SUSE Linux Enterprise

Server and High

Performance Computing
 SUSE

SUSE, headquartered in Nrnberg / Germany, is an independently

operating business unit of The Attachmate Group, Inc.

The Attachmate Group is a privately held 1 billion $ + software

company with four brands:

2 SUSE, All rights reserved.

 From SUSE to Attachmate

2000
1999
2001

1992 2002

2004

2011

3 SUSE, All rights reserved.

 SUSE Linux Enterprise

The most interoperable platform for mission-critical computing,

both physical and virtualfrom the desktop to the data center

4 SUSE, All rights reserved.

SUSE Products
Tools
 Certified Applications: SUSE Linux
Enterprise (v9+) vs Red Hat Enterprise Linux
(v4+) by quarter

March 2011 count: 1575 ISVs and 6812 applications

Novell ISV catalog: www.novell.com/partner/isv
Red Hat ISV catalog: www.redhat.com/apps/isv_catalog/browse_by_vendor.htm

6 SUSE, All rights reserved.

 WebYaST
Key Benefits
Just like YaST, WebYaST is also open source software, with
maintenance and support delivered by Novell as part of the
SUSE Appliance Program. Software vendors and third-party
developers are encouraged to develop modules
specifically aimed at facilitating configuration of their
application.
Key Features
Initial configuration wizard
Time, Time zone, NTP
Status, soft-shutdown, reboot

Monitoring (configurable)

Update

Network

License/EULA and Registration

Users

Service start/stop/status

Log visualization

7 SUSE, All rights reserved.

 Subscription Management Tool
Overview
SMT is a proxy and auditing tool that mirrors the
Novell Customer Center and tightly integrates with
it. It allows you to accurately register and manage
an entire SLE deployment, guaranteeing the
subscription compliance and secure IT process
flow organizations require.

8 SUSE, All rights reserved.

 SUSE Manager

9 SUSE, All rights reserved.

 Open Build Service (OBS)

OBS builds binary

packages for many
distributions and
platforms

OBS makes them

available for
download

10 SUSE, All rights reserved.

 SUSE Studio

Create it Applikationen

Enter New Markets

Konfigurationen

Lower Costs
Operating System

SUSE Studio lets you rapidly scale up or

down as needed. Assemble, build and
maintain complete portable application
stacks in minutes and deploy wherever
you want. Physical, Virtual, Cloud.

11 SUSE, All rights reserved.

 SUSE Lifecycle Management Server

Partner Actions
Application
Updates
Selection
Testing
Quality Assurance
Middleware
Updates
Versioning
Publishing
Pushing
OS Updates
Unified Update

Customer Actions
Applying (if optional)

SUSE Lifecycle Management Server

streamlines the process of updating and
maintaining software appliances.

12 SUSE, All rights reserved.

13 SUSE, All rights reserved.
 SUSE Linux Enterprise
Server and High
Performance Computing
 Where is HPC used?
Yesterday
Universities
Government
Today
Academia & Research Institutions
Financial Services: Risk Analysis
Oil & Gas: Exploration, Production
Semiconductor: Electronics Design
Automation
Life Sciences: Pharmaceutical and
Biomedical
Manufacturing: Aerospace &
Automotive
Tomorrow
Growing movement towards
departmental and workgroup compute
clusters
High Productivity Computing
15 SUSE, All rights reserved.
 What Operating Systems are Used?
Linux runs on 455
(91%) of the top 500
super computers.*
Linux is approaching
90% share of general
clusters.
Linux is on the majority of HPC systems. This is true
even in smaller departmental implementations all
the way up to integrated cluster solutions.
With the move to commodity hardware use in HPC
clusters, Windows is now being leveraged as well.
5 of the Top10 supercomputers of the world are
running an operating system based on SUSE Linux
Enterprise.
*top500.org 06/2011
16 SUSE, All rights reserved.
 Linux HPC Value
Low Cost
Commodity hardware, open source operating system

