GROUP(
CONSUMER_GROUP => 'HIGH_PRIORITY',
COMMENT => 'High priority consumer group');
exec DBMS_RESOURCE_MANAGER.SET_CONSUMER_GROUP_MAPPING(
ATTRIBUTE => DBMS_RESOURCE_MANAGER.SERVICE_NAME,
VALUE => 'AP',
CONSUMER_GROUP => 'HIGH_PRIORITY');
s a
exec DBMS_RESOURCE_MANAGER.SUBMIT_PENDING_AREA;
)h a
m ฺbr
i ฺ c o deฺ
exec - r e d Gui
DBMS_RESOURCE_MANAGER_PRIVS.GRANT_SWITCH_CONSUMER_GROUP(- @ sic dent
GRANTEE_NAME => 'PUBLIC',
e r de Stu
CONSUMER_GROUP => 'HIGH_PRIORITY',
v i lav e this
GRANT_OPTION => FALSE);io_ us
r ic t o
( f ab nse
e rde lCopyright
l a V ab e
i
V Resource
f er Manager: Example
Services and i o the
c tran
ithat s
b r
Assume
F a
n o n- site has two consumer groups called HIGH_PRIORITY and LOW_PRIORITY.
your
These consumer groups map to a resource plan for the database that reflects either the intended ratios
or the intended resource consumption.
Before mapping services to consumer groups, you must first create the consumer groups and the
resource plan for these consumer groups. The resource plan can be priority based or ratio based. The
PL/SQL calls shown in the slide are used to create the HIGH_PRIORITY consumer group, and map
the AP service to the HIGH_PRIORITY consumer group. You can use similar calls to create the
LOW_PRIORITY consumer groups and map the BATCH service to the LOW_PRIORITY consumer
group.
The last PL/SQL call in the example in the slide is executed because sessions are automatically
assigned only to consumer groups for which they have been granted switch privileges. A similar call
should be executed for the LOW_PRIORITY consumer group.
Note: For more information about the Database Resource Manager, refer to the Oracle Database
Administrator’s Guide and PL/SQL Packages and Types Reference.
Use Services with the Scheduler
• Services are associated with Scheduler classes.

• Scheduler jobs have service affinity:
– High Availability
– Load balancing
HOT_BATCH_SERV HOT_BATCH_SERV LOW_BATCH_SERV
s a
Job coordinator Job coordinator Job coordinator
)h a
Job slaves Job slaves Job slaves m ฺbr
i ฺ c o deฺ
r e d Gui
sic dent
Database
Job table @
de Stu
e r
his
lavLOW_BATCH_SERV
Job1 HOT_BATCH_CLASS HOT_BATCH_SERV
Job2 HOT_BATCH_CLASS
v i t
HOT_BATCH_SERV
e
ic i o_ o us
Job3 LOW_BATCH_CLASS
( f abr nse t
e rde lCopyright
l a V ab e
i er
V thesScheduler
f
Use Services i o with
an
b
Justaas
rinicother -environments,
t r the Scheduler in a RAC environment uses one job table for each
F non
database and one job coordinator for each instance. The job coordinators communicate with each
other to keep information current.
The Scheduler can use the services and the benefits they offer in a RAC environment. The service
that a specific job class uses is defined when the job class is created. During execution, jobs are
assigned to job classes and job classes run within services. Using services with job classes ensures
that the work of the Scheduler is identified for workload management and performance tuning.
For example, jobs inherit server-generated alerts and performance thresholds for the service they run
under.
For High Availability, the Scheduler offers service affinity instead of instance affinity. Jobs are not
scheduled to run on any specific instance. They are scheduled to run under a service. So, if an
instance dies, the job can still run on any other instance in the cluster that offers the service.
Note: By specifying the service where you want the jobs to run, the job coordinators balance the load
on your system for better performance.
Services and the Scheduler with EM
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Scheduler f er with EM
i
Services and o the s
atonrun under a specific service, click the Job Classes link in the Database
To a b ric a -job
configure t r
F non
Scheduler section of the Server page. This opens the Scheduler Job Classes page. On the Scheduler
Job Classes page, you can see services assigned to job classes.
When you click the Create button on the Scheduler Job Classes page, the Create Job Class page is
displayed. On this page, you can enter details of a new job class, including which service it must run
under.
Note: Similarly, you can map a service to a job class on the Create Service page as shown at the
bottom of the slide.
Services and the Scheduler with EM
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Scheduler f er with EM (continued)
Services andi o the
icjob class s
aisnset up with the service that you want it to run under, you can create the job.
b r - t r
Fa non
After your
To create the job, click the Jobs link on the Server page. The Scheduler Jobs page appears, on which
you can click the Create button to create a new job. When you click the Create button, the Create Job
page is displayed. This page has different tabs: General, Schedule, and Options. Use the General
tabbed page to assign your job to a job class.
Use the Options page (displayed in the slide) to set the Instance Stickiness attribute for your job.
Basically, this attribute causes the job to be load balanced across the instances for which the service
of the job is running. The job can run only on one instance. If the Instance Stickiness value is set to
TRUE, which is the default value, the Scheduler runs the job on the instance where the service is
offered with the lightest load. If Instance Stickiness is set to FALSE, then the job is run on the first
available instance where the service is offered.
Note: It is possible to set job attributes, such as INSTANCE_STICKINESS, by using the
SET_ATTRIBUTE procedure of the DBMS_SCHEDULER PL/SQL package.
Services and the Scheduler: Example
DBMS_SCHEDULER.CREATE_JOB_CLASS(
JOB_CLASS_NAME => 'HOT_BATCH_CLASS',
RESOURCE_CONSUMER_GROUP => NULL ,
SERVICE => 'HOT_BATCH_SERV' ,
LOGGING_LEVEL => DBMS_SCHEDULER.LOGGING_RUNS,
LOG_HISTORY => 30, COMMENTS => 'P1 batch');
s a
DBMS_SCHEDULER.CREATE_JOB(
h a
JOB_NAME => 'my_report_job',
ฺ b r)
JOB_TYPE => 'stored_procedure',
c o m eฺ
ฺ
di Guid
JOB_ACTION => 'my_name.my_proc();',
r e
NUMBER_OF_ARGUMENTS => 4, START_DATE =>
@ sicSYSDATE+1,
e n t
REPEAT_INTERVAL => 5, END_DATE =>
r d S ud
eSYSDATE+30,
t
i l a
JOB_CLASS => 'HOT_BATCH_CLASS', ve ENABLED
t h is => TRUE,
_v =>se'daily status');
AUTO_DROP => false, COMMENTS
io o u
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
V Scheduler:
f er Example
Services and i o the
ic -example, s
an you define a batch queue, HOT_BATCH_CLASS, managed by the
b
In this r
PL/SQL t r
Fa non
Scheduler. You associate the HOT_BATCH_SERV service to the HOT_BATCH_CLASS queue. It is
assumed that you had already defined the HOT_BATCH_SERV service.
After the class is defined, you can define your job. In this example, the MY_REPORT_JOB job
executes in the HOT_BATCH_CLASS job class at instances offering the HOT_BATCH_SERV
service.
In this example, you do not assign a resource consumer group to the HOT_BATCH_CLASS job class.
However, it is possible to assign a consumer group to a class. Regarding services, this allows you to
combine Scheduler jobs and service prioritization by using the Database Resource Manager.
Note: For more information about the Scheduler, refer to the Oracle Database Administrator’s Guide
and PL/SQL Packages and Types Reference.
Use Services with Parallel Operations
• Slaves inherit the service from the coordinator.

• Slaves only execute on instances running the service.
ERP
ERP ERP
Node 1 ERP Node 2 ERP ERP Node 3 Node 4 s a

)h a
ฺbr
ERP ERP ERP
c m
o deฺ
i ฺ
d Executionu i
r e G
@ sic dencoordinator
t
Shared disks e r de Stu Parallel
v i lav e this Execution
i _
o o us server
ic
( f abr nse t
e rde lCopyright
l a V ab e
i
V Parallel f er Operations
Use Servicesi o with s
anparallel DML operations, the parallel query slaves inherit the service from the
b ric query
Foraparallel - t r and
F non
query coordinator for the duration of the operation. ERP is the name of the service used by the
example shown in the slide.
By default—and starting with Oracle RAC 11gR1—services restrict the set of instances that are used
by a parallel query. Connecting via a service and then issuing a parallel query can use only those
instances that are part of the service that was specified during the connection. This is implemented by
automatically modifying the INSTANCE_GROUPS parameter to reflect SERVICE_NAMES and by
using the service name used to connect to select the instance group to use for parallel operations
(unless you specified PARALLEL_INSTANCE_GROUP to a different value).
To override this behavior, set a value for the INSTANCE_GROUPS and
PARALLEL_INSTANCE_GROUP initialization parameters. In that case, a slave appears to belong
under the service even on an instance that does not support the service, if that slave is being used by a
query coordinator that was started on an instance that does support that service. At the end of the
execution, the slaves revert to the default database service. Note, however, that
INSTANCE_GROUPS is a deprecated initialization parameter and is retained only for backward
compatibility purposes.
Use Services with Metric Thresholds
• You can define service-level thresholds:

– ELAPSED_TIME_PER_CALL
– CPU_TIME_PER_CALL
• Server-generated alerts are triggered on threshold
violations.
• You can react on generated alerts: a
as
– Change priority. )h ฺbr
– Relocate services. m
i ฺ c o deฺ
– Add instances for services.
r e d Gui
@ sic dent
SELECT service_name, elapsedpercall, e r de cpupercall
S tu
FROM V$SERVICEMETRIC;
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Metric f erThresholds
Use Servicesi o with
ic thresholds s
an permit the comparison of actual service levels against the accepted
b r - t r
Fa
Service-level
on
minimumnrequired level. This provides accountability with respect to delivery or failure to deliver an
agreed service level.
You can explicitly specify two metric thresholds for each service on a particular instance:
• The response time for calls, ELAPSED_TIME_PER_CALL: The response time goal indicates
a desire for the elapsed time to be, at most, a certain value. The response time represents the wall
clock time. It is a fundamental measure that reflects all delays and faults that block the call from
running on behalf of the user.
• CPU time for calls: CPU_TIME_PER_CALL
The AWR monitors the service time and publishes AWR alerts when the performance exceeds the
thresholds. You can then respond to these alerts by changing the priority of a job; stopping
overloaded processes; or relocating, expanding, shrinking, starting, or stopping a service. Using
automated tasks, you can automate the reaction. This enables you to maintain service quality despite
changes in demand.
Note: The SELECT statement shown in the slide gives you the accumulated instance statistics for
elapsed time and for CPU-used metrics for each service for the most recent 60-second interval. For
the last-hour history, look at V$SERVICEMETRIC_HISTORY.
Change Service Thresholds by Using EM
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Thresholds
f er by Using EM
i o
Change Service
c tran s
b riset
Youacan - values for your services from the Database Instance Metric and Policy Settings
threshold
F non
page. You can access this page from the Database Instance home page by clicking the Metric and
Policy Settings link in the Related Links section.
Using the Metric and Policy Settings page, you can set the Service CPU Time (per user call) and
Service Response Time (per user call) metrics for your services. If you modify the critical and
warning values on this page, the thresholds apply to all services of the instance. If you want different
thresholds for different services, click the corresponding icon on the last column of the table as
shown in the slide. This takes you to the corresponding Edit Advanced Settings page. There, you can
click Add to add rows to the Monitored Objects table. Each row represents a particular service in this
case.
Note: You can directly set service thresholds from the Create Service page as shown at the bottom of
the slide.
Services and Metric Thresholds: Example
exec DBMS_SERVER_ALERT.SET_THRESHOLD(-
METRICS_ID => dbms_server_alert.elapsed_time_per_call,
WARNING_OPERATOR => dbms_server_alert.operator_ge,
WARNING_VALUE => '500000',
CRITICAL_OPERATOR => dbms_server_alert.operator_ge,
CRITICAL_VALUE => '750000',
OBSERVATION_PERIOD => 15,
s a
CONSECUTIVE_OCCURRENCES => 3,
h a
INSTANCE_NAME => 'I0n',
ฺ b r)
o m eฺ
OBJECT_TYPE => dbms_server_alert.object_type_service,
c
OBJECT_NAME => 'ERP'); di ฺ uid
s icre ent G
r d e@ Stud
Thresholds must be set on eachve instance is supporting the
i l a t h
service.
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Thresholds:f er
Services and
i c i o Metric
a n s Example
r
abexample,
In this r
-tthresholds are added for the ERP service for the ELAPSED_TIME_PER_CALL
F o n
n metric measures the elapsed time for each user call for the corresponding service. The
metric. This
time must be expressed in microseconds.
A warning alert is raised by the server whenever the average elapsed time per call for the ERP
service over a 15-minute period exceeds 0.5 seconds three consecutive times.
A critical alert is raised by the server whenever the average elapsed time per call for the ERP service
over a 15-minute period exceeds 0.75 seconds three consecutive times.
Note: The thresholds must be created for each RAC instance that potentially supports the service.
Service Aggregation and Tracing
• Statistics are always aggregated by service to measure

workloads for performance tuning.
• Statistics can be aggregated at finer levels:
– MODULE
– ACTION
– Combination of SERVICE_NAME, MODULE, ACTION
s a
• Tracing can be done at various levels:
)h a
– SERVICE_NAMES ฺb r
c m
o deฺ
– MODULE i ฺ
d Gui
r e
– ACTION
@ sic dent
e Stu ACTION
– erd MODULE,
Combination of SERVICE_NAME,
v is
ila e ththat use shared
• This is useful for tuning_vsystems
sessions. r ic io o us
b t
e (fa ense
e rd lCopyright
l a V ab
i
V sand f erTracing
i o
Service Aggregation
anstatistics and wait events are collected for the work attributed to every service.
By a b ric important
default, - t r
F noncan further qualify a service by MODULE and ACTION names to identify the
An application
important transactions within the service. This enables you to locate exactly the poorly performing
transactions for categorized workloads. This is especially important when monitoring performance in
systems by using connection pools or transaction processing monitors. For these systems, the
sessions are shared, which makes accountability difficult.
SERVICE_NAME, MODULE, and ACTION are actual columns in V$SESSION. SERVICE_NAME is
set automatically at login time for the user. MODULE and ACTION names are set by the application
by using the DBMS_APPLICATION_INFO PL/SQL package or special OCI calls. MODULE should
be set to a user-recognizable name for the program that is currently executing. Likewise, ACTION
should be set to a specific action or task that a user is performing within a module (for example,
entering a new customer).
Another aspect of this workload aggregation is tracing by service. The traditional method of tracing
each session produces trace files with SQL commands that can span workloads. This results in a hit-
or-miss approach to diagnose problematic SQL. With the criteria that you provide
(SERVICE_NAME, MODULE, or ACTION), specific trace information is captured in a set of trace
files and combined into a single output trace file. This enables you to produce trace files that contain
SQL that is relevant to a specific workload being done.
Top Services Performance Page
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer Page
Top Servicesi o Performance
icPerformance an page, you can access the Top Consumers page by clicking the Top Consumers
b r - t r
Fa non
From the
link.
The Top Consumers page has several tabs for displaying your database as a single-system image. The
Overview tabbed page contains four pie charts: Top Clients, Top Services, Top Modules, and Top
Actions. Each chart provides a different perspective regarding the top resource consumers in your
database.
The Top Services tabbed page displays performance-related information for the services that are
defined in your database. Using this page, you can enable or disable tracing at the service level, as
well as view the resulting SQL trace file.
Service Aggregation Configuration
• Automatic service aggregation level of statistics

• DBMS_MONITOR used for finer granularity of service
aggregations:
– SERV_MOD_ACT_STAT_ENABLE
– SERV_MOD_ACT_STAT_DISABLE
• Possible additional aggregation levels: a
– SERVICE_NAME/MODULE as
)hฺbr
– SERVICE_NAME/MODULE/ACTION m
i ฺ c o deฺ
• Tracing services, modules, and actions:red u i
ic ent G
– SERV_MOD_ACT_TRACE_ENABLE @s
r d e Stud
– SERV_MOD_ACT_TRACE_DISABLE
i l a ve this
• Database settings persist i o _v across u s e instance restarts.
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V sConfiguration
f
i o
Service Aggregation
c tran
b riinstance,
On a
F each
n o n- important statistics and wait events are automatically aggregated and collected by
service. You do not have to do anything to set this up, except connect with different connect strings
using the services you want to connect to. However, to achieve a finer level of granularity of
statistics collection for services, you must use the SERV_MOD_ACT_STAT_ENABLE procedure in
the DBMS_MONITOR package. This procedure enables statistics gathering for additional hierarchical
combinations of SERVICE_NAME/MODULE and SERVICE_NAME/MODULE/ACTION. The
SERV_MOD_ACT_STAT_DISABLE procedure stops the statistics gathering that was turned on.
The enabling and disabling of statistics aggregation within the service applies to every instance
accessing the database. Furthermore, these settings are persistent across instance restarts.
The SERV_MOD_ACT_TRACE_ENABLE procedure enables tracing for services with three
hierarchical possibilities: SERVICE_NAME, SERVICE_NAME/MODULE, and
SERVICE_NAME/MODULE/ACTION. The default is to trace for all instances that access the
database. A parameter is provided that restricts tracing to specified instances where poor
performance is known to exist. This procedure also gives you the option of capturing relevant waits
and bind variable values in the generated trace files. SERV_MOD_ACT_TRACE_DISABLE disables
the tracing at all enabled instances for a given combination of service, module, and action. Like the
statistics gathering mentioned previously, service tracing persists across instance restarts.
Service Aggregation: Example
• Collect statistics on service and module:

exec DBMS_MONITOR.SERV_MOD_ACT_STAT_ENABLE(-
'AP', 'PAYMENTS');
• Collect statistics on service, module, and action:
exec DBMS_MONITOR.SERV_MOD_ACT_STAT_ENABLE(-
'AP', 'PAYMENTS', 'QUERY_DELINQUENT');
• Trace all sessions of an entire service: s a
)h a
ฺ br
om
exec DBMS_MONITOR.SERV_MOD_ACT_TRACE_ENABLE('AP');
i ฺ c d e ฺ
red t Gu i
• Trace on service, module, and action: i
s denc
@
de Stu
exec DBMS_MONITOR.SERV_MOD_ACT_TRACE_ENABLE(-
e r
'AP', 'PAYMENTS', 'QUERY_DELINQUENT');
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sExample f er
i o
Service Aggregation:
c tran
b ripiece
Theafirst
F nof- sample code begins collecting statistics for the PAYMENTS module within the AP
nosecond
service. The example collects statistics only for the QUERY_DELINQUENT program that
runs in the PAYMENTS module under the AP service. This enables statistics collection on specific
tasks that run in the database.
In the third code box, all sessions that log in under the AP service are traced. A trace file is created
for each session that uses the service, regardless of the module and action. To be precise, you can
trace only specific tasks within a service. This is illustrated in the last example, where all sessions of
the AP service that execute the QUERY_DELINQUENT action within the PAYMENTS module are
traced.
Tracing by service, module, and action enables you to focus your tuning efforts on specific SQL,
rather than sifting through trace files with SQL from different programs. Only the SQL statements
that define this task are recorded in the trace file. This complements collecting statistics by service,
module, and action because relevant wait events for an action can be identified.
Note: For more information about the DBMS_MONITOR package, refer to the PL/SQL Packages and
Types Reference.
trcsess Utility
Client Client Client Clients

CRM ERP CRM CRM ERP CRM
Shared Shared Shared

Dedicated Dedicated Dedicated
server server server
server server server
Trace Trace Trace Trace Trace Trace

file file file
file file file
s a
)h a
TRCSESS TRCSESS m ฺbr
i ฺ c o deฺ
r e d Gui
Trace file
TKPROF
@ sicforTrace
e n t
file
for CRM service
r d e Stud one client
i l a ve this
i o _vfile use
Report
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
i o
trcsess Utility
anconsolidates trace output from selected trace files on the basis of several
b ric -utility
Theatrcsess t r
F non
criteria: session ID, client ID, service name, action name, and module name. After trcsess merges
the trace information into a single output file, the output file can be processed by tkprof.
When using the DBMS_MONITOR.SERV_MOD_ACT_TRACE_ENABLE procedure, tracing
information is present in multiple trace files and you must use the trcsess tool to collect it into a
single file.
The trcsess utility is useful for consolidating the tracing of a particular session or service for
performance or debugging purposes.
Tracing a specific session is usually not a problem in the dedicated server model because a single
dedicated process serves a session during its lifetime. All the trace information for the session can be
seen from the trace file belonging to the dedicated server serving it. However, tracing a service might
become a complex task even in the dedicated server model.
Moreover, in a shared-server configuration, a user session is serviced by different processes from
time to time. The trace pertaining to the user session is scattered across different trace files belonging
to different processes. This makes it difficult to get a complete picture of the life cycle of a session.
Service Performance Views
• Service, module, and action information in:

