11g R2 RAC Administration 2

Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 17

Oracle 11g R2 RAC Administration

Presented by

Ebin & Suraj

Objectives
After completing this lesson, you should be able to do the following:
Definitions Differences - Oracle 10g R2, 11g R1, 11g R2 New Features in 11g R2 RAC Terms / Terminology in 11g R2 RAC 11g R2 RAC Architecture Clusterware Scripts ASM Concepts ASM Architecture ASM Sample Scripts

Definitions

Cluster

Clusterware

ASM

ACFS

Grid Infrastructure + Database

Oracle RAC 11g R2

Differences
10g R2 11G R1 11G R2 Node Reboot : OPROCd process NTP file is used for Time synch Across other nodes ASM is handled using Asmcmd and adding/modifying diskgroup was done by logging in as /as sysdba ASM doesnt support other files Including OCR or Voting Disk, but still these were called as ASM files Application Failover takes huge time . Has FAN,ONS to handle this process We had to hardcode details In /etc/hosts file on each node and we used DNS server having each entries NO separate Load Balancing Technique handled Node Reboot : OPROCd process NTP file is used for Time synch Across other nodes ASM is handled using Asmcmd Handled by OHASd process We have NTP as well as CTSSD. Here CTSSD act as observer Additionally Here we have ASMCA to add diskgroups. We can use the foll Sqlplus /as sysasm ASM can support 3rd party files as Well as we can store OCR and Voting Disk file. We have ADVM for this Fast Connection Failover. Less Failover time .Has FAN ,ONS . We have GNS which handles all IP addresses dynamically that is fetched from DHCP.

ASM doesnt support other files Including OCR or Voting Disk, but still these were called as ASM files Application Failover takes huge time . Has FAN,ONS to handle this process We had to hardcode details In /etc/hosts file on each node and we used DNS server having each entries NO separate Load Balancing Technique handled

Here we have SCAN listener To handle requests effectively, hence having more high availability Only 2 Oracle Home . ASM+Clusterware in same home Voting Disk backup is automatic whenever there is structural change and stores info in OCR.

Database, Clusterware,ASM had Separate home Manual Backup of Voting Disk Is highly risky and important

Database, Clusterware,ASM had Separate home Manual Backup of Voting Disk Is highly risky and important

Terms / Terminology in 11g R2 RAC


OCR VIP GNS DHCP DNS GPNP MDNS SCAN ONS FAN Oracle Cluster registry Virtual IP address Grid Naming Service Dynamic Host configuration Protocal Domain Name server Grid Plug and Play Multicast Domain Name Service Single Client Access Name Oracle Notification Service FAST Application Notification

Terms / Terminology in 11g R2 RAC (Continue.)


CTSS CSS EVM Cluster Time Synchronization Service Cluster Synchronization Service Event Manager

ASMCA
ACFS ADVM

ASM Configuration Assistant


Oracle ASM Cluster File System Oracle Dynamic Volume Manager

RAID
LBA TAF

Redundant Array of inexpensive disks


Logical Block Address Transparent Application Failover

SCAN (Single Client Access Name)

Network Requirements For Using SCAN


Define the SCAN in your corporate DNS (Domain Name Service) Use the Grid Naming Service (GNS)

SCAN Configuration in the Cluster

11g R2 RAC Architecture

11g R2 RAC Architecture (by hand)

Clusterware Scripts
Description SCAN STATUS Check config of scan Start Scan Relocate Scan srvctl status scan srvctl config scan $GRID_HOME/bin/srvctl start scan_listener Commands

srvctl relocate scan -i 1 -n plwcdbuat01 -- (plwcdbuat01 is a new node name where u want to reallocate scan 1)
/u01/app/11.2.0/grid/bin/crsctl lsmodules crs /u01/app/11.2.0/grid/bin/crsctl lsmodules evm /u01/app/11.2.0/grid/bin/crsctl lsmodules css /u01/crs/oracle/product/11.2.0/crs/log/oraclebi2/agent/crsd/or arootagent_root ocrcheck cluvfy comp gpnp olsnodes -n -p -i crsctl start cluster -n oraclebi1 crsctl stop resource ora.crsd

To list all the sub-modules

Network issue To check ocr file Grid Plug and play nodes in the cluster Start Cluster services on bi1 Stop resource

Clusterware Scripts (Continue.)


Description start resource Stop diskgroups and ASM instance Commands crsctl start resource ora.crsd crsctl start res ora.registry.acfs -n oracletest1 srvctl stop diskgroup -g data srvctl stop diskgroup -g backup srvctl stop asm

To create a dependency $srvctl modify database -d <unique database_name> -a between FRA diskgroup and a "DATA,FRA" database. $srvctl config database -d <unique database_name> Add database srvctl add database -d live -o $ORACLE_HOME srvctl add instance -d live -i live2 -n prod-rac02 srvctl enable database -d live srvctl enable instance -d live -i live2

srvctl help crsctl help

srvctl -h crvctl -h
alter system set remote_listener='oraclei-scan.est.org.uk:1521' scope=both sid='*'; alter system set local_listener='(DESCRIPTION=(ADDRESS_LIST=(ADDRESS =(PROTOCOL=TCP)(HOST=oraclebi-vip)(PORT=1521))))' scope=both sid='ABC'; ALTER SYSTEM REGISTER;

Scan IP settings and Register Database

ASM Concepts

ASM DISK

Diskgroup
Redundancy Level Fast Mirror Resync IDP

ASM Architecture

ASM Sample Scripts


Description Commands CREATE DISKGROUP data_dg1 EXTERNAL REDUNDANCY DISK /dev/dsk/sdc5, /dev/dsk/sdd5 # vi /etc/sysconfig/rawdevices

create ASM Disk Group

/dev/raw/raw1
/dev/raw/raw2 configure ASM Disks in UNIX / LINUX Environment /dev/raw/raw3 # vi /etc/rc.d/rc.local

/dev/add1
/dev/sdd2 /dev/sdd3

chown oracle:dba /dev/raw/raw*


chmod 660 /dev/raw/raw* a) Download ASMLIB rpm at oracle site b) /etc/init.d/oracleasm createdisk VOL1 /dev/sde1 alter diskgroup data_dg1 add failgroup ctlr3 disk /asmdisks/disk5 name dg5 noforce alter diskgroup data_dg1 resize all size 200m

configure ASM Disks using ASMLIB add new disks to an existing ASM disk group resizing an ASM disk group

ASM Sample Scripts (Continue)


Description Commands check all disks in a specific disk alter diskgroup data_dg2 check all norepair group without repairing

check a specific disk in a specific disk group and repair the disk if needed
mount and dismounting ASM disk groups

alter diskgroup data_dg1 check disk dg_5 alter diskgroup data_dg1 dismount; alter diskgroup data_dg1 mount;

drop just one disk form an ASM alter diskgroup data_dg1 drop disk dg_5 disk group undrop a recently-dropped disk create Tablespaces using ASM Disks alter diskgroup data_dg1 undrop disk dg_5 create tablespace indy_data datafile +DATA_DG1 size 100m set the parameter db_create_file_dest to +DATA_DG1

Questions

Thanks For Joining !!


Special Thanks

Sailaja
[email protected] Organized by Jayakrishnan Pillai [email protected]

Ebin Kadavi Email: [email protected]

Suraj Daga [email protected]

You might also like