– V$SESSION
– V$ACTIVE_SESSION_HISTORY
• Service performance in:
– V$SERVICE_STATS
– V$SERVICE_EVENT
– V$SERVICE_WAIT_CLASS s a
h a
– V$SERVICEMETRIC ฺ b r)
– c o m eฺ
V$SERVICEMETRIC_HISTORY ฺ
di Guid
– r e
sic dent
V$SERV_MOD_ACT_STATS
– DBA_ENABLED_AGGREGATIONS e@
– DBA_ENABLED_TRACES ave
rd s Stu
i il th
_ v e
• Twenty-eight statistics
r ic t us
io foroservices
( f ab nse
e rde lCopyright
l a V ab e
i
V sViews f er
i o
Service Performance
anand action information are visible in V$SESSION and
b ric module,
Theaservice, - t r
F non
V$ACTIVE_SESSION_HISTORY.
The call times and performance statistics are visible in V$SERVICE_STATS,
V$SERVICE_EVENT, V$SERVICE_WAIT_CLASS, V$SERVICEMETRIC, and
V$SERVICEMETRIC_HISTORY.
When statistics collection for specific modules and actions is enabled, performance measures are
visible at each instance in V$SERV_MOD_ACT_STATS.
There are more than 300 performance-related statistics that are tracked and visible in V$SYSSTAT.
Of these, 28 statistics are tracked for services. To see the statistics measured for services, run the
following query: SELECT DISTINCT stat_name FROM v$service_stats
Of the 28 statistics, DB time and DB CPU are worth mentioning. DB time is a statistic that
measures the average response time per call. It represents the actual wall clock time for a call to
complete. DB CPU is an average of the actual CPU time spent per call. The difference between
response time and CPU time is the wait time for the service. After the wait time is known, and if it
consumes a large percentage of response time, then you can trace at the action level to identify the
waits.
Note: DBA_ENABLED_AGGREGATIONS displays information about enabled on-demand statistic
aggregation. DBA_ENABLED_TRACES displays information about enabled traces.
Generalized Trace Enabling
• For all sessions in the database:
EXEC dbms_monitor.DATABASE_TRACE_ENABLE(TRUE,TRUE);
EXEC dbms_monitor.DATABASE_TRACE_DISABLE();
• For a particular session: s a

)h a
EXEC dbms_monitor.SESSION_TRACE_ENABLE(session_id ฺb => r
c m
o=> deฺ
27, serial_num => 60, waits => TRUE, binds i ฺ
d Gui
r e
FALSE); sic t
@ d e n
e r de Stu
EXEC dbms_monitor.SESSION_TRACE_DISABLE(session_id
v i lav e this
=> 27, serial_num => 60);
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Enabling f er
Generalizedi oTrace
c tran s
b riuse
Youacan - to debug performance problems. Trace-enabling procedures have been
tracing
F non
implemented as part of the DBMS_MONITOR package. These procedures enable tracing globally for
a database.
You can use the DATABASE_TRACE_ENABLE procedure to enable instancewide session-level SQL
tracing. The procedure has the following parameters:
• WAITS: Specifies whether wait information is to be traced
• BINDS: Specifies whether bind information is to be traced
• INSTANCE_NAME: Specifies the instance for which tracing is to be enabled. Omitting
INSTANCE_NAME means that the session-level tracing is enabled for the whole database.
Use the DATABASE_TRACE_DISABLE procedure to disable SQL tracing for the whole database or
a specific instance.
Similarly, you can use the SESSION_TRACE_ENABLE procedure to enable tracing for a given
database session identifier (SID), on the local instance. The SID and SERIAL# information can be
found from V$SESSION.
Use the SESSION_TRACE_DISABLE procedure to disable the trace for a given database session
identifier (SID) and serial number.
Manage Services
• Use EM or SRVCTL to manage services:

– Start: Allow connections
– Stop: Prevent connections
– Enable: Allow automatic restart and redistribution
– Disable: Prevent starting and automatic restart
– Relocate: Temporarily change instances on which
services run s a
) h a
– Modify: Modify preferred and available instances ฺbr
c o m eฺ
– Get status information ฺ
di Guid
r e
– Add or remove sic nt
• Use the DBCA: r d e@ Stude
– Add or remove i l a ve this
i o _v use
– Modify services ric to
a b
(f ens e
r lCopyright
l
e
V ab e
i
V sferl a
i o
Manage Services
ic on the an of management tasks that you want to perform, you can use Enterprise
b r - t r
Fa
Depending type
onDBCA, or SRVCTL.
Manager,nthe
The following is a description of the management tasks related to services in a RAC environment:
• Disabling a service is used to disable a specified service on all or specified instances. The
disable state is used when a service is down for maintenance to prevent inappropriate automatic
Oracle Clusterware restarts. Disabling an entire service affects all the instances by disabling the
service at each instance.
• Enabling a service is used to enable a service to run under Oracle Clusterware for automatic
restart and redistribution. You can enable a service even if that service is stopped. Enable is
the default value when a service is created. If the service is already enabled, then the command
is ignored. Enabled services can be started, and disabled services cannot be started. Enabling an
entire service affects the enabling of the service over all the instances by enabling the service at
each instance.
• Starting a service is used to start a service or multiple services on the specified instance. Only
enabled services can be started. The command fails if you attempt to start a service on an
instance and if the number of instances that are currently running the service already reaches its
cardinality.
Manage Services (continued)
• Stopping is used to stop one or more services globally across the cluster database, or on the
specified instance. Only Oracle Clusterware services that are starting or have started are stopped.
You should disable a service that you intend to keep stopped after you stop that service because
if the service is stopped and is not disabled, then it can be restarted automatically as a result of
another planned operation. This operation can force sessions to be disconnected transactionally.
• Removing a service is used to remove its configuration from the cluster database on all or
specified instances. You must first stop the corresponding service before you can remove it. You
can remove a service from specific instances only.
• Relocating a service is used to relocate a service from a source instance to a target instance. The
target instance must be on the preferred or available list for the service. This operation can force
sessions to be disconnected transactionally. The relocated service is temporary until you
permanently modify the configuration.
• Modifying a service configuration is used to permanently modify a service configuration. The
change takes effect when the service is restarted later. This allows you to move a service s a
from
one instance to another. Additionally, this command changes the instances that arer)tohbe the
a
preferred and available instances for a service. m ฺb
i ฺ c o deฺ
• Displaying the current state of a named service
r e d Gui
When you use the DBCA to add services, the DBCA also configures
@ sicthednete t entries for these
nservice
services and starts them. When you use the DBCA to remove
r d e services,
S t uit stops the service, removes
the Oracle Clusterware resource for the service, and v
a e the
removes is net service entries.
When you create a service with SRVCTL, you_must i l t h
v startseit with a separate SRVCTL command.
ic i o o u
SRVCTL does not support concurrentbexecutionsr tof commands on the same object. Therefore, run
only one SRVCTL command e
a neach
at a(ftime fore sedatabase, service, or other object.
Note: The srvctl stop e rddatabase
e l ic command implicitly does a srvctl stop services
l a V ab l
(because services
V i are r on database). However, a subsequent srvctl start database
dependent
f e
requires ian o a n s
ciexplicit srvctl start service.
a b r - tr
F no n
Manage Services with Enterprise Manager
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V withsfEnterprise
i c i
Manage Serviceso a n Manager
r -t r
F b use
Youacan
n o Enterprise
n Manager to manage services within a GUI framework. The screenshot shown
in the slide is the main page for administering services within RAC. It shows you some basic status
information about a defined service.
To access this page, click the Cluster Managed Database Services link on the Cluster Database
Availability page.
You can perform simple service management such as enabling, disabling, starting, stopping, and
relocating services. All possible operations are shown in the slide.
If you choose to start a service on the Cluster Managed Database Services page, then EM attempts to
start the service on every preferred instance. Stopping the service stops it on all instances that it is
currently running.
To relocate a service, choose the service that you want to administer, select the Manage option from
the Actions drop-down list, and then click Go.
Note: On the Cluster Managed Database Services page, you can test the connection for a service.
Manage Services with Enterprise Manager
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
V withsfEnterprise
i c i
Manage Serviceso a n Manager (continued)
To a
r r
-t Managed Database Service page for an individual service, you must choose a
b thenCluster
F access n o
service from the Cluster Managed Database Services page, select the Manage option from the
Actions drop-down list, and then click Go.
This is the Cluster Managed Database Service page for an individual service. It offers you the same
functionality as the previous page, except that actions performed here apply to specific instances of a
service.
This page also offers you the added functionality of relocating a service to an available instance.
Relocating a service from one instance to another stops the service on the first instance and then
starts it on the second.
Note: This page also shows you the TAF policy set for this particular service. You can directly edit
the service’s properties, or link to the Top Consumers page.
Manage Services: Example
• Start a named service on all preferred instances:

$ srvctl start service –d PROD –s AP
• Stop a service on selected instances:

$ srvctl stop service –d PROD –s AP –i RAC03,RAC04
• Disable a service at a named instance: a

as
$ srvctl disable service –d PROD –s AP –i RAC04 )h
ฺbr
c m
o deฺ
• Set an available instance as a preferred instance: i ฺ
d Gui
r e
$ srvctl modify service –d PROD –s AP @ sicRAC05
-i e n t –r
r d e Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Example f er
i o
Manage Services: s
an some management tasks with services by using SRVCTL.
b ricdemonstrates
Theaslide - t r
F non
Assume that an AP service has been created with four preferred instances: RAC01, RAC02, RAC03,
and RAC04. An available instance, RAC05, has also been defined for AP.
In the first example, the AP service is started on all preferred instances. If any of the preferred or
available instances that support AP are not running but are enabled, then they are started.
The stop command stops the AP service on instances RAC03 and RAC04. The instances
themselves are not shut down, but remain running possibly supporting other services. The AP service
continues to run on RAC01 and RAC02. The intention might have been to perform maintenance on
RAC04, and so the AP service was disabled on that instance to prevent automatic restart of the
service on that instance. The OCR records the fact that AP is disabled for RAC04. Thus, Oracle
Clusterware will not run AP on RAC04 until the service is enabled later.
The last command in the slide changes RAC05 from being an available instance to a preferred one.
This is beneficial if the intent is to always have four instances run the service because RAC04 was
previously disabled.
Do not perform other service operations while the online service modification is in progress.
Note: For more information, refer to the Oracle Real Application Clusters Administrator’s Guide.
Manage Services: Scenario
DW HR DW HR
I1 I2 I3 I4 I5 I6 I1 I2 I3 I4 I5 I6
Payday Holiday season
srvctl modify service –d PROD –s DW –n –i I1,I2,I3,I4 –a I5,I6
s a
srvctl modify service –d PROD –s HR –n –i I5,I6 –a I1,I2,I3,I4
)h a
m ฺbr
i ฺ c o deฺ
srvctl stop service –d PROD –s DW,HR -f
r e d Gui
@ sic dent
e r de Stu
srvctl start service –d PROD –s DW,HR
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Scenario f er
i o
Manage Services: n s
b ricpresents
Theaslide - t r apossible
a change in a service configuration with a minimum down time for your
F
workload. non
It is assumed that you have a six-node cluster where you run two services: DW and HR. Originally, DW
is configured to have I1 and I2 instances as its preferred instances, and I3, I4, I5, and I6 as its
available instances. Similarly, HR is originally configured to have I3, I4, I5, and I6 as its
preferred instances, and I1 and I2 as its available instances. This initial configuration corresponds
to the Payday period (as shown in the left part of the graphic).
During the Holiday season, you need to change your services configuration so that DW is now run on
the first four instances, and HR on the remaining two.
From the top going down, the slide shows you the commands you need to execute to switch your
services configuration.
Note that the –n option of the srvctl modify service commands is used to remove the
initial configuration of your services. The changes take effect when the services are next restarted.
You can also use the –f option for these commands so that the next stop command disconnects
corresponding sessions. Here, you prefer using the –f option with the srvctl stop service
commands, which stop the services globally on your cluster.
You then use the srvctl start service commands to use the newly created service
configuration.
Using Distributed Transactions with RAC
• Scope of application: XA or MS DTC

• Fully supported by default on RAC (GTXn background)
• Or all transaction branches occur on same instance:
dbms_service.modify_service(service_name=>'S1', DTP=>TRUE)
dbms_service.modify_service(service_name=>'S2', DTP=>TRUE)
Mid-tier S1
partition 1 S0 S1 S2 RAC01
s a
)h a
m ฺbr
Mid-tier S0
i ฺ c o deฺ
non-DT S0 S1 S2 RAC02
r e d Gui
@ sic dent
e r de Stu
Mid-tier
partition 2
S2
v i lavS0 e S1 t his S2 RAC03
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Transactions
f er
i c i o
Using Distributed
a n s with RAC
r r
-t with distributed transactions (compliant with XA standard or coordinated by
Fabyounuse
When
onRAC
Microsoft Distributed Transaction Coordinator), it is possible for two application components in the
same transaction to connect to different nodes of a RAC cluster. This situation can occur on systems
with automatic load balancing where the application cannot control which database nodes a
distributed transaction branch gets processed.
Starting with Oracle RAC 11gR1, these types of distributed transactions are automatically controlled
by the system through the use of new background processes called GTX0…GTXj.
GLOBAL_TXN_PROCESSES specifies the initial number of GTXn background processes on an
instance, and its default value is 1. Letting the database handle distributed transactions automatically
is useful for systems that process global transactions heavily.
To provide improved application performance with distributed transaction processing in Oracle
RAC, you may want to take advantage of the specialized service referred to as a DTP service. Using
DTP services, you can direct all branches of a distributed transaction to a single instance in the
cluster. For load balancing across the cluster, it is better to have several groups of smaller application
servers with each group directing its transactions to a single service (or set of services) than to have
one or two larger application servers. The graphic in the slide presents a possible solution. Assume
that you have three RAC nodes—RAC01, RAC02, and RAC03—with each node capable of servicing
any nondistributed transaction coming from a middle tier.
Using Distributed Transactions with RAC (continued)
For distributed transactions from other middle tiers, they are partitioned statically via Oracle Net
Services to one of these three nodes. Thus, each node publishes itself as an S0 service for
nondistributed transactions. In addition, RAC01 and RAC03 publish themselves as a DTP service: S1
and S2, respectively.
As shown in the slide, a DTP service is one that has its DTP flag set to TRUE. In addition, you should
always define a DTP service as a singleton service (that is, with only one preferred instance).
Each mid-tier client has Oracle Net Service Names configuration and accesses the Oracle database
through Oracle Net. Each Oracle Net Service that is being used for distributed transactions is
configured with one DTP service.
The Oracle server ensures the cardinality of a DTP service to be 1 across the RAC cluster,
prohibiting more than one instance of the same DTP service from running in a RAC database. Each
distributed transaction is processed by one of the DTP services via Oracle Net, so that all tightly
coupled branches of the distributed transaction are routed to the same node of the RAC database. s a
Different distributed transactions can be load balanced to different RAC nodes via different ) a
h DTP
services. If one of these database nodes fails, Oracle Clusterware and RAC automatically b r
ฺ detect the
c o m e ฺone of the
failure and do the transaction recovery before starting the corresponding DTP i ฺ servicedon
i
available RAC nodes. When the node comes back, the same DTP service
i c redmayt beGrelocated
u back
automatically depending on the workload. @ s d e n
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Restricted Session and Services
ALTER SYSTEM 2
1
ENABLE RESTRICTED SESSION;
RAC01 RAC02
ERP ERP ERP ERP
s a
)h a
Oracle
4 m ฺbr
Clusterware
i ฺ c o deฺ
r e d Gui
ERP 3
@ sic deERP nt
e r de Stu
5 v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V and f r
eServices
i o
Restricted Session
icyou put n
aone s
b r - t r
Fa non
Whenever instance of the cluster in restricted mode, Oracle Clusterware stops the
services running on the restricted instance, and it starts them on available instances if they exist. That
way, the listeners are dynamically informed of the changes, and they no longer attempt to route
requests to the restricted instance, regardless of its current load. In effect, the listeners exempt the
restricted instance from their connection load-balancing algorithm.
This feature comes with two important considerations:
• First, even users with RESTRICTED SESSION privilege are not able to connect remotely
through the listeners to an instance that is in the restricted mode. They need to connect locally to
the node supporting the instance and use the bequeath protocol.
• Second, this new feature works only when the restricted instance dynamically registers with the
listeners. That is, if you configure the listener.ora file with SID_LIST entries, and you
do not use dynamic registration, the listener cannot block connection attempts to a restricted
instance. In this case, and because the unrestricted instances of the cluster are still accessible, the
restricted instance will eventually become least loaded, and the listener will start routing
connection requests to that instance. Unable to accept the connection request because of its
restricted status, the instance will deny the connection and return an error. This situation has the
potential for blocking access to an entire service.
Note: The listener uses dynamic service registration information before static configurations.
Summary

• Configure and manage services in a RAC environment
• Use services with client applications
• Use services with the Database Resource Manager
• Use services with the Scheduler
• Set performance-metric thresholds on services s a
h a
• Configure services aggregation and tracing ฺbr) m eฺ
ฺ c o
r e di Guid
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non

• Defining services by using DBCA
• Managing services by using Enterprise Manager
• Using server-generated alerts in combination with
services
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
High Availability
of Connections
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Configure client-side connect-time load balancing
• Configure client-side connect-time failover
• Configure server-side connect-time load balancing
• Use the Load Balancing Advisory (LBA)
• Describe the benefits of Fast Application Notification s a
(FAN) r ) ha
m ฺb
• Configure server-side callouts
i ฺ c o deฺ
r e d Gui
• Configure the server- and client-sidesONS
ic ent
@ d
rde Failover
• Configure Transparent Application Stu (TAF) ve this
i l a
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V ssee f erthe following page:
i o
For more information,
ic -tran
b r
http://www.oracle.com/technology/products/database/clustering/pdf/awmrac11g.pdf
Fa non
Types of Workload Distribution
• Connection balancing is rendered possible by

configuring multiple listeners on multiple nodes:
– Client-side connect-time load balancing
– Client-side connect-time failover
– Server-side connect-time load balancing
• Run-time connection load balancing is rendered
s a
possible by using connection pools: a
b of
ฺpool r)h
– Work requests automatically balanced across the m
connections i ฺ c o deฺ
r e d Gui
sic dent cache and
– Native feature of the JDBC implicit connection
@
ODP.NET connection pool de tu e r sS
i v
la e thi
v
o_ o us
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
V Distribution
f er
i o
Types of Workload
ic -tran s
a b r
With
F RAC,
n o n
multiple listeners on multiple nodes can be configured to handle client connection
requests for the same database service.
A multiple-listener configuration enables you to leverage the following failover and load-
balancing features:
• Client-side connect-time load balancing
• Client-side connect-time failover
• Server-side connect-time load balancing
These features can be implemented either one by one, or in combination with each other.
Moreover, if you are using connection pools, you can benefit from readily available run-time
connection load balancing to distribute the client work requests across the pool of connections
established by the middle tier. This possibility is offered by the Oracle JDBC implicit
connection cache feature as well as Oracle Data Provider for .NET (ODP.NET) connection pool.
Client-Side Connect-Time Load Balancing
ERP =
(DESCRIPTION =
(ADDRESS_LIST =
(LOAD_BALANCE=ON)
(ADDRESS=(PROTOCOL=TCP)(HOST=node1vip)(PORT=1521))
)
s a
)h a
Random
m ฺbr
access i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
node1 vila node2
v this
i o _ use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer Load Balancing
Client-SideioConnect-Time
b r ic -tran
Fa non connect-time load balancing feature enables clients to randomize connection
The client-side
requests among a list of available listeners. Oracle Net progresses through the list of protocol
addresses in a random sequence, balancing the load on the various listeners. Without this
feature, Oracle Net always takes the first protocol address to attempt a connection.
You enable this feature by setting the LOAD_BALANCE=ON clause in the corresponding client-
side TNS entry.
Note: For a small number of connections, the random sequence is not always even.
Client-Side Connect-Time Failover
ERP =
(DESCRIPTION =
(ADDRESS_LIST =
(LOAD_BALANCE=ON)
(FAILOVER=ON) 3
) 4
(CONNECT_DATA=(SERVICE_NAME=ERP))) s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
1
e r de Stu
i lav e this
node2vip
v
_ us
2onode1vip
i
b ic
r e to
a
(f ens
r lCopyright
l
e
V ab e
i l a
V sfer Failover
Client-SideioConnect-Time
b r ic -tran
Fa nonenables clients to connect to another listener if the initial connection to the first
This feature
listener fails. The number of listener protocol addresses in the connect descriptor determines
how many listeners are tried. Without client-side connect-time failover, Oracle Net attempts a
connection with only one listener. As shown by the example in the slide, client-side connect-
time failover is enabled by setting the FAILOVER=ON clause in the corresponding client-side
TNS entry.
In the example, you expect the client to randomly attempt connections to either NODE1VIP or
NODE2VIP, because LOAD_BALANCE is set to ON. In the case where one of the nodes is down,
the client cannot know this. If a connection attempt is made to a down node, the client needs to
wait until it receives the notification that the node is not accessible, before an alternate address
in the ADDRESS_LIST is tried.
Therefore, it is highly recommended to use virtual host names in the ADDRESS_LIST of your
connect descriptors. If a failure of a node occurs (1), the virtual IP address assigned to that node
is failed over and brought online on another node in the cluster (2). Thus, all client connection
attempts are still able to get a response from the IP address, without the need to wait for the
operating system TCP/IP timeout (3). Therefore, clients get an immediate acknowledgement
from the IP address, and are notified that the service on that node is not available. The next
address in the ADDRESS_LIST can then be tried immediately with no delay (4).
Note: If you use connect-time failover, do not set GLOBAL_DBNAME in your listener.ora
Unauthorized
file. reproduction or distribution prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
Server-Side Connect-Time Load Balancing
ERP = (DESCRIPTION=
(ADDRESS_LIST=(LOAD_BALANCE=ON)(FAILOVER=ON)
(ADDRESS=(PROTOCOL=TCP)(HOST=node2vip)(PORT=1521)))
6 4 3 2
ERP started on both instances
Listener Listener
1
5 1 1 s a
)h a
PMON
ฺbr
PMON
m
*.REMOTE_LISTENER=RACDB_LISTENERS
i ฺ c o Node2
d e ฺ
Node1
red t Gu i
i
s denc
RACDB_LISTENERS= @
de Stu
(DESCRIPTION=
e r
i lav e this
(ADDRESS=(PROTOCOL=tcp)(HOST=node1vip)(PORT=1521))
v
i o_ o us
(ADDRESS=(PROTOCOL=tcp)(HOST=node2vip)(PORT=1521)))
ic
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer Load Balancing
Server-Sidei oConnect-Time
ic -tran
b r
a shows
The
F slide
n o n you how listeners distribute service connection requests across a RAC cluster.
Here, the client application connects to the ERP service. On the server side, the database is using
the dynamic service registration feature. This allows the PMON process of each instance in the
cluster to register service performance information with each listener in the cluster (1). Each
listener is then aware of which instance has a particular service started, as well as how that
service is performing on each instance.
You configure this feature by setting the REMOTE_LISTENER initialization parameter of each
instance to a TNS name that describes the list of all available listeners. The slide shows the
shared entry in the SPFILE as well as its corresponding server-side TNS entry.
Depending on the load information, as computed by the Load Balancing Advisory, and sent by
each PMON process, a listener redirects the incoming connection request (2) to the listener of
the node where the corresponding service is performing the best (3).
In the example, the listener on NODE2 is tried first. Based on workload information dynamically
updated by PMON processes, the listener determines that the best instance is the one residing on
NODE1. The listener redirects the connection request to the listener on NODE1 (4). That listener
then starts a dedicated server process (5), and the connection is made to that process (6).
Note: For more information, refer to the Net Services Administrator’s Guide.
Fast Application Notification: Overview
.NET app C app C app Java app Java app
ODP.NET OCI API ONS OCI ONS Java JDBC

API API
ONS ONS
AQ
s a
HA
)h a
Proxy Events
ONS
m ฺbr
app
i ฺ c o deฺ
Callout
HA
CRS
HA
EMD r e d GDB u i
script Events Events
@ sic dent Control
e r de Stu
Callout
exec HA vil
av this
i o _ use Node1
Events
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V Notification:
f
i c i o
Fast Application
a n s Overview
r r
-t Notification (FAN) enables end-to-end, lights-out recovery of applications and
FabApplication
Fast
n o n
load balancing based on real transaction performance in a RAC environment. With FAN, the
continuous service built into Oracle Real Application Clusters 11g is extended to applications
and mid-tier servers. When the state of a database service changes, (for example, up, down, or
not restarting), the new status is posted to interested subscribers through FAN events.
Applications use these events to achieve very fast detection of failures, and rebalancing of
connection pools following failures and recovery.
The easiest way to receive all the benefits of FAN, with no effort, is to use a client that is
integrated with FAN:
• JDBC Implicit Connection Cache
• User extensible callouts
• Connection Manager (CMAN)
• Listeners
• Oracle Notification Service (ONS) API
• OCI Connection Pool or Session Pool
• Transparent Application Failover (TAF)
• ODP.NET Connection Pool
Note: Not all the above applications can receive all types of FAN events.
Fast Application Notification: Benefits
• No need for connections to rely on connection

timeouts
• Used by Load Balancing Advisory to propagate load
information
• Designed for enterprise application and management
console integration
• Reliable distributed system that: s a
)h a
– Detects high-availability event occurrences in a timely
m ฺ br
manner o deฺ
i ฺ c
red Gui
– Pushes notification directly to your applications
• Tightly integrated with: @ sic dent
– Oracle JDBC applications using e r deconnection
S tu pools
– Enterprise Manager _vil
av this
i o u s e
– Data Guard Broker
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V Notification:
f
i c i o
Fast Application
a n s Benefits
r t r
Fab non-client or mid-tier applications connected to the database have relied on connection
Traditionally,
timeouts, out-of-band polling mechanisms, or other custom solutions to realize that a system
component has failed. This approach has huge implications in application availability, because
down times are extended and more noticeable.
With FAN, important high-availability events are pushed as soon as they are detected, which
results in a more efficient use of existing computing resources, and a better integration with your
enterprise applications, including mid-tier connection managers, or IT management consoles,
including trouble ticket loggers and e-mail/paging servers.
FAN is, in fact, a distributed system that is enabled on each participating node. This makes it
very reliable and fault tolerant because the failure of one component is detected by another.
Therefore, event notification can be detected and pushed by any of the participating nodes.
FAN events are tightly integrated with Oracle Data Guard Broker, Oracle JDBC implicit
connection cache, ODP.NET, TAF, and Enterprise Manager. For example, Oracle Database 11g
JDBC applications managing connection pools do not need custom code development. They are
automatically integrated with the ONS if implicit connection cache and fast connection failover
are enabled.
Note: For more information about FAN and Data Guard integration, refer to the lesson titled
“Design for High Availability” in this course.
FAN-Supported Event Types
Event type Description

SERVICE Primary application service
SRV_PRECONNECT Shadow application service event

(mid-tiers and TAF using primary and secondary
instances)
SERVICEMEMBER Application service on a specific instance
DATABASE Oracle database

s a
INSTANCE Oracle instance )h a
m ฺbr
ASM Oracle ASM instance
i ฺ c o deฺ
r e d Gui
NODE Oracle cluster node
@ sic dent
e r de Stu
SERVICE_METRICS Load Balancing
v i lav e this
Advisory
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Event f r
eTypes
i
FAN-Supportedo
ic -tran s
b r
Fa nonevents pertaining to the list of managed cluster resources shown in the slide. The
FAN delivers
table describes each of the resources.
Note: SRV_PRECONNECT and SERVICE_METRICS are discussed later in this lesson.
FAN Event Status
Event status Description

up Managed resource comes up.
down Managed resource goes down.
preconn_up Shadow application service comes up.
preconn_down Shadow application service goes down.

s a
Managed node goes down. )h a
ฺbr
nodedown
Managed resource cannot fail over to a c m

o deฺ
remote
not_restarting
i ฺ
d Gui
node. r e
restart_failed Managed resource fails @ siclocally
to start e n t
after a
d
discrete number of rretries. e Stud
Unknown i l a ve this
Status is unrecognized.
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
FAN Eventio Status
b r ic -tran
Fa nodescribes
This table n the event status for each of the managed cluster resources seen previously.
FAN Event Reasons
Event Reason Description

user User-initiated commands, such as srvctl and sqlplus
failure Managed resource polling checks detecting a failure
dependency Dependency of another managed resource that triggered

a failure condition a
as
unknown Unknown or internal application state when event is
triggered
)h
ฺbr
c m
o attribute
e ฺ
autostart Initial cluster boot: Managed resource has iprofile ฺ i
d lastGOracle
u d
AUTO_START=1, and was offline before
i c rethe t
Clusterware shutdown.
@ s den
Boot Initial cluster boot: Managede r deresourceS tuwas running before
i lav e tshutdown.
the last Oracle Clusterware
v his
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
FAN Eventio Reasons
b r ic -tran
Fa nostatus
The event n for each managed resource is associated with an event reason. The reason
further describes what triggered the event. The table in the slide gives you the list of possible
reasons with a corresponding description.
FAN Event Format
<Event_Type>
VERSION=<n.n>
[service=<serviceName.dbDomainName>]
[database=<dbName>] [instance=<sid>]
[host=<hostname>]
status=<Event_Status>
reason=<Event_Reason>
[card=<n>]
timestamp=<eventDate> <eventTime> s a
)h a
SERVICE VERSION=1.0 service=ERP.oracle.com om
ฺbr
i ฺ c d e ฺ
red t Gu
database=RACDB status=up reason=user card=4 i
timestamp=16-Mar-2004 19:08:15 s i c n
r d e@ Stude
NODE VERSION=1.0 host=strac-1
i l a ve this
status=nodedown timestamp=16-Mar-2004
i o _v use 17:35:53
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
FAN Eventio Format
In a b ric to its
- t r an status, and reason, a FAN event has other payload fields to further
F addition
describen
on type,
the unique cluster resource whose status is being monitored and published:
• The event payload version, which is currently 1.0
• The name of the primary or shadow application service. This name is excluded from NODE
events.
• The name of the RAC database, which is also excluded from NODE events
• The name of the RAC instance, which is excluded from SERVICE, DATABASE, and NODE
events
• The name of the cluster host machine, which is excluded from SERVICE and DATABASE
events
• The service cardinality, which is excluded from all events except for SERVICE
status=up events
• The server-side date and time when the event is detected
The general FAN event format is described in the slide along with possible FAN event examples.
Note the differences in event payload for each FAN event type.
Load Balancing Advisory: FAN Event
Parameter Description
Version Version of the event payload
Event type SERVICE_METRICS
Service Matches DBA_SERVICES
Database unique name Unique DB name supporting the service
Time stamp Date and time stamp (local time zone)

s a
)h a
Repeated
m ฺbr
Instance Instance name supporting the serviceฺco
i d e ฺ
reddatabase i
uand instance
Percent Percentage of work to send to i
s dencthis t G
@
dNOe DATA, u
tUNKNOWN
Flag GOOD, VIOLATING,
e r S
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Advisory:
f er FAN Event
i
Load Balancingo
ic -tran s
a b r
The
F Load
n o n
Balancing Advisory FAN event is described in the slide. Basically, it contains a
calculated percentage of work requests that should be sent to each instance. The flag indicates
the behavior of the service on the corresponding instance relating to the thresholds set on that
instance for the service.
Server-Side Callouts Implementation
• The callout directory:

– <CRS Home>/racg/usrco
– Can store more than one callout
– Grants execution on callout directory and callouts only to
the Oracle Clusterware user
• Callouts execution order is nondeterministic.
s a
• Writing callouts involves:
)h a
1. Parsing callout arguments: The event payload mฺb
r
i ฺ c o deฺ
2. Filtering incoming FAN events
r e d Gui
c
3. Executing event-handling programs si ent e@ Stud
r d
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V sfImplementation
Server-Sidei oCallouts
ic -tran
a b r
Each
F database
n o n event detected by the RAC High Availability (HA) framework results in the
execution of each executable script or program deployed in the standard Oracle Clusterware
callout directory. On UNIX, it is $ORA_CRS_HOME/racg/usrco. Unless your Oracle
Clusterware home directory is shared across the network, you must deploy each new callout on
each RAC node.
The order in which these callouts are executed is nondeterministic. However, RAC guarantees
that all callouts are invoked once for each recognized event in an asynchronous fashion. Thus, it
is recommended to merge callouts whose executions need to be in a particular order.
You can install as many callout scripts or programs as your business requires, provided that each
callout does not incur expensive operations that delay the propagation of HA events. If many
callouts are going to be written to perform different operations based on the event received, it
might be more efficient to write a single callout program that merges each single callout.
Writing server-side callouts involves the steps shown in the slide. In order for your callout to
identify an event, it must parse the event payload sent by the RAC HA framework to your
callout. After the sent event is identified, your callout can filter it to avoid execution on each
event notification. Then, your callout needs to implement a corresponding event handler that
depends on the event itself and the recovery process required by your business.
Note: As a security measure, make sure that the callout directory and its contained callouts have
write permissions
Unauthorized only to
reproduction orthe system user
distribution who installed
prohibitedฺ Oracle Clusterware.
Server-Side Callout Parse: Example
#!/bin/sh
NOTIFY_EVENTTYPE=$1
for ARGS in $*; do
PROPERTY=`echo $ARGS | $AWK -F"=" '{print $1}'`
VALUE=`echo $ARGS | $AWK -F"=" '{print $2}'`
case $PROPERTY in
VERSION|version) NOTIFY_VERSION=$VALUE ;;
SERVICE|service) NOTIFY_SERVICE=$VALUE ;;
DATABASE|database) NOTIFY_DATABASE=$VALUE ;;
INSTANCE|instance) NOTIFY_INSTANCE=$VALUE ;;
s a
HOST|host) NOTIFY_HOST=$VALUE ;;
) h a
STATUS|status) NOTIFY_STATUS=$VALUE ;; ฺbr
c o m eฺ
REASON|reason) NOTIFY_REASON=$VALUE ฺ ;;
di Guid;;
CARD|card) r e
NOTIFY_CARDINALITY=$VALUE
sic dent ;;
TIMESTAMP|timestamp) NOTIFY_LOGDATE=$VALUE
@
??:??:??) r de Stu
NOTIFY_LOGTIME=$PROPERTY
e ;;
esac
v i lav e this
done io_ r ic
s
o u
b t
e (fa ense
rd lCopyright
l a V ab
i
V sParse:
f er Example
Server-Sidei oCallout
ic -tran
b r
Fa nowant
Unless you n your callouts to be executed on each event notification, you must first identify
the event parameters that are passed automatically to your callout during its execution. The
example in the slide shows you how to parse these arguments by using a sample Bourne shell
script.
The first argument that is passed to your callout is the type of event that is detected. Then,
depending on the event type, a set of PROPERTY=VALUE strings are passed to identify exactly
the event itself.
The script given in the slide identifies the event type and each pair of PROPERTY=VALUE
string. The data is then dispatched into a set of variables that can be used later in the callout for
filtering purposes.
As mentioned in the previous slide, it might be better to have a single callout that parses the
event payload, and then executes a function or another program on the basis of information in
the event, as opposed to having to filter information in each callout. This becomes necessary
only if many callouts are required.
Note: Make sure that executable permissions are set correctly on the callout script.
Server-Side Callout Filter: Example
if ((( [ $NOTIFY_EVENTTYPE = "SERVICE" ] ||

[ $NOTIFY_EVENTTYPE = "DATABASE" ] || \
[ $NOTIFY_EVENTTYPE = "NODE" ] \
) && \
( [ $NOTIFY_STATUS = "not_restarting" ] || \
[ $NOTIFY_STATUS = "restart_failed" ] \
)) && \
( [ $NOTIFY_DATABASE = "HQPROD" ] || \
[ $NOTIFY_SERVICE = "ERP" ] \
s a
)) h a
then ฺ b r)
/usr/local/bin/logTicket $NOTIFY_LOGDATEco\ m eฺ
ฺ
di G \ uid
r
$NOTIFY_LOGTIMEe
sic dent \
$NOTIFY_SERVICE
@
r de Stu
$NOTIFY_DBNAME
e
\
v i l e t his
av$NOTIFY_HOST
fi o_ s
ici to u
b r
e (fa ense
rd lCopyright
l a V ab
i
V sFilter:
f er Example
Server-Sidei oCallout
ic -tran
b r
Fa nonin the slide shows you a way to filter FAN events from a callout script. This
The example
example is based on the example in the previous slide.
Now that the event characteristics are identified, this script triggers the execution of the trouble-
logging program /usr/local/bin/logTicket only when the RAC HA framework posts
a SERVICE, DATABASE, or NODE event type, with a status set to either not_restarting or
restart_failed, and only for the production HQPROD RAC database or the ERP service.
It is assumed that the logTicket program is already created and that it takes the arguments
shown in the slide.
It is also assumed that a ticket is logged only for not_restarting or restart_failed
events, because they are the ones that exceeded internally monitored timeouts and seriously need
human intervention for full resolution.
Configuring the Server-Side ONS
Mid-tier1
localport=6100
remoteport=6200 ONS
useocr=on
$ racgons add_config node1:6200 node2:6200

2
$ racgons add_config midtier1:6200
$ onsctl reconfig 3 3 $ onsctl reconfig s a

)h a
m ฺbr
ONS i ฺ c
ONS o deฺ
r e d Gui
OCR
@ sic dent
ons.config 1 e r de S1tuons.config
v i lav e tNode2 his
Node1
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Server-Side
f er ONS
Configuringi o the
ic -tran s
b r
Fa nconfiguration
The ONS
on is controlled by the <CRS HOME>/opmn/conf/ons.config configuration
file. This file is automatically created during installation. There are three important parameters
that should always be configured for each ONS:
• The first is localport, the port that ONS uses to talk to local clients.
• The second is remoteport, the port that ONS uses to talk to other ONS daemons.
• The third parameter is called nodes. It specifies the list of other ONS daemons to talk to.
This list should include all RAC ONS daemons, and all mid-tier ONS daemons. Node
values are given as either host names or IP addresses followed by its remoteport. Instead,
you can store this data in Oracle Cluster Registry (OCR) using the racgons add_config
command and having the useocr parameter set to on in the ons.config file. By storing
nodes information in OCR, you do not need to edit a file on every node to change the
configuration. Instead, you need to run only a single command on one of the cluster nodes.
In the slide, it is assumed that ONS daemons are already started on each cluster node. This
should be the default situation after a correct RAC installation. However, if you want to use
OCR, you should edit the ons.config file on each node, and then add the configuration to OCR
before reloading it on each cluster node. This is illustrated in the slide.
Note: You should run racgons whenever you add or remove a node that runs an ONS daemon.
To remove a node from OCR, you can use the racgons remove_config command.
Optionally Configure the Client-Side ONS
Mid-tier1
ons.config ONS $ onsctl start 2
localport=6100
remoteport=6200 1
nodes=node1:6200,node2:6200
s a
)h a
m ฺbr
ONS i ฺ c
ONSo deฺ
r e d Gui
OCR
@ sic dent
ons.config e r de Stuons.config
v i lav e tNode2 his
Node1
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfthe er Client-Side ONS
Optionallyio
Configure
b r ic -tran
Fa non 10g Release 1 FAN uses Oracle Notification Service (ONS) on the mid-tier to
Oracle Database
receive FAN events when you are using the Java Database Connectivity (JDBC) Implicit
Connection Cache (ICC). To use ONS on the mid-tier, you need to install ONS on each host
where you have client applications that need to be integrated with FAN. Most of the time, these
hosts play the role of a mid-tier application server. Therefore, on the client side, you must
configure all the RAC nodes in the ONS configuration file. A sample configuration file might
look like the one shown in the slide.
After configuring ONS, you start the ONS daemon with the onsctl start command. It is
your responsibility to make sure that an ONS daemon is running at all times. You can check that
the ONS daemon is active by executing the onsctl ping command.
Note: With Oracle Database 10g Release 2 and later, there is no requirement to use ONS
daemons on the mid-tier when using the JDBC Implicit Connection Cache. To configure this
option, use either the OracleDataSource property or a setter API
setONSConfiguration(configStr). The input to this API is the contents of the
ons.config file specified as a string. For example,
setONSConfiguration("nodes=host1:port1,host2:port2 ");
The ons.jar file must be on the client’s CLASSPATH. There are no daemons to start or
manage.
JDBC Fast Connection Failover: Overview
Mid-tier1
Service Service or node
ONS
UP event DOWN event
JDBC ICC
Event
handler
Connections Connections
reconnected Connection Cache marked down &
s a
cleaned up
)h a
Connections Connections ฺbr
using usingcom e ฺ
Listeners
i ฺ i d
service names
Connections i cr nted namesu
service
G
load balancing @ s de
ONS ……… e r d eONS S t u
v i lav e this
Node1
ic i o_ o us Noden
( f abr nse t
e rde lCopyright
l a V ab e
i r
V sfeFailover:
JDBC Fast i o
ic -tran
Connection Overview
b r
Fa Application
Oracle
non
Server 10g integrates JDBC ICC with the ONS API by having application
developers enable Fast Connection Failover (FCF). FCF works in conjunction with the JDBC
ICC to quickly and automatically recover lost or damaged connections. This automatic
connection management results from FAN events received by the local ONS daemon, or by a
remote ONS if a local one is not used, and handled by a special event handler thread. Both
JDBC thin and JDBC OCI drivers are supported.
Therefore, if JDBC ICC and FCF are enabled, your Java program automatically becomes an
ONS subscriber without having to manage FAN events directly.
Whenever a service or node down event is received by the mid-tier ONS, the event handler
automatically marks the corresponding connections as down and cleans them up. This prevents
applications that request connections from the cache from receiving invalid or bad connections.
Whenever a service up event is received by the mid-tier ONS, the event handler recycles some
unused connections, and reconnects them using the event service name. The number of recycled
connections is automatically determined by the connection cache. Because the listeners perform
connection load balancing, this automatically rebalances the connections across the preferred
instances of the service without waiting for application connection requests or retries.
For more information, refer to the Oracle Database JDBC Developer’s Guide and Reference.
Note: Similarly,
Unauthorized ODP.NET
reproduction also allows prohibitedฺ
or distribution you to use FCF using AQ2009,
Copyright© for FAN notifications.
Oracle and/or its affiliatesฺ
Using Oracle Streams Advanced Queuing for FAN
• Use AQ to publish FAN to ODP.NET and OCI.

• Turn on FAN notification to alert queue.
•
exec DBMS_SERVICE.MODIFY_SERVICE (
service_name => 'SELF-SERVICE', aq_ha_notification => TRUE);
• View published FAN events:

SQL> select object_name,reason
2 from dba_outstanding_alerts;
s a
)h a
OBJECT_NAME REASON
m ฺbr
----------- ----------------------------------------
i ฺ c o deฺ
xwkE Database xwkE (domain ) up as of time
r e d Gui
2005-12-30 11:57:29.000000000 -05:00;
@ sic dent
reason code: user
e r de Stu
lav e this
JFSERV Composite service xwkE up as of time
v i
2006-01-02 05:27:46.000000000 -05:00;
reason code: user
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VStreams f r
eAdvanced
Using Oracle
i c i o a n s Queuing for FAN
r r
-t Clusters publish FAN events to a system alert queue in the database using
FabApplication
Real
n o n
Oracle Streams Advanced Queuing (AQ). ODP.NET and OCI client integration uses this method
to subscribe to FAN events.
To have FAN events for a service posted to that alert queue, the notification must be turned on
for the service using either the DBMS_SERVICE PL/SQL package as shown in the slide, or by
using the Enterprise Manager interface.
To view FAN events that are published, you can use the DBA_OUTSTANDING_ALERTS or
DBA_ALERT_HISTORY views. An example using DBA_OUTSTANDING_ALERTS is shown
in the slide.
JDBC/ODP.NET FCF Benefits
• Database connections are balanced across preferred

instances according to LBA.
• Database work requests are balanced across preferred
instances according to LBA.
• Database connections are anticipated.
• Database connection failures are immediately detected a
and stopped. has r)
ฺ b
m eฺ
ฺ c o
r e di Guid
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V FCFsBenefits
f er
i
JDBC/ODP.NETo n
b ric FCF,
Byaenabling - t r ayour
non
existing Java applications connecting through Oracle JDBC and
Fapplication services, or your .NET applications using ODP.NET connection pools and
application services benefit from the following:
• All database connections are balanced across all RAC instances that support the new service
name, instead of having the first batch of sessions routed to the first RAC instance. This is
done according to the Load Balancing Advisory algorithm you use (see the next slide).
Connection pools are rebalanced upon service, instance, or node up events.
• The connection cache immediately starts placing connections to a particular RAC instance
when a new service is started on that instance.
• The connection cache immediately shuts down stale connections to RAC instances where
the service is stopped on that instance, or whose node goes down.
• Your application automatically becomes a FAN subscriber without having to manage FAN
events directly by just setting up flags in your connection descriptors.
• An exception is immediately thrown as soon as the service status becomes
not_restarting, which avoids wasteful service connection retries.
Note: For more information about how to subscribe to FAN events, refer to the Oracle Database
JDBC Developer’s Guide and Oracle Data Provider for .NET Developer’s Guide.
Load Balancing Advisory
• The Load Balancing Advisory (LBA) is an advisory for

sending work across RAC instances.
• The LBA advice is available to all applications that
send work:
– JDBC and ODP connection pools
– Connection load balancing
• The LBA advice sends work to where services are s a
h a
executing well and resources are available: ฺbr) m
– Relies on service goodness i ฺ c o deฺ
r e d Gui
– Adjusts distribution for different power
@ sicnodes,
d e n t different
priority and shape workloads, changing
d e Studemand
e r
lav hung,
– Stops sending work to slow,
_ thisor failed nodes
vi se
o
ici to u
b r
e (fa ense
rd lCopyright
l a V ab
i
V Advisoryf er
i o
Load Balancing
ic -tran s
a b r
Well-written
F n o napplications use persistent connections that span the instances of RAC offering a
service. Connections are created infrequently and exist for a long duration. Work comes into the
system with high frequency, borrows these connections, and exists for a relatively short
duration.
The Load Balancing Advisory has the task of advising the direction of incoming work to the
RAC instances that provide optimal quality of service for that work. The LBA algorithm uses
metrics sensitive to the current performance of services across the system.
The Load Balancing Advisory is deployed with Oracle’s key clients, such as Connection Load
Balancing, JDBC Implicit Connection Cache, OCI Session Pool, Oracle Data Provider (ODP)
Connection Pool for .NET, and is open for third-party subscription via ONS.
Using the Load Balancing Advisory for load balancing recognizes machine power differences,
sessions that are blocked in wait, failures that block processing, as well as competing services of
different importance. Using the Load Balancing Advisory prevents sending work to nodes that
are overworked, hung, or failed.
JDBC/ODP.NET Runtime Connection
Load Balancing: Overview
CRM work requests
Connection Cache
10%
?
s a
60%
30%
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
RAC RAC erd
e Stu RAC
CRM is CRM is vi l a v t h i s
CRM is
Inst1 very busy. _ Inst2 se Inst3
i
not busy.o u busy.
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Runtime f er Connection Load Balancing: Overview
i
JDBC/ODP.NET o
ic -tran s
b r
Fa nusing
Without
on the Load Balancing Advisory, work requests to RAC instances are assigned on a
random basis, which is suitable when each instance is performing equally well. However, if one
of the instances becomes more burdened than the others because of the amount of work resulting
from each connection assignment, the random model does not perform optimally.
The Runtime Connection Load Balancing feature provides assignment of connections based on
feedback from the instances in the RAC cluster. The Connection Cache assigns connections to
clients on the basis of a relative number indicating what percentage of work requests each
instance should handle.
In the diagram in the slide, the feedback indicates that the CRM service on Inst1 is so busy that
it should service only 10% of the CRM work requests; Inst2 is so lightly loaded that it should
service 60%; and Inst3 is somewhere in the middle, servicing 30% of requests. Note that these
percentages apply to, and the decision is made on, a per service basis. In this example, CRM is
the service in question.
Connection Load Balancing in RAC
• Connection load balancing allows listeners to

distribute connection requests to the best instances.
• Three metrics are available for listeners to decide:
– Session count by instance
– Run queue length of the node
– Service goodness
s a
• The metric used depends on the connection load- ) h a
r
balancing goal defined for the service: mฺb c o deฺ
– LONG i ฺ
d Gui
r e
– NONE
@ sic dent
– SHORT e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Balancing
f er in RAC
Connection i o Load
ic -tran s
a b r
Balancing
F n o n
connection requests is referred to as connection load balancing. Connections are
routed to different instances in the cluster on the basis of load information available to the
listener. Three metrics are available for the listeners to use when selecting the best instance, as
follows:
• Session count by instance: For services that span RAC instances uniformly and similar
capacity nodes, the session count evenly distributes the sessions across RAC. This method
is used when the service’s connection load-balancing goal is set to LONG.
• Run queue length of the node: For services that use a subset of RAC instances and
different capacity nodes, the run queue length places more sessions on the node with least
load at the time of connection creation.
• Goodness by service: The goodness of the service is a ranking of the quality of service that
the service is experiencing at an instance. It also considers states such as restricted access to
an instance. This method is used when the service’s connection load-balancing goal is set to
SHORT. To prevent a listener from routing all connections to the same instance between
updates to the goodness values, each listener adjusts its local ratings by a delta as
connections are distributed. The delta value used represents an average of resource time that
connections consume by using a service. To further reduce login storms, the listener uses a
threshold delta when the computed delta is too low because no work was sent over the
connections yet.
Load Balancing Advisory: Summary
• Uses DBMS_SERVICE.GOAL
– Service time: weighted moving average of elapsed time
– Throughput: weighted moving average of throughput
• AWR
– Calculates goodness locally (MMNL), forwards to master MMON
– Master MMON builds advisory for distribution of work across
RAC, and posts load balancing advice to AQ
s a
– IMON retrieves advice and send it to ONS h a
ฺ b r)
– EMON retrieves advice and send it to OCI
c o m eฺ
ฺ
– Local MMNL post goodness to PMON edi uid r G
• sic ent
Listeners use DBMS_SERVICE.CLB_GOAL=SHORT
– Use goodness from PMON to distribute r d e@connections.
S t ud
• Load Balancing Advisory users i l a ve(inside t histhe pools)
v
o_ too send e
– Use percentages and r ic i
flags
t us work.
( f ab nse
e rde lCopyright
l a V ab e
i
V Advisory:
f er Summary
i o
Load Balancing
ic -tran s
a b r
You
F enable
n o n Load Balancing Advisory when setting the service’s goal to
the
DBMS_SERVICE.GOAL_SERVICE_TIME or to DBMS_SERVIVCE.GOAL_THROUGHPUT.
MMNL (Manageability MoNitor Light) calculates the service metrics for service goal and
resource consumption every five seconds. MMNL derives the service goodness from these data.
MMON computes and posts the LBA FAN event to a system queue, and MMNL forwards the
service goodness and delta to PMON.
IMON (Instance Monitor) and EMON (Event MONitor) retrieve the event from the queue, and
PMON forwards the goodness and delta values to the listeners.
IMON posts the LBA FAN event to the local ONS daemon, and EMON posts it to AQ
subscribers.
The server ONS sends the event to the mid-tier ONS (if used).
The mid-tier receives the event and forwards them to subscribers. Each connection pool
subscribes to receive events for its own services. On receipt of each event, the Connection Pool
Manager refreshes the ratio of work to forward to each RAC instance connection part of the
pool. It also ranks the instances to use when aging out connections.
Work requests are routed to RAC instances according to the ratios calculated previously.
Monitor LBA FAN Events
SQL> SELECT TO_CHAR(enq_time, 'HH:MI:SS') Enq_time, user_data

2 FROM sys.sys$service_metrics_tab
3 ORDER BY 1 ;
ENQ_TIME USER_DATA
-------- -----------------------------------------------------
…
04:19:46 SYS$RLBTYP('JFSERV', 'VERSION=1.0 database=xwkE
s a
service=JFSERV { {instance=xwkE2 percent=50
)h a
ฺbr
flag=UNKNOWN}{instance=xwkE1 percent=50 flag=UNKNOWN}
} timestamp=2006-01-02 06:19:46')
c m
o deฺ
04:20:16 SYS$RLBTYP('JFSERV', 'VERSION=1.0 database=xwkE i ฺ
d Gui
r e
sic dent
service=JFSERV { {instance=xwkE2 percent=80
flag=UNKNOWN}{instance=xwkE1 percent=20 flag=UNKNOWN}
@
de Stu
} timestamp=2006-01-02 06:20:16')
e r
SQL>
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
Vi fer
Monitor LBAi o FAN n
ic -tra s
Events
b r
Fa nuse
You can
onthe SQL query shown in the slide to monitor the Load Balancing Advisory FAN
events for each of your services.
FAN Release Map
Oracle FAN Product Event FAN Event Received and Used

Release Integration System
10.1.0.2 JDBC ICC&FCF ONS Up/down/Load Balancing Advisory
Server-side callouts RAC Up/down
10.1.0.3 CMAN ONS Down

Listeners PMON Up/down/LBA
10.2 OCI connection pool AQ Down

s a
OCI session pool AQ Down
)h a
TAF AQ Down
m ฺbr
ODP.NET AQ Up/down/LBA i ฺ c o deฺ
r e d Gui
Custom OCI
DG Broker (10.2.0.3)
AQ
AQ
All
Down @ sic dent
e r de Stu
lav e this
11.1 OCI session pool AQ LBA
v i
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
Vi er
FAN Releasei o Map nsf
b r ic -tra
Fa nomap
The release n for FAN is shown in the slide.
Note: LBA stands for Load Balancing Advisory.
Transparent Application Failover: Overview
TAF Basic TAF Preconnect
2 Application 2 Application
OCI Library 6 OCI Library 6
4 Net Services 4 Net Services
3 7 3
5 5
s a
)h a
8
m ฺ br7
AP 3 ERP 3
i ฺ c o deฺ
7 r e d 3 Gui
1 1
@ sic dent
AP e r
ERPde ERP_PRECONNECT
S tu
v i lav e this
ic i o_ o us 1
r
ab nse t
( f
e rde lCopyright
l a V ab e
i
V sfer Failover (TAF): Overview
Transparenti o Application
ic -tran
b r
Fa anrun-time
TAF is on feature of the OCI driver. It enables your application to automatically
reconnect to the service if the initial connection fails. During the reconnection, although your
active transactions are rolled back, TAF can optionally resume the execution of a SELECT
statement that was in progress. TAF supports two failover methods:
• With the BASIC method, the reconnection is established at failover time. After the service
has been started on the nodes (1), the initial connection (2) is made. The listener establishes
the connection (3), and your application accesses the database (4) until the connection fails
(5) for any reason. Your application then receives an error the next time it tries to access the
database (6). Then, the OCI driver reconnects to the same service (7), and the next time your
application tries to access the database, it transparently uses the newly created connection
(8). TAF can be enabled to receive FAN events for faster down events detection and
failover.
• The PRECONNECT method is similar to the BASIC method except that it is during the
initial connection that a shadow connection is also created to anticipate the failover. TAF
guarantees that the shadow connection is always created on the available instances of your
service by using an automatically created and maintained shadow service.
Note: Optionally, you can register TAF callbacks with the OCI layer. These callback functions
are automatically invoked at failover detection and allow you to have some control of the
failover process. For more information, refer to the Oracle Call Interface Programmer’s Guide.
TAF Basic Configuration Without FAN: Example
$ srvctl add service -d RACDB -s AP -r I1,I2 \

> -P BASIC
$ srvctl start service -d RACDB -s AP
AP =
(DESCRIPTION =(FAILOVER=ON)(LOAD_BALANCE=ON)
(ADDRESS=(PROTOCOL=TCP)(HOST=N1VIP)(PORT=1521))
s a
h a
(CONNECT_DATA =
ฺ b r)
(SERVICE_NAME = AP)
c o m eฺ
(FAILOVER_MODE = ฺ
di Guid
r e
ic ent
(TYPE=SESSION)
@ s
(METHOD=BASIC)
r d e Stud
(RETRIES=180)
i l a ve this
(DELAY=5)))) _v e
io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
V sfer Without FAN: Example
TAF Basicio Configuration
b r ic -tran
Fa nonTAF, it is recommended that you create and start a service that is used during
Before using
connections. By doing so, you benefit from the integration of TAF and services. When you want
to use BASIC TAF with a service, you should have the -P BASIC option when creating the
service. After the service is created, you simply start it on your database.
Then, your application needs to connect to the service by using a connection descriptor similar
to the one shown in the slide. The FAILOVER_MODE parameter must be included in the
CONNECT_DATA section of your connection descriptor:
• TYPE specifies the type of failover. The SESSION value means that only the user session
is reauthenticated on the server side, whereas open cursors in the OCI application need to be
reexecuted. The SELECT value means that not only the user session is reauthenticated on
the server side, but also the open cursors in the OCI can continue fetching. This implies that
the client-side logic maintains fetch-state of each open cursor. A SELECT statement is
reexecuted by using the same snapshot, discarding those rows already fetched, and
retrieving those rows that were not fetched initially. TAF verifies that the discarded rows
are those that were returned initially, or it returns an error message.
• METHOD=BASIC is used to reconnect at failover time.
• RETRIES specifies the number of times to attempt to connect after a failover.
• DELAY specifies the amount of time in seconds to wait between connect attempts.
Note: If using TAF, do not set the GLOBAL_DBNAME parameter in your listener.ora file.
TAF Basic Configuration with FAN: Example
$ srvctl add service -d RACDB -s AP -r I1,I2
$ srvctl start service -d RACDB -s AP
execute dbms_service.modify_service ( ,-
service_name => 'AP' ,-
aq_ha_notifications => true ,-
failover_method => dbms_service.failover_method_basic ,-
failover_type => dbms_service.failover_type_session ,- s a
)h a
ฺbr
failover_retries => 180, failover_delay => 5 ,-
clb_goal => dbms_service.clb_goal_long);
c m
o deฺ
i ฺ
d Gui
r e
sic dent
AP =
@
de Stu
e r
v i la=v AP)))
e t his
(CONNECT_DATA = (SERVICE_NAME
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer with FAN: Example
TAF Basicio Configuration
b r ic -tran
F a
Oracle n 10g Release 2 supports server-side TAF with FAN. To use server-side TAF,
Database
o
n
create and start your service using SRVCTL, then configure TAF in the RDBMS by using the
DBMS_SERVICE package as shown in the slide. When done, make sure that you define a TNS
entry for it in your tnsnames.ora file. Note that this TNS name does not need to specify TAF
parameters as with the previous slide.
TAF Preconnect Configuration: Example
$ srvctl add service -d RACDB -s ERP -r I1 –a I2 \

> -P PRECONNECT
$ srvctl start service -d RACDB -s ERP
ERP =
(CONNECT_DATA = (SERVICE_NAME = ERP) s a
h a
(FAILOVER_MODE = (BACKUP=ERP_PRECONNECT)
ฺ b r)
m
(TYPE=SESSION)(METHOD=PRECONNECT))))
c o deฺ
i ฺ
d Gui
r e
ERP_PRECONNECT =
(DESCRIPTION =(FAILOVER=ON)(LOAD_BALANCE=ON) @ sic dent
(ADDRESS=(PROTOCOL=TCP)(HOST=N1VIP)(PORT=1521)) e r de Stu
v i
(ADDRESS=(PROTOCOL=TCP)(HOST=N2VIP)(PORT=1521))lav e this
(CONNECT_DATA = (SERVICE_NAME
ic i o_ o us = ERP_PRECONNECT)))
( f abr nse t
e rde lCopyright
l a V ab e
i er
V Configuration:
f
i c i
TAF Preconnect o a n s Example
In a
r
b to use t r
-PRECONNECT
F order n o n TAF, it is recommended that you create a service with preferred
and available instances. Also, in order for the shadow service to be created and managed
automatically by Oracle Clusterware, you must define the service with the –P PRECONNECT
option. The shadow service is always named using the format
<service_name>_PRECONNECT.
Like with the BASIC method without FAN, you need to use a special connection descriptor to
use the PRECONNECT method while connecting to the service. One such connection descriptor
is shown in the slide.
The main differences with the previous example are that METHOD is set to PRECONNECT and
an addition parameter is added. This parameter is called BACKUP and must be set to another
entry in your tnsnames.ora file that points to the shadow service.
Note: In all cases where TAF cannot use the PRECONNECT method, TAF falls back to the
BASIC method automatically.
TAF Verification
SELECT machine, failover_method, failover_type,
failed_over, service_name, COUNT(*)
FROM v$session
GROUP BY machine, failover_method, failover_type,
failed_over, service_name;
MACHINE FAILOVER_M FAILOVER_T FAI SERVICE_N COUNT(*)

1st ------- ---------- ---------- --- -------- --------
node node1 BASIC SESSION NO AP 1
node1 PRECONNECT SESSION NO ERP 1
s a
MACHINE FAILOVER_M FAILOVER_T FAI SERVICE_N COUNT(*) ) h a
2nd ฺ b r
node
------- ---------- ---------- --- --------- --------
c o m e ฺ
node2 NONE NONE NO ERP_PRECO i ฺ 1 i d
i c red t Gu
MACHINE FAILOVER_M FAILOVER_T FAI@ s den COUNT(*)
SERVICE_N
2nd
------- ---------- ---------- rd e --------
S tu
node v e ---
i s
--------
after
node2 BASIC SESSION
_ v ila eYES th AP 1
node2 PRECONNECT SESSION
r ic io o us YES ERP_PRECO 1
( f ab nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
TAF Verification
an TAF is correctly configured and that connections are associated with a
b ric whether
Toadetermine - t r
F non
failover option, you can examine the V$SESSION view. To obtain information about the
connected clients and their TAF status, examine the FAILOVER_TYPE, FAILOVER_METHOD,
FAILED_OVER, and SERVICE_NAME columns. The example includes one query that you
could execute to verify that you have correctly configured TAF.
This example is based on the previously configured AP and ERP services, and their
corresponding connection descriptors.
The first output in the slide is the result of the execution of the query on the first node after two
SQL*Plus sessions from the first node have connected to the AP and ERP services, respectively.
The output shows that the AP connection ended up on the first instance. Because of the load-
balancing algorithm, it can end up on the second instance. Alternatively, the ERP connection
must end up on the first instance because it is the only preferred one.
The second output is the result of the execution of the query on the second node before any
connection failure. Note that there is currently one unused connection established under the
ERP_PRECONNECT service that is automatically started on the ERP available instance.
The third output is the one corresponding to the execution of the query on the second node after
the failure of the first instance. A second connection has been created automatically for the AP
service connection,
Unauthorized reproductionand or
thedistribution
original ERPprohibitedฺ
connectionCopyright©
now uses the2009,
preconnected connection.
Oracle and/or its affiliatesฺ
FAN Connection Pools and TAF Considerations
• Both techniques are integrated with services and

provide service connection load balancing.
• Do not use FCF when working with TAF, and vice
versa.
• Connection pools that use FAN are always
preconnected.
a
• TAF may rely on operating system (OS) timeouts to has
detect failures. ฺ b r)
c o m eฺ
• FAN never relies on OS timeouts to detect i ฺ uid
edfailures. icr ent G
s
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Pools f r
eand
i c i
FAN Connectiono a n s TAF Considerations
r -t r
ab theonconnection
Because load balancing is a listener functionality, both FCF and TAF
F n
automatically benefit from connection load balancing for services.
When you use FCF, there is no need to use TAF. Moreover, FCF and TAF cannot work together.
For example, you do not need to preconnect if you use FAN in conjunction with connection
pools. The connection pool is always preconnected.
With both techniques, you automatically benefit from VIPs at connection time. This means that
your application does not rely on lengthy operating system connection timeouts at connect time,
or when issuing a SQL statement. However, when in the SQL stack, and the application is
blocked on a read/write call, the application needs to be integrated with FAN in order to receive
an interrupt if a node goes down. In a similar case, TAF may rely on OS timeouts to detect the
failure. This takes much more time to fail over the connection than when using FAN.
Summary

• Configure client-side connect-time load balancing
• Configure client-side connect-time failover
• Configure server-side connect-time load balancing
• Use the Load Balancing Advisory
• Describe the benefits of Fast Application Notification s a
• Configure server-side callouts r ) ha
m ฺb
• Configure the server- and client-side ONS iฺco ideฺ
i c
• Configure Transparent Application Failover red t Gu
s n
r d e@ Stude
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non

• Monitoring high availability of connections
• Creating and using callout scripts
• Using the Load Balancing Advisory
• Using the Transparent Application Failover feature
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Oracle Clusterware Administration
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Manually control the Oracle Clusterware stack
• Change voting disk configuration
• Back up or recover your voting disks
• Manually back up OCR
• Recover OCR s a
h a
• Replace an OCR mirror ฺ b r)
• Repair the OCR configuration c o m eฺ
ฺ
di Guid
r e
• Change VIP addresses
@ sic dent
• Use the CRS framework e r de Stu
• Prevent automatic instancev i lavrestarts
e t his
_ io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i er
Objectivesio V nsf
b r ic -tra
Fa nofothis
The goal n lesson is to make sure you understand the various administration tasks you can
operate at the Oracle Clusterware level. Although some important procedures are clearly
detailed in the lesson, the complete syntax for each command-line tool used is not systematically
explained. In this lesson you are using the following tools to administer Oracle Clusterware:
• crsctl
• crs_stat
• ocrconfig
• ocrcheck
• ocrdump
• srvctl
• oifcfg
• crs_profile, crs_register, crs_setperm, crs_start, crs_relocate, crs_stop, crs_unregister
For more information about the various options of the commands you can see in this lesson,
refer to the Oracle Database Oracle Clusterware and Oracle Real Application Clusters
Administration and Deployment Guide.
Oracle Clusterware: Overview
Portable cluster infrastructure that provides HA to RAC

databases and/or other applications:
• Monitors applications’ health
• Restarts applications on failure
• Can fail over applications on node failure
Protected App A
s a
)h a
ฺbr
RAC DB Inst RAC DB Inst
Listener Listener
c m
o deฺ
ORACLE_HOME ORACLE_HOME i ฺ
d Gui
Protected App B
r e
sic dent
Node 1 Node 2 Node 3
CRS HOME CRS HOME CRS HOME
@
de Stu
e r
v i lav e this
Oracle Clusterware
i _ files
osystem us
r ic t o
( f ab nse
e rde lCopyright
l a V ab e
i
V sOverviewf er
i o
Oracle Clusterware:
ic -tran
b r
Fa non
Oracle Clusterware is a portable cluster infrastructure that provides High Availability (HA) to
RAC databases and other applications. Oracle Clusterware makes applications highly available
by monitoring the health of the applications, by restarting applications on failure, by relocating
applications to another cluster node when the currently used node fails or when the application
can no longer run in the current node. In the case of node failure, certain type of protected
applications, such as a database instance, are not failed over surviving nodes.
Here, a cluster is a collection of two or more nodes where the nodes share a common pool of
storage used by the Oracle Clusterware system files (OCR and voting disk), a common network
interconnect, and a common operating system.
The graphic in the slide describes a possible three-node configuration where Node1 runs a RAC
database instance, a listener, and application A, all protected by Oracle Clusterware.
On Node2, only one RAC database instance and a listener are protected by Oracle Clusterware.
On Node3, one application B is protected by Oracle Clusterware.
Oracle Clusterware monitors all protected applications periodically, and based on the defined
failover policy, it can restart them either on the same node or relocate them to another node, or it
can decide to not restart them at all.
Note: Although Oracle Clusterware is a required component for using RAC, it does not require a
RAC license when used only to protect applications other than RAC databases.
Oracle Clusterware Run-Time View
init
oprocd evmd ocssd crsd

racgwrap
s a
+
)h a
evmlogger Voting
racgmain
m ฺbr
OCR iฺco d e ฺ
disk u i
red t racgimon
racgimon
i
s denc G racgimon
racgevtf @
de Stu
callout e r
callout
callout v i lav e this action
action
i o _ u s action
ic o
( f abr nse t
e rde lCopyright
l a V ab e
i
V sRun-Time
f er
i c i o
Oracle Clusterware
a n View
r t r
F b nothe
OnaUNIX, n-Oracle Clusterware stack is run from entries in /etc/inittab with respawn.
On Windows, it is run using the services controller. Here is a basic description of each process:
• Cluster Synchronization Services Daemon (OCSSD): This process runs in both vendor
clusterware and nonvendor clusterware environments. It integrates with existing vendor
clusterware, when present. OCSSD’s primary job is internode health monitoring, primarily
using the network interconnect as well as voting disks, and database/ASM instance endpoint
discovery via group services. OCSSD runs as user oracle, and failure exit causes
machine reboot to prevent data corruption in the event of a split brain.
• Process Monitor Daemon (OPROCD): This process is spawned in any nonvendor
clusterware environment. If OPROCD detects problems, it kills a node. It runs as root.
This daemon is used to detect hardware and driver freezes on the machine. If a machine was
frozen long enough for the other nodes to evict it from the cluster, it needs to kill itself to
prevent any I/O from being reissued to the disk after the rest of the cluster has remastered
locks.
Oracle Clusterware Run-Time View (continued)
• Cluster Ready Services Daemon (CRSD): This process is the engine for High Availability
operations. It manages Oracle Clusterware registered applications and starts, stops, check,
and fails them over via special action scripts. CRSD spawns dedicated processes called
RACGIMON that monitor the health of the database and ASM instances and host various
feature threads such as Fast Application Notification (FAN). One RACGIMON process is
spawned for each instance. CRSD maintains configuration profiles as well as resource
statuses in OCR (Oracle Cluster Registry). It runs as root and is restarted automatically on
failure. In addition, CRSD can spawn temporary children to execute particular actions such
as:
- racgeut (Execute Under Timer), to kill actions that do not complete after a certain
amount of time
- racgmdb (Manage Database), to start/stop/check instances
- racgchsn (Change Service Name), to add/delete/check service names for instances a
- racgons, to add/remove ONS configuration to OCR h as
- racgvip, to start/stop/check instance virtual IP ฺ b r )
• Event Management Daemon (EVMD): This process forwards cluster c o m
events e ฺ things
when
i ฺ i d
happen. It spawns a permanent child evmlogger that, on demand,
i c red spawns
t G u children such
as racgevtf to invoke callouts. It runs as oracle, andsis restarted e n automatically on
failure.
@
de Stu d
e r
Note: The RACG infrastructure is used to deploy v i e hisdatabase in highly available
lavthe Oracle
t
clustered environment. This infrastructure
ic i ois_ mainly
o usimplemented using the racgwrap script
that invokes the racgmain program. r
ab Itnissused t
e by CRS to execute actions for all node-centric
( f
resources as well as to proxy
e r deactionslicfor
e all instance-centric resources to RACGIMON.
Basically, this infrastructure
V e
is lresponsible for managing all ora.* resources.
V ila ferab
r i c io rans
t
Fab non-
Manually Control Oracle Clusterware Stack
Might be needed for planned outages:
# crsctl stop crs -wait
# crsctl start crs -wait
# crsctl disable crs

s a
)h a
# crsctl enable crs m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V Oracle f er Clusterware Stack
i o
Manually Control
ic -tran s
b r
Fa node
When a
non
of Oracle Clusterware comes up, the Oracle Clusterware processes start up
automatically. You can control this by using crsctl commands. You may have to manually
control the Oracle Clusterware stack while applying patches or during any planned outages. In
addition, these commands can be used by third-party clusteware when used in combination with
Oracle Clusterware.
You can stop the Oracle Clusterware stack by using the crsctl stop crs command.
You can also start the Oracle Clusterware stack by using the crsctl start crs command.
The –wait option displays progress and status for each daemon. Without this option, the
command returns immediately.
Use the crsctl disable crs command to disable Oracle Clusterware from being started
in a subsequent reboot. This command does not stop the currently running Oracle Clusterware
stack.
Use the crsctl enable crs command to enable Oracle Clusterware to be started in a
subsequent reboot.
Note: You must run these commands as root.
CRS Resources
• A resource is a CRS-managed application.

• Application profile attributes are stored in OCR:
– Check interval – Action script – Dependencies
– Failure policies – Privileges – …
• An action script must be able to:
– Start the application a
as
– Stop the application )h ฺbr
– Check the application
c m
o deฺ
i ฺ
d Gui
• Life cycle of a resource: r e
@ sic dent
crs_profile crs_register crs_start e tu
crs_stat
e r d S
v i lav e this
ic i o_ o uscrs_stop
crs_relocate crs_unregister
r
ab nse t
( f
e rde lCopyright
l a V ab e
i
V sfer
i
CRS Resourceso
ic -tran
b r
Fa the
CRS is
non
primary program for managing High Availability operations of applications within
the cluster. Applications that CRS manages are called resources. By default, CRS can manage
RAC resources such as database instance, ASM instances, listeners, instance VIPs, services,
ONS, and GSD. However, CRS is also able to manage other type of application processes and
application VIPs. CRS resources are managed according to their configuration parameters
(resource profile) stored in OCR and an action script stored anywhere you want. The resource
profile contains information such as the check interval, failure policies, the name of the action
script, privileges that CRS should use to manage the application, and resource dependencies.
The action script must be able to start, stop, and check the application.
CRS provides the following facilities to support the life cycle of a resource:
• crs_profile creates and edit a resource profile.
• crs_register adds the resource to the list of applications managed by CRS.
• crs_start starts the resource according to its profile. After a resource is started, its
application process is continuously monitored by CRS using a check action at regular
intervals. Also, when the application goes offline unexpectedly, it is restarted and/or failed
over to another node according to its resource profile.
• crs_stat informs you about the current status of a list of resources.
• crs_relocate moves the resource to another node of the cluster.
• crs_unregister removes the resource from the monitoring scope of CRS.
RAC Resources
$ <CRS HOME>/bin/crs_stat -t
Name Type Target State Host
----------------------------------------------------------------
ora.atlhp8.ASM1.asm application ONLINE ONLINE atlhp8
ora.atlhp8.LISTENER_ATLHP8.lsnr application ONLINE ONLINE atlhp8
ora.atlhp8.gsd application ONLINE ONLINE atlhp8
ora.atlhp8.ons application ONLINE ONLINE atlhp8
ora.atlhp8.vip application ONLINE ONLINE atlhp8
ora.atlhp9.ASM2.asm application ONLINE ONLINE atlhp9
ora.atlhp9.LISTENER_ATLHP9.lsnr application ONLINE ONLINE atlhp9
s a
ora.atlhp9.gsd application ONLINE ONLINE atlhp9
)h a
ora.atlhp9.ons application ONLINE ONLINE atlhp9
m ฺbr
ora.atlhp9.vip application ONLINE ONLINE atlhp9
i ฺ c o deฺ
ora.xwkE.JF1.cs application ONLINE ONLINE atlhp8
r e d Gui
sic dent
ora.xwkE.JF1.xwkE1.srv application ONLINE ONLINE atlhp8
ora.xwkE.JF1.xwkE2.srv @
application ONLINE ONLINE atlhp9
de Stu
ora.xwkE.db r
e
ora.xwkE.xwkE1.inst
lav e this
v i
ora.xwkE.xwkE2.inst
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i
RAC Resourceso
ic -tra–tn command shows you all the resources that are currently under Oracle
b r
Fa non
The crs_stat
Clusterware control. In the example shown in the slide, only resources starting with the prefix
ora. exist. These are the resources that implement RAC high availability in a clustered
environment.
You can see that, by default, Oracle Clusterware can control databases, database and ASM
instances, VIP/ONS/GSD/Listener (also called nodeapps), services, and service members.
In the slide, the Target status for the resources is ONLINE, which means that at next node
restart, Oracle Clusterware will try to start them up automatically.
State shows you the current status of the resource.
Target can be ONLINE or OFFLINE.
State can be ONLINE, OFFLINE, or UNKNOWN. UNKNOWN results from a failed start/stop
action, and can be reset only by a crs_stop -f resourceName command. The
combination of Target and State can be used to derive whether a resource is starting or
stopping.
Host shows you the name of the host on which the resource is managed.
Note: Using the crs_stat –t command truncates the resource names for formatting reasons.
The output example reestablishes entire names for clarity purposes.
Resource Attributes: Example
$ <CRS HOME>/bin/crs_stat -p ora.JFDB.JFDB1.inst

NAME=ora.JFDB.JFDB1.inst
TYPE=application
ACTION_SCRIPT=/u01/app/oracle/product/11g/bin/racgwrap
ACTIVE_PLACEMENT=0
AUTO_START=1
CHECK_INTERVAL=600
DESCRIPTION=CRS application for Instance
FAILOVER_DELAY=0
FAILURE_INTERVAL=0
FAILURE_THRESHOLD=0
s a
HOSTING_MEMBERS=atlhp8
)h a
PLACEMENT=restricted
m ฺbr
REQUIRED_RESOURCES=ora.atlhp8.ASM1.asm
i ฺ c o deฺ
RESTART_ATTEMPTS=5
r e d Gui
…
@ sic dent
$ <CRS HOME>/bin/crs_stat –t ora.xwkE.xwkE1.inst
e r de Stu
lav e this
Name Type Target State Host
v i
-----------------------------------------------------
ora....E1.inst application ONLINE
ic i o_ o us
ONLINE atlhp8
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfExample er
o
Resource iAttributes:
an
icuse the-trcrs_stat
b r
Fa non
You can –p resource_name command to print the OCR contents for
the named resource. The example in the slide shows you what you get for a RAC database
instance. Not all attributes are mandatory for each resource. Here is a brief description of the
most important attributes shown on the output above:
• NAME is the name of the application resource.
• TYPE must be APPLICATION for all CRS resources.
• ACTION_SCRIPT is the name and location of the action script used by CRS to start,
check, and stop the application. The default path is <CRS HOME>/crs/script.
• ACTIVE_PLACEMENT defaults to 0. When set to 1, Oracle Clusterware reevaluates the
placement of a resource during addition or restart of a cluster node.
• AUTO_START is a flag indicating whether Oracle Clusterware should automatically start a
resource after a cluster restart, regardless of whether the resource was running before the
cluster restart. When set to 0, Oracle Clusterware starts the resource only if it was running
before the restart. When set to 1, Oracle Clusterware always starts the resource after a
restart. When set to 2, Oracle Clusterware never restarts the resource (regardless of the
resource’s state when the node stopped).
• CHECK_INTERVAL is the time interval, in seconds, between repeated executions of the
check command for the application.
Resource Attributes: Example (continued)
• DESCRIPTION is a description of the resource.
• FAILOVER_DELAY is the amount of time, in seconds, that Oracle Clusterware waits
before attempting to restart or fail over a resource.
• FAILURE_INTERVAL is the interval, in seconds, during which Oracle Clusterware applies
the failure threshold. If the value is zero (0), then tracking of failures is disabled.
• FAILURE_THRESHOLD is the number of failures detected within a specified
FAILURE_INTERVAL before Oracle Clusterware marks the resource as unavailable and
no longer monitors it. If a resource’s check script fails this several times, then the resource
is stopped and set offline. If the value is zero (0), then tracking of failures is disabled. The
maximum value is 20.
• HOSTING_MEMBERS is an ordered list of cluster nodes separated by blank spaces that can
host the resource. Run the olsnodes commands to see your node names.
• PLACEMENT defines the placement policy (balanced, favored, or restricted)a that
specifies how Oracle Clusterware chooses the cluster node on which to start the resource: ha s
- balanced: Oracle Clusterware favors starting or restarting the application r )
ฺb on the
node that is currently running the fewest resources. The host with c m
othe feweste ฺresources
i ฺ i d
running is chosen. If no node is favored by these criteria, rthen
i c ed anyt available
G u node is
chosen.
@ s den
- favored: Oracle Clusterware refers to the rlist
e deof nodes
S tuin the HOSTING_MEMBERS
attribute of the application profile. Only
v i lavcluster
e t his that are in this list and that
nodes
satisfy the resource requirements i _ eligible
oare u s for placement consideration. The order of
the hosting nodes determines ic
br which o
t runs the application. If none of the nodes in
node
( f a s e
the hosting node list
r d l i en then Oracle Clusterware places the application on
e are available,
c
any availableV enode. bThis
l e node may or may not be included in the
a
il fera list.
HOSTING_MEMBERS
V
i o
- criestricted:
a n s Similar to the favored policy, except that if none of the nodes on the
r r
Fab nhosting
o n-t list are available, then Oracle Clusterware does not start or restart the
application. A restricted placement policy ensures that the application never runs
on a node that is not on the list, even if you manually relocate it to that node.
• REQUIRED_RESOURCES is an ordered list of resource names separated by blank spaces
that this resource depends on. Oracle Clusterware relocates or stops an application if a
required resource becomes unavailable. Therefore, in the example on the previous page, it is
clear that to start the JFDB1 instance, the ASM instance ASM1 must be started first.
• RESTART_ATTEMPTS is the number of times that Oracle Clusterware only attempts to
restart a resource on a single cluster node before attempting to relocate the resource. After
the time period that you have indicated by the setting for UPTIME_THRESHOLD has
elapsed, Oracle Clusterware resets the value for the restart counter (RESTART_COUNTS) to
0. Basically, RESTART_COUNTS cannot exceed RESTART_ATTEMPTS for the
UPTIME_THRESHOLD period.
The crs_stat –t resource_name command shows you the named resource’s states. In
the slide, the Target status for the resource is ONLINE meaning that at the next node restart,
Oracle Clusterware will try to start up the instance. State shows you the current status of the
instance.
Note: The output shown in the slide is truncated for formatting reasons.
Main Voting Disk Function
Node1 Node2 Node3

CSS CSS CSS
We all see 1&2&3 Nodes can

We all see We all see
see each other.
1&2&3 1&2&3
Voting disk
Split-brain s a
h a
)Node3
ฺbr
Node1 Node2
Node3 can no longer CSS CSS
c m
o deฺ CSS
communicate through i ฺ i
1&I see 1 2 red u
private interconnect. I do not see 3 &
G
sic dent I’d evicted!
I’ve been
Others no longer see I see 1 2 &
its heartbeats and 2 see 1 2 &

d @
e Stu I see 3 better stop
evict that node by
=> r
ve this
We should
evict 3! _v
i l a
using the voting disk.
i o u s eVoting disk
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V Functionf er
Main Votingi o Disk
ic -tran s
b r
Fa noservice
CSS is the n that determines which nodes in the cluster are available, and provides cluster
group membership and simple locking services to the other processes. CSS typically determines
node availability via communication through a dedicated private network with a voting disk
used as a secondary communication mechanism. Basically, this is done by sending heartbeat
messages through the network and the voting disk as illustrated by the top graphic in the slide.
The voting disk is a shared raw disk partition or file on a clustered file system that is accessible
to all nodes in the cluster. Its primary purpose is to help in situations where the private network
communication fails. When that happens, the cluster is unable to have all nodes remain available
because they are no longer able to synchronize I/O to the shared disks. Therefore, some of the
nodes must go offline. The voting disk is then used to communicate the node state information
used to determine which nodes go offline. Without the voting disk, it can become impossible for
an isolated node(s), to determine whether it is experiencing a network failure or whether the
other nodes are no longer available. It would then be possible for the cluster to get into a state
where multiple subclusters of nodes would have unsynchronized access to the same database
files. This situation is commonly referred to as the cluster split-brain problem.
The graphic at the bottom of the slide illustrates what happens when node3 can no longer send
heartbeats to other members of the cluster. When others can no longer see node3’s heartbeats,
they decide to evict that node by using the voting disk. When node3 reads the removal message,
it generally
Unauthorized reboots itselfortodistribution
reproduction make sure all outstanding
prohibitedฺ write I/Os are
Copyright© lost.Oracle and/or its affiliatesฺ
2009,
Main Voting Disk Function (continued)
Note: In addition to the voting disk mechanism, a similar mechanism also exists for RAC
database instances. At the instance level, the control file is used by all participating instances for
voting. This is necessary because there can be cases where instances should be evicted, even if
network connectivity between nodes is still in good shape.
For example, if LMON or LMD is stuck on one instance, it could then be possible to end up with
a frozen cluster database. Therefore, instead of allowing a clusterwide hang to occur, RAC
evicts the problematic instance(s) from the cluster.
When the problem is detected, the instances race to get a lock on the control file. The instance
that obtains the lock tallies the votes of the instances to decide membership. This is called
Instance Membership Reconfiguration (IMR).
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Important CSS Parameters
• MISSCOUNT:
– Represents network heartbeat timeouts
– Determines disk I/O timeouts during reconfiguration
– Defaults to 30 seconds
– Should not be changed
• DISKTIMEOUT:
– Represents disk I/O timeouts outside reconfiguration
– Defaults to 200 seconds
– Can be temporarily changed when experiencing very long I/Ohas
a
latencies to voting disks: ฺbr) m eฺ
1. Shut down Oracle Clusterware on all nodes but one.
ฺ c o
2. As root on available node, use: crsctl set css disktimeout r e di Guid M+1
3. Reboot available node.
@ sic dent
4. Restart all other nodes.
e r de Stu
• Can be changed ONLY under i v thguidance
laexplicit is from Oracle
o _ v s e
Support
b r ici to u
e (fa ense
rd lCopyright
l a V ab
i
V Parameters
f er
o
Important iCSS
icmisscount s
an parameter represents the maximum time, in seconds, that a network
b r - t r
Fa
The CSS
on
heartbeatnacross the interconnect can be missed before entering into a cluster reconfiguration for
node eviction purposes. The default value for the misscount parameter value is 30 seconds.
The misscount parameter’s value drives cluster membership reconfigurations and directly
effects the availability of the cluster. Its default settings should be acceptable. Modifying this
value not only can influence the timeout interval for the I/O to the voting disk, but also
influences the tolerance for missed network heartbeats across the interconnect. This directly
affects database and cluster availability. The CSS misscount default value, when using
vendor (non-Oracle) clusterware, is also 30 seconds, and you should not change the default
misscount value if you are using vendor clusterware.
The CSS disktimeout parameter represents the maximum time, in seconds, that a disk
heartbeat can be missed (outside cluster reconfiguration events) before entering into a cluster
reconfiguration for node eviction purposes. Its default value is 200 seconds. However, if I/O
latencies to the voting disk are greater than the default internal I/O timeout, the cluster may
experience CSS node evictions. The most common cause in these latencies relate to multipath
I/O software drivers and the reconfiguration times resulting from a failure in the I/O path.
Therefore, until the underlying storage I/O latency is resolved, disktimeout could be
temporarily modified based only on “maximum I/O latency to the voting disk” including
latencies resulting from I/O path reconfiguration plus one second (M+1).
Multiplexing Voting Disks
• Voting disk is a vital resource for your cluster

availability.
• Use one voting disk if it is stored on a reliable disk.
• Otherwise, use multiplexed voting disks:
– There is no need to rely on multipathing solutions.
– Multiplexed copies should be stored on independent a
devices. h as
ฺbr)
– Make sure that there is no I/O starvation for yourmvoting
disks devices. i ฺ c o deฺ
r e d Gui
sic
– Use at least three multiplexed copies.
@ nt de
• CSS uses a simple majority rule r e decide
dto S tu whether
e
v this
vila
voting disk reads are consistent.
o _ use
i
ric e to
a b
(f ens
d e
r lCopyright
e
V ab e
i l a
V sDisks er
i
Multiplexing o Voting f
b r ic -tran
F a
CSS n can be improved by configuring it with multiple voting disks. Using only one
availability
o
n
voting disk is adequate for clusters that are configured to use a single, highly available shared
disk, where both the database files and the CSS voting disk reside. However, it is desirable to
use multiple copies of the voting disk when using less reliable storage. Also, you can use
multiple voting disks so that you do not have to rely on a multipathing solution.
The way voting disk multiplexing is implemented forces you to have at least three voting disks.
To avoid a single point of failure, your multiplexed voting disk should be located on physically
independent storage devices with a predictable load well below saturation.
When using multiplexed copies of the voting disk, CSS multiplexes voting data to all the voting
disks. When CSS needs to read the voting disk, it reads all the information from all the voting
disks. If strictly more than half of the voting disks are up and contain consistent information,
CSS can use that consistent data in the same way as a single voting disk configuration. If less
than half of the voting disks have readable consistent data, CSS will need to self-terminate like
in the situation where a single voting disk cannot be read by CSS. This self-termination is to
prevent disjoint subclusters from forming. You can have up to 32 voting disks, but use the
following formula to determine the number of voting disks you should use: v = f*2+1, where
v is the number of voting disks, and f is the number of disk failures you want to survive.
Note: A typical voting disk configuration comprises between three and five disks.
Change Voting Disk Configuration
• Voting disk configuration can be changed dynamically.

• To add a new voting disk:
# crsctl add css votedisk <new voting disk path>
• To remove a voting disk:

s a
# crsctl delete css votedisk <old voting disk path>
)h a
m ฺbr
• If Oracle Clusterware is down on all nodes,iฺuse c o thedeฺ
–force option: r e d Gui
@ sic dent
# crsctl add css votedisk <new voting
e r de diskStupath> -force
v i lav e this
# crsctl delete css votedisk
ic i o_ <old o u svoting disk path> -force
( f abr nse t
e rde lCopyright
l a V ab e
i er
V DisksConfiguration
f
i
Change Votingo
ic -tran
b r
Fa Oracle
During
non
Clusterware installation, you can multiplex your voting disk by using the Specify
Voting Disk Location screen of the Oracle Universal Installer. This screen allows you to specify
three voting disk locations. However, you can dynamically add and remove voting disks after
installing Oracle Clusterware by using the following commands as the root user:
• To add a voting disk: crsctl add css votedisk path
• To remove a voting disk: crsctl delete css votedisk path
Here, path is the fully qualified path.
If your cluster is down, you can use the -force option (at the very end of the crsctl
command) to modify the voting disk configuration with either of these commands without
interacting with active Oracle Clusterware daemons. However, using the -force option while
any cluster node is active may corrupt your configuration.
Back Up and Recover Your Voting Disks
• Should not be needed. Instead, you should add/remove.

• Recommendation is to use symbolic links.
• Back up one voting disk by using the dd command.
– After Oracle Clusterware installation
– After node addition or deletion
– Cannot be done online
$ crsctl query css votedisk
s a
)h a
$ dd if=<voting disk path> of=<backup path> bs=4k
m ฺbr
i ฺ c o deฺ
• Recover voting disks by restoring the c e
first
r ui the
doneGusing
dd command, and then multiplex it@
i ent
ifsnecessary.
r d e Stud
• If no voting disk backup islaavailable,
i ve this reinstall Oracle
Clusterware. i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
V i f r
eYour
Back Up and
i c i o Recover
a n s Voting Disks
r r
-t no need to back up a voting disk. Simply add a new one and drop a bad one.
Fab should
There
n o nbe
It is recommended to use symbolic links to specify your voting disk paths. This is because the
voting disk paths are directly stored in OCR, and editing the OCR file directly is not supported.
By using symbolic links to your voting disks, it becomes easier to restore your voting disks if
their original locations can no longer be used as a restore location.
A new backup of one of your available voting disks should be taken any time a new node is
added, or an existing node is removed. The recommended way to do that is to use the dd
command (ocopy in Windows environments). As a general rule on most platforms, including
Linux and Sun, the block size for the dd command should be at least 4 KB to ensure that the
backup of the voting disk gets complete blocks.
Before backing up your voting disk with the dd command, make sure that you stopped Oracle
Clusterware on all nodes.
The crsctl query css votedisk command lists the voting disks currently used by
CSS. This can help you to determine which voting disk to backup.
The slide shows you the procedure you can follow to back up and restore your voting disk.
Note: If you lose all your voting disks and you do not have any backup, you must reinstall
Oracle Clusterware.
OCR Architecture
Node1 Node2 Node3
OCR cache OCR cache OCR cache
CRS CRS CRS

process process process
s a
)h a
Client Client
m ฺbr
process process
i ฺ c o deฺ
r e d Gui
Shared @ sic dent
storage OCR r d e SOCR tu
e
lav e this file
primary mirror
v i
file
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
OCR Architecture
ic -tran
b r
Fa configuration
Cluster
non
information is maintained in Oracle Cluster Registry (OCR). OCR relies
on a distributed shared-cache architecture for optimizing queries, and clusterwide atomic
updates against the cluster repository. Each node in the cluster maintains an in-memory copy of
OCR, along with the Cluster Ready Services Daemon (CRSD) that accesses its OCR cache.
Only one of the CRS processes actually reads from and writes to the OCR file on shared storage.
This process is responsible for refreshing its own local cache, as well as the OCR cache on other
nodes in the cluster. For queries against the cluster repository, the OCR clients communicate
directly with the local OCR process on the node from which they originate. When clients need to
update OCR, they communicate through their local CRS process to the CRS process that is
performing input/output (I/O) for writing to the repository on disk.
The main OCR client applications are the Oracle Universal Installer (OUI), SRVCTL, Enterprise
Manager (EM), the Database Configuration Assistant (DBCA), the Database Upgrade Assistant
(DBUA), NetCA, and the Virtual Internet Protocol Configuration Assistant (VIPCA).
Furthermore, OCR maintains dependency and status information for application resources
defined within Oracle Clusterware, specifically databases, instances, services, and node
applications.
OCR Architecture (continued)
The installation process for Oracle Clusterware gives you the option of automatically mirroring
OCR. This creates a second OCR file (the OCR mirror file) to duplicate the original OCR file
(the primary OCR file). You can put the OCR mirror file on a cluster file system or on a shared
raw device. Although it is recommended to mirror your OCR, you are not forced to do it during
installation.
The name of the OCR configuration file on UNIX-based system is ocr.loc, and the OCR file
location variables are ocrconfig_loc and ocrmirrorconfig_loc.
It is strongly recommended that you use mirrored OCR files if the underlying storage is not
RAID. This prevents OCR from becoming a single point of failure.
Note: OCR also serves as a configuration file in a single instance with the ASM, where there is
one OCR per node.
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
OCR Contents and Organization
root
css
CRS HOME
SYSTEM evm
crs
OCR
NODEAPPS
s a
LOG
)h a
DATABASE ASM
m ฺbr
DATABASES i ฺ c o deฺ
r e d Gui
ONS SERVICE
@ sic dent
CRS
e r de Stu
INSTANCE
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er
Vand Organization
f
i
OCR Contents o
ic -tran s
b r
Fa non technology requires a repository through which the clustering software and
Every clustering
other cluster-aware application processes can share information. Oracle Clusterware uses Oracle
Cluster Registry to store information about resources it manages. This information is stored in a
treelike structure using key–value pairs.
The slide shows you the main branches composing the OCR structure:
• The SYSTEM keys contain data related to the main Oracle Clusterware processes such as
CSSD, CRSD, and EVMD. For example, CSSD keys contain information about the
misscount parameter and voting disk paths.
• The DATABASE keys contain data related to the RAC databases that you registered with
Oracle Clusterware. As shown, you have information about instances, nodeapps, services,
and so on.
• The last category of keys that you can find in OCR relate to the resource profiles used by
Oracle Clusterware to maintain availability of the additional application you registered.
These resources include the additional application VIPs, the monitoring scripts, and the
check interval values.
Note: The XML data on the right side of the slide were obtained by using the ocrdump –xml
command.
Managing OCR Files and Locations: Overview
-export
-import
-repair ocrmirror -upgrade

-repair ocr -downgrade
ocrconfig
s a
-backuploc h a
-overwrite -showbackup ฺ b r)
o
-manualbackup
c m eฺ
di ฺ
-restore uid
s icre ent G
-replace ocrmirror
r d e@ Stud
e
ocrdump -replace
v i l a vocr
e t his ocrcheck
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i er Locations: Overview
V Filessfand
Managing iOCRo an tool (the main configuration tool for Oracle Cluster Registry) to:
icthe ocrconfig
b r - t r
F•a Generate
You use
non logical backups of OCR using the –export option, and use them later to restore
your OCR information using the –import option
• Upgrade or downgrade OCR
• Use the –showbackup option to view the generated backups (by default, OCR is backed
up on a regular basis). These backups are generated in a default location that you can
change using the –backuploc option. If need be, you can then restore physical copies of
your OCR using the –restore option. You can also manually create OCR backups using
the -manualbackup option.
• Use the –replace ocr or –replace ocrmirror options to add, remove, or replace
the primary OCR files or the OCR mirror file
• Use the –overwrite option under the guidance of Support Services because it allows you
to overwrite some OCR protection mechanisms when one or more nodes in your cluster
cannot start because of an OCR corruption
• Use the –repair option to change the parameters listing the OCR and OCR mirror
locations
The ocrcheck tool enables you to verify the OCR integrity of both OCR and its mirror. Use
the ocrdump utility to write the OCR contents (or part of it) to a text or XML file.
Automatic OCR Backups
• The OCR content is critical to Oracle Clusterware.

• OCR is automatically backed up physically:
– Every four hours: CRS keeps the last three copies.
– At the end of every day: CRS keeps the last two copies.
– At the end of every week: CRS keeps the last two copies.
$ cd $ORACLE_BASE/Crs/cdata/jfv_clus
$ ls -lt
-rw-r--r-- 1 root root 4784128 Jan 9 02:54 backup00.ocr
s a
-rw-r--r-- 1 root root 4784128 Jan 9 02:54 day_.ocr
)h a
m ฺbr
i ฺ c o deฺ
-rw-r--r-- 1 root root 4784128 Jan 8 02:54 day.ocr
r e d Gui
-rw-r--r-- 1 root root 4784128 Jan 6
-rw-r--r-- 1 root root 4005888 Dec 30
02:54
14:54 @ sic dent
week_.ocr
week.ocr
e r de Stu
• Change the default automatic
v i lavbackup e t hislocation:
# ocrconfig –backuploc
ic i o_/shared/bak
o us
r
ab nse t
( f
e rde lCopyright
l a V ab e
i
V Backupsf er
AutomaticioOCR
ic -tran s
b r
Fa non important cluster and database configuration information for RAC and Oracle
OCR contains
Clusterware. One of the Oracle Clusterware instances (CRSD master) in the cluster
automatically creates OCR backups every four hours, and CRS retains the last three copies. That
CRSD process also creates an OCR backup at the beginning of each day and of each week, and
retains the last two copies. This is illustrated in the slide where you can see the content of the
default backup directory of the CRSD master.
Although you cannot customize the backup frequencies or the number of retained copies, you
have the possibility to identify the name and location of the automatically retained copies by
using the ocrconfig -showbackup command.
The default target location of each automatically generated OCR backup file is the
<CRS Home>/cdata/<cluster name> directory. It is recommended to change this
location to one that is shared by all nodes in the cluster by using the
ocrconfig -backuploc <new location> command. This command takes one
argument that is the full path directory name of the new location.
Back Up OCR Manually
• Daily backups of your automatic OCR backups to a

different storage device:
– Use your favorite backup tool.
• On demand physical backups:
# ocrconfig –manualbackup
s a
• Logical backups of your OCR before and after making ) h a
ฺ b r
significant changes: m
i ฺ c o d e ฺ
# ocrconfig –export file name red t Gu i
i
s denc
@
debackups tu that match
• Make sure that you restore OCR e r S
your current system configuration. v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VManually f er
Back Up OCRi o
ic -tran s
b r
a ofothe
Because
F n n importance of OCR information, it is also recommended to manually create
copies of the automatically generated physical backups. You can use any backup software to
copy the automatically generated backup files, and it is recommended to do that at least once
daily to a different device from where the primary OCR resides.
You can performs an OCR backup on demand using the –manualbackup option. The backup
is generated in the location that you specify with the -backuploc option. .
In addition, you should also export the OCR contents before and after making significant
configuration changes such as adding or deleting nodes from your environment, modifying
Oracle Clusterware resources, or creating a database. Use the ocrconfig -export
command as the root user to generate OCR logical backups. You need to specify a file name as
the argument of the command, and it generates a binary file that you should not try to edit.
Most configuration changes that you make not only change the OCR contents but also cause file
and database object creation. Some of these changes are often not restored when you restore
OCR. Do not perform an OCR restore as a correction to revert to previous configurations if
some of these configuration changes fail. This may result in an OCR with contents that do not
match the state of the rest of your system.
Note: If you try to export OCR while an OCR client is running, you get an error.
Recover OCR Using Physical Backups
1. Locate a physical backup: $ ocrconfig –showbackup
2. Review its contents: # ocrdump –backupfile file_name
3. Stop Oracle Clusterware

# crsctl stop crs
on all nodes:
s a
4. Restore the physical OCR backup:
)h a
ฺbr
om ฺ
# ocrconfig –restore <CRS HOME>/cdata/jfv_clus/day.ocr
iฺc uide
e d
5. Restart Oracle Clusterware s icr ent G
on all nodes: r d e@ Stustart
# crsctl d crs
i l a ve this
6. Check OCR integrity: _v $ cluvfy e comp ocr -n all
o s
b r ici to u
e (fa ense
rd lCopyright
l a V ab
i er
VUsingsPhysical
f
Recover c
i i
OCRo a n Backups
b r t r
F1.athe
Use n- procedure to restore OCR on UNIX-based systems:
following
n o
Identify the OCR backups by using the ocrconfig -showbackup command. You can
execute this command from any node as user oracle. The output tells you on which node
and which path to retrieve both automatically and manually generated backups. Use the
auto or manual argument to display only one category.
2. Review the contents of the backup by using ocrdump -backupfile file_name,
where file_name is the name of the backup file.
3. Stop Oracle Clusterware on all the nodes of your cluster by executing the
crsctl stop crs command on all the nodes as the root user.
4. Perform the restore by applying an OCR backup file that you identified in step one using
the following command as the root user, where file_name is the name of the OCR file
that you want to restore. Make sure that the OCR devices that you specify in the OCR
configuration file (/etc/oracle/ocr.loc) exist and that these OCR devices are valid
before running this command: ocrconfig -restore file_name
5. Restart Oracle Clusterware on all the nodes in your cluster by restarting each node or by
running the crsctl start crs command as the root user.
6. Run the following command to verify OCR integrity, where the -n all argument
retrieves a listing of all the cluster nodes that are configured as part of your cluster:
cluvfy
Unauthorized comp ocr
reproduction -n all prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
or distribution
Recover OCR Using Logical Backups
1. Locate a logical backup created using an OCR export.
2. Stop Oracle Clusterware on all nodes:

# crsctl stop crs
3. Restore the logical OCR backup:

s a
# ocrconfig –import /shared/export/ocrback.dmp
)h a
m ฺbr
4. Restart Oracle Clusterware on all nodes: iฺco ideฺ
i c r ed t Gu
# crsctl start crs
@ s den
e r de Stu
5. Check OCR integrity: _$vicluvfy lav e tcomp his ocr -n all
r ic io o us
( f ab nse t
e rde lCopyright
l a V ab e
i er Backups
VUsingsLogical
f
Recover c
i i
OCRo an
b r - t r
Fa non procedure to import OCR on UNIX-based systems:
Use the following
1. Identify the OCR export file that you want to import by identifying the OCR export file
that you previously created using the ocrconfig -export file_name command.
2. Stop Oracle Clusterware on all the nodes in your RAC database by executing the
crsctl stop crs command on all the nodes as the root user.
3. Perform the import by applying an OCR export file that you identified in step one using the
following command, where file_name is the name of the OCR file from which you want
to import OCR information: ocrconfig -import file_name
4. Restart Oracle Clusterware on all the nodes in your cluster by restarting each node using
the crsctl start crs command as the root user.
5. Run the following Cluster Verification Utility (CVU) command to verify OCR integrity,
where the -n all argument retrieves a listing of all the cluster nodes that are configured as
part of your cluster: cluvfy comp ocr -n all
Replace an OCR Mirror: Example
# ocrcheck
Status of Oracle Cluster Registry is as follows:
Version : 2
Total space (kbytes) : 200692
Used space (kbytes) : 3752
Available space (kbytes) : 196940
ID : 495185602
s a
Device/File Name : /oradata/OCR1
)h a
Device/File integrity check succeeded
m ฺbr
Device/File Name : /oradata/OCR2
i ฺ c o deฺ
Device/File needs to be synchronized with the other device
r e d Gui
# ocrconfig –replace ocrmirror /oradata/OCR2 @ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
Vor Remove
f er an OCR File
i
Replace, Add,o
ic -tran s
b r
Fa nexample
The code
on in the slide shows you how to replace the existing OCR mirror file. It is
assumed that you already have an OCR mirror, and that this mirror is no longer working as
expected. Such a reorganization can be triggered because you received an OCR failure alert in
Enterprise Manager, or because you saw an alert directly in the Oracle Clusterware alert log file.
Using the ocrcheck command, you clearly see that the OCR mirror is no longer in sync with the
primary OCR. You then issue the ocrconfig –replace ocrmirror filename command to
replace the existing mirror with a copy of your primary OCR. In the example, filename can be a
new file name if you decide to also relocate your OCR mirror file.
If it is the primary OCR file that is failing, and if your OCR mirror is still in good health, you
can use the ocrconfig –replace ocr filename command instead.
Note: The example in the slide shows you a replace scenario. However, you can also use a
similar command to add or remove either the primary or the mirror OCR file:
• Executing ocrconfig –replace ocr|ocrmirror filename adds the primary or mirror
OCR file to your environment if it does not already exist.
• Executing ocrconfig –replace ocr|ocrmirror removes the primary or the mirror OCR
file.
Repair OCR Configuration: Example
1. Stop Oracle Clusterware on Node2:

# crsctl stop crs
2. Add OCR mirror from Node1:

# ocrconfig –replace ocrmirror /OCRMirror
s a
3. Repair OCR mirror location on Node2: )h a
m ฺ br
# ocrconfig –repair ocrmirror /OCRMirror
i ฺ c o deฺ
r e d Gui
@ sic dent
4. Start Oracle Clusterware on Node2: de Stu
e r
v i lav e this# crsctl start crs
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer Example
Repair OCR i o Configuration:
ic -tran
Use b r
a ocrconfig
the –repair command to repair inconsistent OCR configuration information.
F n o n
The OCR configuration information is stored in:
• /etc/oracle/ocr.loc on Linux and AIX
• /var/opt/oracle/ocr.loc on Solaris and HP-UX
• Registry key HKEY_LOCAL_MACHINE\SOFTWARE\Oracle\ocr on Windows
You may need to repair an OCR configuration on a particular node if your OCR configuration
changes while that node is stopped. For example, you may need to repair the OCR on a node that
was not up while you were adding, replacing, or removing an OCR.
The example in the slide illustrates the case where the OCR mirror file is added on the first node
of your cluster while the second node is not running Oracle Clusterware.
You cannot perform this operation on a node on which Oracle Clusterware is running.
Note: This repairs the OCR configuration information only; it does not repair OCR itself.
OCR Considerations
• If using raw devices to store OCR files, make sure they

exist before add or replace operations.
• You must be the root user to be able to add, replace,
or remove an OCR file while using ocrconfig.
• While adding or replacing an OCR file, its mirror needs
to be online.
a
• If you remove a primary OCR file, the mirror OCR file has
becomes primary. ฺ b r)
c o m eฺ
ฺ
• Never remove the last remaining OCR file.edi uid
icr ent G
s
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i er
V Considerations
f
ReplacingiOCRo
ic -tran s
b r
F•a If you
Here is a listnof important considerations when you use the ocrconfig –replace command:
noare using raw devices, make sure that the file name exists before issuing an add or
replace operation using ocrconfig.
• To be able to execute an add, replace, or remove operation using ocrconfig, you must be
logged in as the root user.
• The OCR file that you are replacing can be either online or offline.
• If you remove a primary OCR file, then the mirrored OCR file becomes the primary OCR
file.
• Do not perform an OCR removal operation unless there is at least one other active OCR file
online.
Change VIP Addresses
1. Determine the interface used to support your VIP:

$ ifconfig -a
2. Stop all resources depending on the VIP:

$ srvctl stop instance -d DB -i DB1
$ srvctl stop asm -n node1
s a
)h a
ฺbr
# srvctl stop nodeapps -n node1
o m ฺ
3. Verify that the VIP is no longer running:ediฺc uide
$ ifconfig -a [ + $ crs_stat s icr en] t G
r d e@ Stud
i l a ve this
4. Change IP in /etc/hosts i o _v and u s eDNS.
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer
Change VIP i oAddresses
ic -tran
b r
a address
The
F VIP
n o n is a static IP address with a virtual host name defined and resolved through
either the DNS or your hosts file. During Oracle Clusterware installation, you are prompted to
enter a Virtual IP and virtual host name for each of the nodes in the cluster. These are stored in
OCR, and different components within the Oracle Clusterware HA framework depend on these
VIPs. If, for some reasons, you want to change the VIP address, use the following procedure on
each node, one at a time:
1. Confirm the current IP address for the VIP by running the ifconfig –a command. On
Windows, run the ipconfig /all command. This should show you the current VIP bound
to one of the network interfaces.
2. Stop all resources that are dependent on the VIP on that node: First, stop the database
instance, and then the ASM instance. When done, stop nodeapps.
3. Verify that the VIP is no longer running by executing the ifconfig -a command again,
and confirm that its interface is no longer listed in the output. If the interface still shows as
online, this is an indication that a resource which is dependent on the VIP is still running.
The crs_stat -t command can help to show resources that are still online.
4. Make any changes necessary to all nodes’ /etc/hosts files (on UNIX), or
\WINNT\System32\drivers\etc\hosts files on Windows, and make the necessary DNS
changes, to associate the new IP address with the old host name.
Change VIP Addresses
5. Modify your VIP address using srvctl:

# srvctl modify nodeapps -n node1 -A
192.168.2.125/255.255.255.0/eth0
6. Start nodeapps and all resources depending on it:

# srvctl start nodeapps -n node1
s a
)h a
7. Repeat from step 1 for the next node.
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer(continued)
Change VIP i oAddresses
ic -tran
b r
Fa nonnodeapps and provide the new virtual IP address. Use the srvctl modify
5. Modify
nodeapps command with the –A option. This command should be run as root. In the slide
example, you specify the new IP address (192.168.2.125), then the corresponding netmask
(255.255.255.0), and the interface that you want the VIP to use (eth0).
6. Start nodeapps again.
7. Repeat the same steps for all the nodes in the cluster. You can stay connected from the first
node because srvctl is a clusterwide management tool.
Note: If only the IP address is changed, it is not necessary to make changes to the
listener.ora, tnsnames.ora and initialization parameter files, provided they are using the
virtual host names. If changing both the virtual host name and the VIP address for a node, it will
be necessary to modify those files with the new virtual host name. For the listener.ora file,
you can use netca to remove the old listener and create a new listener. In addition, changes will
need to be made to the tnsnames.ora file of any clients connecting to the old virtual host name.
Change Public/Interconnect IP Subnet
Configuration: Example
Use oifcfg to add or delete network interface information

in OCR:
$ <CRS HOME>/bin/oifcfg getif 1
eth0 139.2.156.0 global public
eth1 192.168.0.0 global cluster_interconnect
$ oifcfg delif -global eth0 2s a

$ oifcfg setif –global eth0/139.2.166.0:public
)h a
m ฺbr
$ oifcfg delif –global eth1 i ฺ c o deฺ3
$ oifcfg setif –global eth1/192.168.1.0:cluster_interconnect r e d Gui
@ sic dent
e r de Stu
$ oifcfg getif
eth0 139.2.166.0 global public ila
v this 4
i o _v use
eth1 192.168.1.0 global cluster_interconnect
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfer IP Subnet Configuration
i o
Change Public/Interconnect
ic -trOraclean Clusterware and RAC, it is possible for you to specify wrong
b r
Fa nonduring the OUI interview regarding the public and interconnect interfaces that
When installing
information
Oracle Clusterware should use. If that happens, Oracle Clusterware will be able to start at the
end of the installation process, but you might end up having trouble later to communicate with
other nodes in your cluster. If either the interface, IP subnet, or IP address for both your public
network and interconnect are incorrect or need to be changed, you should make the changes
using the Oracle Interface Configuration Tool (oifcfg) because this will update the
corresponding OCR information.
An example is shown in the slide, where both IP subnet for the public and private network are
incorrect:
1. You get the current interfaces information by using the getif option.
2. You delete the entry corresponding to public interface first by using the delif option, and
then enter the correct information by using the setif option.
3. You do the same for your private interconnect.
4. You check that the new information is correct.
Note: A network interface can be stored as a global interface or as a node-specific interface. An
interface is stored as a global interface when all the nodes of a RAC cluster have the same
interface connected to the same subnet (recommended). It is stored as a node-specific interface
only when there are some nodes in the cluster that have a different set of interfaces and subnets.
Third-Party Application Protection: Overview
• High Availability framework:

– Command-line tools to register applications with CRS
– Calls control application agents to manage applications
– OCR used to describe CRS attributes for the applications
• High Availability C API:
– Modify directly CRS attributes in OCR
– Modify CRS attributes on the fly
• Application VIPs: s a
)h a
– Used for applications accessed by network means ฺbr
c o m eฺ
– NIC redundancy ฺ
di Guid
r e
– NIC failover sic t
@ d e n
• OCFS:
e r de Stu
– Store application configuration
v i lav efiles t his
– Share files between ic i o_ onodes
cluster us
r
ab nse t
( f
e rde lCopyright
l a V ab e
i
V sferProtection: Overview
Third-Partyi oApplication
ic -tran
b r
Fa non
Oracle Clusterware provides two publicly available components that can be used to help protect
any application on a cluster:
• The High Availability framework provides facilities to manage your applications under
CRS protection via command-line tools such as crs_register, crs_start, and crs_stop.
This framework is also used to automatically invoke control scripts that you created so that
CRS can start, stop, and monitor your applications. OCR is used as a repository to define
failover policies and other important parameters for CRS to control your applications.
• The C API can be used to directly manipulate OCR to define how CRS should protect an
application. This API can be used to modify, at run time, how the application should be
managed by CRS. Discussing the C API is out of the scope of this course.
If the application you want CRS to protect is accessed by way of a network, you have the
possibility to create a Virtual Internet Protocol address for your application. This is referred to as
an application VIP. Application VIPs created by Oracle Clusterware are able to fail over from
one network interface card (NIC) to another on the same node as well as from one NIC to
another one located on another node in case all public networks are down on a given node.
In addition, your application might need to store configuration files on disk. To share these files
among nodes, Oracle Corporation also provides you with the Oracle Cluster File System
(OCFS).
Application VIP and RAC VIP Differences
• RAC VIP is mainly used in case of node down events:

– VIP is failed over to a surviving node.
– From there it returns NAK to clients forcing them to
reconnect.
– There is no need to fail over resources associated to the
VIP.
• Application VIP is mainly used in case of application as a
r ) h
down events: ฺ b
– VIP is failed over to another node together with c othe deฺ m
i ฺ
d Gui
application(s). r e
@ sic dent
– From there, clients can still connect
d e through
t u the VIP.
r
veonethVIP S
– Although not recommended, i l a is can serve many
v e
applications. io_ us
ric e to
a b
(f ens
r lCopyright
l
e
V ab e
i l a
V andsRAC er VIP Differences
Application i oVIP f
b r ic -tran
F a theodifferences
Most of n between resources attached to application VIPs and RAC VIPs reside in
n
the fact that they are configured differently within Oracle Clusterware. For example, it makes no
sense from a RAC perspective to fail over either a database instance or listener because there is
already a listener and an instance waiting on another node. Therefore, the listener does not listen
on any other VIPs than the one node-specific VIP. Looking at the CRS profile of those
resources, you will see the differences. Also, most of the time, there are many applications
attached to a RAC VIP such as listeners, database instances, and ASM instances. Although it is
possible to associate an application VIP to multiple applications, this is not recommended
because if one of the applications cannot be started or restarted on a node, it will be failed over
to another node with the VIP, which in turn will force the other applications to be also relocated.
This is especially true if the applications are independent. However, one noticeable difference
between a RAC VIP and an application VIP is that after a RAC VIP is failed over to a surviving
node, it no longer accepts connections (NAK), thus forcing clients that are trying to access that
address, to reconnect using another address. If it accepts new connections, then if a failback
occurs, after the node is back again, then current connections going through the VIP on the
failed-over node are lost because the interface is gone. Application VIPs, on the other side, are
fully functional after they are failed over, and continue to accept connections.
RAC VIPs are mainly used when there is a node failure because clients can use other nodes to
connect. Application
Unauthorized reproduction VIPs are mainly used
or distribution when theCopyright©
prohibitedฺ application cannot be restarted
2009, Oracle and/oron aitsnode.
affiliatesฺ
Use CRS Framework: Overview
1. Create an application VIP, if necessary:

a. Create a profile: Network data + usrvip predefined script
b. Register the application VIP.
c. Set user permissions on the application VIP.
d. Start the application VIP by using crs_start.
2. Write an application action script that accepts three a
parameters: has r)
• start: Script should start the application. m ฺb
i ฺ c o deฺ
r e d Guisi up.
• check: Script should confirm that the application
sic ent
• stop: Script should stop the application.
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i r
V sfeOverview
i o
Use CRS Framework:
ic -tran
b r
a presents
The
F slide
n o n you the basic steps you need to follow to register an application that is
monitored by the CRS framework:
1. If your application is accessed via the network, and if you want your application to be still
available after some network problems, it is recommended that you create an application
VIP for your application.
a. First, you should create an application profile to define the network information
relating to this VIP—for example, the name of the public network adapter to use,
the IP address, and the netmask. In the profile, you should also specify the usrvip
action script provided by Oracle Clusterware. You can then use the default values for
the failover policies.
b. Use the crs_register command to add this application VIP to the list of managed
applications.
c. On UNIX-based operating systems, the application VIP script must run as the root
user. So, using crs_setperm, you can change the owner of the VIP to root. Using the
same command tool, you can also enable another user, such as oracle, to start the
application VIP.
d. When done, you can use the crs_start command to start the VIP application.
2. You can now create an action script to support the start, check, and stop actions on your
application.
Use CRS Framework: Overview
3. Create an application profile:

• Action script location
• Check interval
• Failover policies
• Application VIP, if necessary
4. Set permissions on your application. a
5. Register the profile with Oracle Clusterware. h as
ฺ b r)
6. Start your application by using crs_start. om ฺ
e d iฺc uide
s icr ent G
r d e@ Stud
i l a ve this
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfeOverview r
i c i o
Use CRS Framework:
a n (continued)
r r
-tprofile for your application. You should use enough resource attributes to define
F3.abCreate
n o n
the
at least the action script location and name, the check interval, the failover policies, and the
required application VIP resource (if necessary). You can manage application availability as
follows:
- Specify starting resources during cluster or node startup.
- Restart applications that fail.
- Relocate applications to other nodes if they cannot run in their current location.
4. Like for the VIP application, you can define under which user your application should be
running as well as which user can start your application. That is why on UNIX-based
platforms, Oracle Clusterware must run as the root user, and on Windows-based platforms,
Oracle Clusterware must run as Administrator.
5. When done, you can register your application by using the crs_register command.
6. You are then ready to start your application that is going to be monitored by Oracle
Clusterware. Do this by executing the crs_start command.
Use CRS Framework: Example
# crs_profile –create AppVIP1 –t application \ 1

–a <CRS HOME>/bin/usrvip \
–o oi=eth0,ov=144.25.214.49,on=255.255.252.0
# crs_register AppVIP1 2
# crs_setperm AppVIP1 –o root 3

s a
)h a
# crs_setperm AppVIP1 –u user:oracle:r-x m ฺbr 4
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
$ crs_start AppVIP1
i l a v this 5
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i r
V sfeExample
i o
Use CRS Framework:
ic -tran
b r
Fa nothen previous overview slides, here is an example that protects the apache application
Following
using Oracle Clusterware:
1. You create the AppVP1 application VIP profile by using the crs_profile –create
command. In order, here are the parameters specified in the example:
- Name of the application VIP
- The application type
- The predefined action script usrvip located in <CRS HOME>/bin
- The name of the public network adapter, the VIP address used to locate your
application regardless of the node it is running on, and the netmask used for the VIP
The result of this command is to create a text file called AppVIP1.cap in
<CRS HOME>/crs/profile. This file contains the attributes and is read by crs_register.
If your session is not running as the root user, the .cap file is created in <CRS
HOME>/crs/public.
2. Use the crs_register command to register your application VIP with Oracle Clusterware.
3. On UNIX-based operating systems, the application VIP action script must run as the root
user. As the root user, change the owner of the resource as shown using the
crs_setperm –o command.
4. As the root user, enable the oracle user to manage your application VIP via CRS
commands. Use the crs_setperm –u command.
5. As the oracle user, start the application VIP using the crs_start command.
#!/bin/sh
6
VIPADD=144.25.214.49
HTTDCONFLOC=/etc/httpd/conf/httpd.conf
WEBCHECK=http://$VIPADD:80/icons/apache_pb.gif
case $1 in
'start')
/usr/bin/apachectl –k start –f $HTTDCONFLOC
RET=$?
;;
'stop')
/usr/bin/apachectl –k stop
s a
RET=$?
)h a
ฺbr
;;
'check')
c m
o deฺ
/usr/bin/wget –q –delete-after $WEBCHECK
i ฺ
d Gui
RET=$?
r e
*)
;;
@ sic dent
RET=0
e r de Stu
;;
v i lav e this
o_ o us
esac
exit $RET
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfeExample r
i c i o
Use CRS Framework:
a n (continued)
r t r
F6.abAfternothen-application VIP is functional, you can write the action script for your application.
The example shown in the slide can be used by Oracle Clusterware as an action script to
protect the apache application. It is a shell script that can parse one argument with three
different values. It uses the apachectl command tool to start and stop the apache
application on your node. It uses the wget command to check whether a Web page can be
accessed. These are the three actions CRS will perform while protecting your application.
For the next steps, it is supposed that this script is called myApp1.scr.
Note: Make sure you distribute this script on all nodes of your cluster in the same location. The
default location is assumed to be <CRS HOME>/crs/script in this case.
# crs_profile –create myApp1 –t application –r AppVIP1 \

–a myapp1.scr –o ci=5,ra=2 7
# crs_register myApp1 8
# crs_setperm myApp1 –o root 9 s a

)h a
# crs_setperm myApp1 –u user:oracle:r-x m ฺbr 10
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
$ crs_start myApp1
i l a v this 11
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfeExampler
i c i o
Use CRS Framework:
a n (continued)
r t r
F7.abYounocann-now create a profile for your application. Here your resource is called myApp1. It
uses myApp1.scr as its action script and depends on the AppVIP1 application. If AppVIP1
fails or if it is relocated to another node, then Oracle Clusterware stops or moves the
myApp1 application. The example also defines its check interval to be five seconds, and
the number of attempts to restart the application to 2. This means that Oracle Clusterware
will fail over the application to another node after a second local failure happens.
8. The crs_register command registers myApp1 with Oracle Clusterware.
9. Because you want the apache server listening on the default port 80, you want the
application to execute as the root user. As the root user, change the owner of the
resource, as shown, using the crs_setperm –o command.
10. As the root user, enable the oracle user to manage your application VIP via CRS
commands. Use the crs_setperm –u command.
11. As the oracle user, start myApp1 by using the crs_start command.
Summary

• Manually control the Oracle Clusterware stack
• Change voting disk configuration
• Backup and recover your voting disks
• Manually back up OCR
• Recover OCR s a
h a
• Replace an OCR mirror ฺ b r)
• Repair the OCR configuration c o m eฺ
ฺ
di Guid
r e
• Change VIP addresses
@ sic dent
• Use the CRS framework e r de Stu
• Prevent automatic instancev i lavrestarts
e t his
_ io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non

• Mirroring the OCR
• Backing up and restoring OCR
• Multiplexing the voting disk
• Using Oracle Clusterware to protect the xclock
application a
as
)h
ฺbr
c m
o deฺ
i ฺ
d Gui
r e
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Diagnosing Oracle Clusterware
and RAC Components
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
Objectives

• Collect Oracle Clusterware diagnostic files
• Use Cluster Verify
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
One Golden Rule in RAC Debugging
• Always make sure that your nodes have exactly the

same system time to:
– Facilitate log information analysis
– Ensure accurate results when reading GV$ views
– Avoid untimely instance evictions
• The best recommendation is to synchronize nodes
s a
using Network Time Protocol. a
)h ฺbr
c m
o deฺ
i ฺ
d Gui
r e
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V insRAC f er Debugging
One Golden i o Rule
ic -tran
b r
Fa nonrecommended to set up Network Time Protocol (NTP) on all cluster nodes, even
It is strongly
before you install RAC. This will synchronize the clocks among all nodes, and facilitate analysis
of tracing information based on time stamps as well as results from queries issued on GV$ views.
Note: Adjusting clocks by more than 15 minutes can cause instance evictions. It is strongly
advised to shut down all instances before date/time adjustments.
Oracle Clusterware Main Log Files
/etc
ORA_CRS_HOME ORACLE_HOME
oracle
log log
<nodename>.oprocd.log
<hostname> <hostname>
s a
crsd cssd evmd racg client client racg
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
racgeut racgevtf racgmain
@ sic dent
e r de Stu
alert<nodename>.log
v i lavracgeute t hisracgimon racgmain racgmdb
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sMain f er Log Files
i o
Oracle Clusterware
ic -tran
b r
Fa non
Oracle Clusterware uses a unified log directory structure to consolidate the Oracle Clusterware
component log files. This consolidated structure simplifies diagnostic information collection and
assists during data retrieval and problem analysis.
The slide shows you the main directories used by Oracle Clusterware to store its log files:
• CRS logs are in $ORA_CRS_HOME/log/<hostname>/crsd/. The crsd.log file is
archived every 10 MB (crsd.l01, crsd.l02, …).
• CSS logs are in $ORA_CRS_HOME/log/<hostname>/cssd/. The cssd.log file is
archived every 20 MB (cssd.l01, cssd.l02, …).
• EVM logs are in $ORA_CRS_HOME/log/<hostname>/evmd.
• Depending on the resource, specific logs are in
$ORA_CRS_HOME/log/<hostname>/racg and in
$ORACLE_HOME/log/<hostname>/racg. In the last directory, imon_<service>.log
is archived every 10 MB for each service. Each RACG executable has a subdirectory
assigned exclusively for that executable. The name of the RACG executable subdirectory is
the same as the name of the executable.
• SRVM (srvctl) and OCR (ocrdump, ocrconfig, ocrcheck) logs are in
$ORA_CRS_HOME/log/<hostname>/client/ and in
$ORACLE_HOME/log/<hostname>/client/.
• Important Oracle Clusterware alerts can be found in alert<nodename>.log in the
$ORA_CRS_HOME/log/<hostname> directory.
Diagnostics Collection Script
• Script to collect all important log files:

– Must be executed as root
– Is located in $ORA_CRS_HOME/bin/
– Is called diagcollection.pl
• Generates the following files in the local directory:
– basData_<hostname>.tar.gz – ocrData _<hostname>. tar.gz
– crsData _<hostname>. tar.gz – s a
oraData _<hostname>. tar.gz
)h a
m ฺbr
i ฺ c o deฺ
# export ORACLE_HOME=/u01/app/oracle/product/10.2.0/db_1
r e d Gui
# export ORA_CRS_HOME=/u01/crs1020 s ic ent
# export ORACLE_BASE=/u01/app/oracle
d @
e Stud
# cd $ORA_CRS_HOME/bin r
ve this
i l a
# ./diagcollection.pl -collect
i o _v use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
V sfeScript r
Diagnostics i o Collection
c tran
idiagcollection.pl
b r
Use
F a the
n o n- script to collect diagnostic information from an Oracle
Clusterware installation. The diagnostics provide additional information so that Oracle Support
can resolve problems. This script is located in $ORA_CRS_HOME/bin. Before executing the
script, you must be logged in as the root user, and you must set the following environment
variables: ORACLE_BASE, ORACLE_HOME, ORA_CRS_HOME, HOSTNAME. The example in the
slide shows you how to invoke the script to collect the diagnostic information. When invoked
with the –collect option, the script generates, in the local directory, the four files mentioned
in the slide. Mainly, basData.tar.gz contains log files from the $ORACLE_BASE/admin
directory. crsData.tar.gz contains log files from $ORA_CRS_HOME/log/<hostname>.
The ocrData.tar.gz files contain the results of an ocrdump, ocrcheck, and the list of ocr
backups. oraData.tar.gz contains log files from $ORACLE_HOME/log/<hostname>.
If you invoke the script with the -collect option, and you already have the four files generated
from a previous run in the local directory, the script asks you if you want to overwrite the
existing files. You can also invoke the script with the –clean option to clean out the files
generated from a previous run in your local directory. Alternatively, you can invoke the script to
just capture a subset of the log files. You can do so by adding extra options after the –collect
option: -crs for collecting Oracle Clusterware logs, -oh for collecting ORACLE_HOME logs, -ob
for collecting ORACLE_BASE logs, or –all for collecting all logs. The –all option is the
default. The –coreanalyze option enables you to extract to text files only core files found in
the generated
Unauthorized files.
reproduction or distribution prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
Managing Diagnostic Data in RAC
Support Workbench
DIAGNOSTIC_DEST
ADR Shared
Base ADR base
diag
asm rdbms
s a
DB
)h a
ฺbr
+asm
Name
c m
o deฺ
+ASM1 +ASMn SID1 SIDn
ADR
i ฺ
d Gui
r e
Home
@ sic dent
e r de Stu
ADRCI
v i lav e this GV$DIAG_INFO
_
io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
V sfData er in RAC
o
Managing iDiagnostic
ic -tran
b r
Fa nthat
Problems
on span Oracle RAC instances can be the most difficult types of problems to
diagnose. For example, you may need to correlate the trace files from across multiple instances,
and merge the trace files. Oracle Database Release 11g includes an advanced fault diagnosibility
infrastructure for collecting and managing diagnostic data, and uses the Automatic Diagnostic
Repository (ADR) file-based repository for storing the database diagnostic data. When you
create the ADR base on a shared disk, you can place ADR homes for all instances of the same
Oracle RAC database and the all corresponding ASM instances under the same ADR Base. With
shared storage, you can use the ADRCI command-line tool to correlate diagnostics across all
instances because some ADRCI commands (such as SHOW INCIDENT) can work with multiple
ADR homes simultaneously.
Note: Although not required, it is recommended that you share ADR base with your RAC
databases. However, if you are using shared Oracle homes, you must share your ADR base.
Cluster Verify: Overview
• To verify that you have a well-formed cluster for Oracle

Clusterware and RAC:
– Installation
– Configuration
– Operation
• Full stack verification
s a
• Nonintrusive verification h a
ฺ b r)
• Diagnostic mode seeks to establish a reason ofor m theeฺ
failure of any verification task. ฺ c
di Guid
ic r e t
• Easy-to-use interface: @s den de Stu
– Stage commands e r
– Component commands_ v
v
ila e this
io s u
b r ic t o
e (fa ense
rd lCopyright
l a V ab
i
VOverviewf er
i o
Cluster Verify:
ic -tran s
b r
Fa Verification
Cluster
non
Utility (CVU) is provided with Oracle Clusterware and Oracle Database
10g Release 2 (and later) with Real Application Clusters. The purpose of CVU is to enable you
to verify during setup and configuration that all components required for a successful installation
of Oracle Clusterware or Oracle Clusterware and a RAC database are installed and configured
correctly, and to provide you with ongoing assistance any time you need to make changes to
your RAC cluster.
There are two types of CVU commands:
• Stage commands are CVU commands used to test system setup and readiness for successful
software installation, database creation, or configuration change steps. These commands are
also used to validate successful completion of specific cluster configuration steps.
• Component commands are CVU commands used to check individual cluster components,
and determine their state.
It is recommended to use stage checks during the installation of Oracle Clusterware and RAC.
In addition, you can use CVU to verify a particular component while the stack is running or to
isolate a cluster subsystem for diagnosis. During the diagnostic mode of operation, CVU tries to
establish a reason for the failure of any verification task to help diagnose a problem.
Note: CVU is a nonintrusive tool in the sense that it does not try to fix any issues it finds.
Cluster Verify Stages
-pre dbcfg
$ cluvfy stage -list
Configures
RAC DB
-pre dbinst
Installs
RAC
-pre crsinst
Installs
CRS
s a
-pre cfs
a
-post crsinst
)h
Sets up OCFS
m ฺbr
(OPT)
i ฺ c o deฺ
-post cfs
r e d Gui
@ sic dent
User sets up the
e r de Stu
lav e this
hardware,
network, & storage -post ihwos
v
o_ o us
ic i
( f abr nse t
e rde lCopyright
l a V ab e
i
VStagessfer
i o
Cluster Verify
ic -tran
b r
Fa isnoa n
A stage specific phase of an Oracle Clusterware or RAC deployment. Before performing any
operations in a stage, a predefined set of checks must be performed to ensure the readiness of
cluster for that stage. These checks are known as “pre” checks for that stage. Similarly, a
predefined set of checks must be performed after completion of a stage to ensure the correct
execution of operations within that stage. These checks are known as “post” checks for that
stage. You can list verifiable stages with the cluvfy stage -list command. All stages have
pre or post steps and some stages have both. Valid stage options and stage names are:
• -post hwos: Postcheck for hardware and operating system
• -pre cfs: Precheck for CFS setup
• -post cfs: Postcheck for CFS setup
• -pre crsinst: Precheck for CRS installation
• -post crsinst: Postcheck for CRS installation
• -pre dbinst: Precheck for database installation
• -pre dbcfg: Precheck for database configuration
Cluster Verify Components
• An individual subsystem or a module of the RAC

cluster is known as a component in CVU.
• The availability and integrity of a cluster component
can be verified.
• Components can be simple like a specific storage
device, or complex like the Oracle Clusterware stack:
s a
– Space availability – Cluster integrity h a
– Shared storage – Administrative ฺ b r)
c o m eฺ
accessibility privileges ฺ
di Guid
r e
– Node connectivity sic compatibility
– Peer
@ e n t
– Cluster File System integrity rd– e System
S t ud requirements
i l
– Oracle Clusterware integrity a ve this
v o _ use $ cluvfy comp -list
i
ric e to
a b
(f ens
r lCopyright
l
e
V ab e
i l a
VComponents er
i
Cluster Verifyo n s f
b r ic -thet r anotion
CVU
F a supports
o n of component verification. The verifications in this category are not
n
associated with any specific stage. A component can range from a basic one, such as free disk
space, to a complex one (spanning over multiple subcomponents), such as the Oracle
Clusterware stack. Availability, integrity, or any other specific behavior of a cluster component
can be verified. You can list verifiable CVU components with the cluvfy comp -list
command.
• nodereach: Checks reachability between nodes
• nodecon: Checks node connectivity
• cfs: Checks Oracle Cluster File System integrity (The sharedness check for the file system
is supported for OCFS2 versions 1.2.1 and later.)
• ssa: Checks shared storage accessibility
• space: Checks space availability
• sys: Checks minimum system requirements
• clu: Checks cluster integrity
• clumgr: Checks cluster manager integrity
• ocr: Checks OCR integrity
• crs: Checks CRS integrity
• nodeapp: Checks node applications existence
• admprv: Checks administrative privileges
• peer:
Unauthorized Compares properties
reproduction with peers
or distribution prohibitedฺ Copyright© 2009, Oracle and/or its affiliatesฺ
Cluster Verify Locations
• Download it from OTN:

– Create a local directory.
– Copy and extract cvu_<OS>.zip.
• Oracle software DVD:
– Disk1 directory
– runcluvfy.sh
s a
• Oracle Clusterware home: )h a
– $ORA_CRS_HOME/bin/cluvfy m ฺbr
i ฺ c o deฺ
• Oracle Home: r e d Gui
– $ORACLE_HOME/bin/cluvfy @ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
VLocationsf er
i o
Cluster Verify
ic -tran s
b r
Fa nonVerification Utility (CVU) is first released in Oracle Clusterware release 10.2.0.1.0.
The Cluster
CVU supports 11gR1, 10gR2 as well as 10gR1 for Oracle Clusterware and RAC products. CVU
is available in three different forms:
• Available on Oracle Technology Network (OTN) at:
http://www.oracle.com/technology/products/database/clustering/cvu/cvu_download_homep
age.html
From there, you need to download the package and unzip it to a local directory
(<cvhome>). You can use the cluvfy command from the <cvhome>/bin. Optionally,
you can set the CV_DESTLOC environment variable. This should point to a writable area
on all nodes. CVU attempts to copy the necessary bits as required to this location. If this
variable is not set, CVU uses /tmp as the default.
• Available in 11.1 Oracle software DVD as packaged version. Make use of
runcluvfy.sh, which is needed when nothing is installed. You can find it in Disk1.
• Installed in both 11.1 Oracle Clusterware and RAC homes. Make use of cluvfy if the
CRS software stack is installed. If the CRS software is installed, you can find cluvfy
under $ORA_CRS_HOME/bin.
Note: For manual installation, you need to install CVU on only one node. CVU deploys itself on
remote nodes during executions that require access to remote nodes.
Cluster Verify Configuration File
$ cat cvu_config
# Configuration file for CVU
# Version: 011405
#
#CV_ORACLE_RELEASE=11gR1
#CV_NODE_ALL=
CV_RAW_CHECK_ENABLED=TRUE
s a
)h a
CV_ASSUME_DISTID=RHEL4
m ฺbr
i ฺ c o deฺ
#CV_XCHK_FOR_SSH_ENABLED=TRUE r e d Gui
@ sic dent
#ORACLE_SRVM_REMOTESHELL=/usr/bin/ssh
e r de Stu
v i l av this
#ORACLE_SRVM_REMOTECOPY=/usr/bin/scp
i o _ use
a b ric e to
e (f ens
e d
r lCopyright
l a V ab
i
VConfiguration
f er File
i
Cluster Verifyo s
an configuration file to define specific inputs for the execution of the CVU.
icuse the-trCVU’s
b r
Fa
You can
onthe configuration file is $CV_HOME/cv/admin/cvu_config. The following is
The pathnfor
the list of keys supported in cvu_config:
• CV_NODE_ALL: If set, it specifies the list of nodes that should be picked up when Oracle
Clusterware is not installed and the -n all option has been used in the command line.
• CV_RAW_CHECK_ENABLED: If set to TRUE, it enables the check for accessibility of
shared SCSI disks on Red Hat release 3.0 and higher. This shared disk accessibility check
requires that you install a cvuqdisk rpm on all the nodes. By default, this key is set to TRUE
and shared disk check is enabled.
• CV_ASSUME_DISTID: Specifies the distribution ID that CVU uses. For example, to make
CVU working with SuSE 9 ES, set it to Pensacola.
• CV_XCHK_FOR_SSH_ENABLED: If set to TRUE, it enables the X-Windows check for
verifying user equivalence with ssh. By default, this entry is commented out and X-
Windows check is disabled.
• ORACLE_SRVM_REMOTESHELL: If set, it specifies the location for the ssh/rsh
command to override CVU’s default value. By default, this entry is commented out and the
tool uses /usr/sbin/ssh and /usr/sbin/rsh.
Note: If CVU does not find a key entry defined in the configuration file, the CVU searches for
the environment variable that matches the name of the key; otherwise, the CVU uses a default.
Cluster Verify Configuration File (continued)
• ORACLE_SRVM_REMOTECOPY: If set, it specifies the location for the scp or rcp
command to override the CVU default value. By default, this entry is commented out and
the CVU uses /usr/bin/scp and /usr/sbin/rcp.
If the CVU does not find a key entry defined in the configuration file, the CVU searches for the
environment variable that matches the name of the key. If the environment variable is set, the
CVU uses its value. Otherwise it uses a default value for that entity.
To provide the CVU with a list of all the nodes of a cluster, you can use the -n all option
while executing a command. The CVU attempts to obtain the node list in the following
sequence:
1. If vendor clusterware is available, the CVU selects all the configured nodes from the vendor
clusterware using the lsnodes utility.
2. If Oracle Clusterware is installed, the CVU selects all the configured nodes from Oracle
Clusterware using the olsnodes utility. s a
3. If neither the vendor nor Oracle Clusterware is installed, the CVU searches forra) value
a
h for
the CV_NODE_ALL key in the configuration file. m ฺ b
i ฺ c o d e ฺ
If the vendor and Oracle Clusterware are not installed and if no key e
r d CV_NODE_ALL
named u i
exists in the configuration file, the CVU searches for a value fori c t G
s the dCV_NODE_ALL
n
@ e
e r sS dethe CVUtureports an error.
environmental variable. If you have not set this variable,
i v
la e thi
v
o_ o us
ic i
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non
Cluster Verify: Examples
$ cluvfy comp sys -n node1,node2 -p crs -verbose 1
$ cluvfy comp ssa -n all -s /dev/sda1 2
$ cluvfy comp space -n all -l /home/product -z 5G 3
$ cluvfy comp nodereach -n node2 -srcnode node1 4
s a
5a
$ cluvfy comp nodecon -n node1,node2 –i eth0 -verbose h
ฺ b r)
c o m eฺ6
$ cluvfy comp admprv -n all -o user_equiv -verbose ฺ
di Guid
r e
$ cluvfy comp nodeapp -n all -verbose @ sic dent 7
e
rd s S t u
v e
$ cluvfy comp peer -n all –verbose
v i la e th| i more 8
_
io o us
r ic
( f ab nse t
e rde lCopyright
l a V ab e
i
VExamplesf er
i o
Cluster Verify:
ic -tran s
b r
Fa nshows
The slide
on you some possible interesting examples:
1. To verify the minimal system requirements on the nodes before installing Oracle
Clusterware or RAC, use the sys component verification command. To check the system
requirements for installing RAC, use the -p database argument, and to check the system
requirements for installing Oracle Clusterware, use the -p crs argument. To check the
system requirements for installing Oracle Clusterware or RAC from Oracle Database 10g
release 1 (10.1), use the -r 10gR1 argument. The example verifies the system requirements
for installing Oracle Clusterware on the cluster nodes known as node1 and node2.
2. To verify whether storage is shared among the nodes in your cluster database or to identify
all of the storage that is available on the system and can be shared across the cluster nodes,
use the component verification command ssa. The example uses the –s option to specify
the path to check.
3. You are planning to install more software on the local /home/product file system of each
node in the cluster, and that software will take up 5 GB on each node. This command is
successful if 5 GB is available in /home/product of every node; otherwise, it fails.
Note: The –verbose option can be used with any command. It basically gives you more
information in the output.
Cluster Verify: Examples (continued)
4. To verify the reachability of the cluster nodes from the local node or from any other cluster
node, use the component verification command nodereach. The example tries to check
whether node2 can be reached from node1.
5. To verify the connectivity between the cluster nodes through all of the available network
interfaces or through specific network interfaces, use the component verification command
nodecon. The example checks whether node1 and node2 can communicate through the
eth0 network interface. Without the -i option, the CVU discovers all the network
interfaces that are available on the cluster nodes, reviews the interfaces’ corresponding IP
addresses and subnets, obtains the list of interfaces that are suitable for use as VIPs and the
list of interfaces suitable for use as private interconnects, and verifies the connectivity
between all the nodes through those interfaces.
6. To verify user accounts and administrative permissions–related issues for user equivalence,
Oracle Clusterware installation, and RAC installation, use the component verification a
command admprv. On Linux and UNIX platforms, the example verifies user equivalence h a s
for all the nodes by first using ssh and then using rsh if the ssh check fails.ฺbTo r)verify the
equivalence only through ssh, use the -sshonly option. By default,ฺc o
the m e ฺ check
equivalence
does not verify X-Windows configurations, such as when yourhave
i
ed disabled i d
u X-forwarding
i c t G
with the setting of the DISPLAY environment variable. To s n
verify X-Windows
e aspects during
e @ t u d
user equivalence checks, set the CV_XCHK_FOR_SSH_ENABLED
v e rd s S key to TRUE in the
configuration file before you run the command.
v i la e tthe
Use hi -o crs_inst argument to verify
whether you have permissions to install
ic i o_ Oracle
o u sClusterware. You can use the -o db_inst
argument to verify the permissions
( f a br that s earet required for installing RAC and the -o
d c n
db_config argument toeverify theepermissions that are required for creating a RAC
r l i
e blaeRAC database’s configuration.
database or for modifying
a V a existence of node applications, namely VIP, ONS, and GSD, on all
7. The example
V il verifies
f e rthe
theic o
i To
nodes.
a s
n the integrity of all the Oracle Clusterware components, use the
verify
b r
component t r
- verification crs command. To verify the integrity of each individual Cluster
Fa Manager
nonsubcomponent (CSS), use the component verification command clumgr. To verify
the integrity of Oracle Cluster Registry, use the component verification ocr command. To
check the integrity of your entire cluster, which means to verify that all the nodes in the
cluster have the same view of the cluster configuration, use the component verification clu
command.
8. The example compares all the nodes and determines whether any differences exist between
the values of preselected properties. This is successful if the same setup is found across all
the nodes. You can also use the comp peer command with the -refnode option to compare
the properties of other nodes against the reference node. This command allows you to
specify the –r 10gR1 option. Here is a truncated list of the preselected properties: Total
memory, Swap space, Kernel version, System architecture, Package existence for
various components (glibc, make, binutils, gcc, compat-db, ...), Group existence for
"oinstall", Group existence for "dba", User existence for "nobody".
Note: For stage examples, refer to the installation lessons in this course.
Cluster Verify Output: Example
$ cluvfy comp crs -n all -verbose
Verifying CRS integrity

Checking CRS integrity...
…
Liveness of all the daemons
Node Name CRS daemon CSS daemon EVM daemon
------------ ------------- --------------- ----------
atlhp9 yes yes yes
atlhp8 yes yes yes
s a
Checking CRS health...
)h a
Check: Health of CRS
m ฺbr
Node Name CRS OK?
i ฺ c o deฺ
---------------------------------- ------------------------
r e d Gui
atlhp9 yes
@ sic dent
atlhp8 yes
e r de Stu
lav e this
Result: CRS health check passed.
CRS integrity check passed.
v i
i o_ o us
Verification of CRS integrity was successful.
ic
( f abr nse t
e rde lCopyright
l a V ab e
i
VOutput: f r
eExample
i o
Cluster Verify
ic -tran s
b r
Fa nshows
The slide
on you the output of the cluvfy comp crs –n all –verbose command. This
command checks the complete Oracle Clusterware stack.
Note: The output is truncated for formatting reasons.
Summary

• Collect Oracle Clusterware diagnostic files
• Use Cluster Verify
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non

• Identifying Oracle Clusterware log files
• Fixing voting disk corruptions
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde lCopyright
l a V ab e
i
V sfer
i o
ic -tran
b r
Fa non
s a
)h a
m ฺbr
i ฺ c o deฺ
r e d Gui
@ sic dent
e r de Stu
v i lav e this
ic i o_ o us
( f abr nse t
e rde le lice
l a V ab
i
V sfer
i o
ic -tran
b r
Fa non

D50311GC11 sg1

Uploaded by

Copyright:

Available Formats

D50311GC11 sg1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

D50311GC11 sg1

Uploaded by

Copyright:

Available Formats

s a

James Womack Disclaimer

This document contains proprietary information and is protected by copyright and

1 Oracle Clusterware Installation and Configuration

2 RAC Software Installation

Summary 2-31 ( f abr nse t

8 High Availability of Connections

10 Diagnosing Oracle Clusterware and RAC Components

11 Node Addition and Removal

12 Design for High Availability

Appendix A: Practices and Solutions

Appendix B: Miscellaneous Topics

Appendix C: Practice Environment Scripts

• This course is designed for anyone interested in

In this course, you:

Topics Lessons Day

Introduction and installation I, 1, 2, 3 1

RAC administration and tuning 4, 5, 6 2

• High availability: Surviving node and instance failures

SMP model RAC model

• Hardware: Disk input/output (I/O)

Hardware Time 100% of task

Cluster system scaleup Cluster system speedup

Workload Speedup Scaleup

OLTP and Internet No Yes

DSS with parallel query Yes Yes

Each machine has 2 CPUs:

Each machine has 2 HBAs:

Each switch needs to support 800 MB/s

9i RAC 11g Oracle Clusterware

SGA1 SGA2 SGA1 SGA2

Node1 Node2 Node3

• Retrieve information about all started instances

• Heuristics for scalability cases:

Node1 Node2 Node1 Node2

Execution slaves have node affinity with the execution

Node 1 Node 2 Node 3 Node 4

Oracle Clusterware Oracle Clusterware s a

Node1 Noden Node1 Noden

Voting files Voting files

• Storage is a critical component of grids:

• Is a shared disk cluster file system for Linux and

• Provides the first portable and high-performance

High-speed Interconnect: Gigabit Ethernet Proprietary

1. Connect and log in to http://metalink.oracle.com.

Modify service to instance mapping.

• RAC scalability and transaction throughput

After completing this lesson, you should be able to:

Oracle RAC 11g incorporates a two-phase installation

1. Complete preinstallation tasks:

• Startup and shutdown services

Check system requirements.

• At least 1 GB of physical memory is needed.

• A minimum of 1 GB of swap space is required.

• Each node must have at least two network adapters.

clnode-1 clnode-2 clnode-1vip

• Supported interconnect software protocols are

• Package versions are checked by the cluvfy utility.

• Create an oracle user, a dba, and an oinstall group