Open Source
Very easy to customize, maintain, and improve

HPC Clusters born on Linux

Beowulf

Modular system
GUI overhead not required. Appliance ready form factors

Linux Standards
Huge base of tools, including very powerful remote management
tools
Most all hardware is developed first on Linux
Vendors and eco-system around Linux HPCCs
17 SUSE, All rights reserved.
 HPC
Your Business Issues

How can I:
Solve computational,
data-intensive, or
numerically intensive tasks
at reasonable costs?
Reduce the efforts and
time to set-up and
maintain Linux clusters for
HPC?
Ensure that all
components of my HPC
stack work perfectly
together?
18 SUSE, All rights reserved.
 SUSE Linux Enterprise Server
Overview

Highly reliable, interoperable and manageable

server operating system
Built to power mission-critical workloads in physical
and virtual environments
The natural successor to UNIX, backed by proven
services for UNIX migration
The only Linux recommended by Microsoft and SAP
Complemented by product extensions that deliver
advanced capabilities

19 SUSE, All rights reserved.

 HPC Feature Highlights in SLES
Advanced Input/Output Processing
Asynchronous I/O (AIO)
A form of input/output processing that permits other
processing to continue before the transmission has finished
Modular I/O Scheduler
Algorithm most suitable for workload can be chosen
dynamically
Multi-core/hyper-threading processor support
Execute threads in parallel within each individual processor
Intel I/O Acceleration
Offloads the CPU towards the network card, thus allowing
the system to continue processing data while I/O is taking
place

20 SUSE, All rights reserved.

 HPC Feature Highlights in SLES
Flexible and Pluggable I/O Scheduler
CFQ
Complete Fair Queuing, suitable for a wide
variety of applications, default I/O scheduler
Deadline I/O scheduler
A valuable choice on a server with intelligent
hardware, or if latency is the primary goal
Anticipatory I/O scheduler
Suitable for file servers but does not work as well
for database workloads
No-op
Better suited for systems with storage solutions
that have their own cache and their own
virtualization Modular I/O Schedule

21 SUSE, All rights reserved.

 HPC Feature Highlights in SLES
Non-Uniform Memory Architecture
Non-Uniform Memory Access or Non-Uniform
Memory Architecture (NUMA)
A computer memory design used in multiprocessors,
where the memory access time depends on the
memory location relative to a processor
Under NUMA, a processor can access its own local
memory faster than non-local memory, that is,
memory local to another processor or memory
shared between processors
NUMA provides excellent scalability and performance
for Intel Itanium and AMD Opteron architectures

22 SUSE, All rights reserved.

 HPC Feature Hightlights in SLES
More Features
Enhanced and tuned 2.6 series Linux Kernel for optimal
performance
CPU Management and System Activity
CPUset System, CPUset command line tool
Sysstat package
IRQbalance
OpenFabrics Enterprise Distribution (OFED)
Remote Direct Memory Access (RDMA) switched fabric
technologies, high-speed data transport technologies for server
and storage connectivity
SystemTap

23 SUSE, All rights reserved.

 SUSE Linux Enterprise 11 Service Pack 1

Scalability
The only enterprise Linux operating system that helps customers
Manage several workloads within one Linux instance in a lightweight
manner by providing Kernel Resource Management with Control Groups
Scale with their hardware by supporting 4096 CPUs on Intel64,1024 CPUs
on POWER, 4096 CPUs on Itanium
Compute huge amounts of data in memory, e.g. in data- warehouse and
ERP systems, by supporting 16TiB RAM
(and beyond) on certified hardware
Deploy huge amounts of data by supporting SGI's XFS for
filesystem and file sizes up to 8 EiB in the 4th generation of the OS
Improve efficiency by leveraging HW support for power saving features
due to the tick less idle Kernel, i.e. individual cores can be sent to
sleep completely

24 SUSE, All rights reserved.

 SUSE Linux Enterprise

Scalability Tomorrow
SUSE Linux Enterprise 11 Service Pack 2:
Container Technology (based on lxc) extends Kernel Resource
Management and enables you to create lightweight virtualization-
like separations for better load management and higher security
(soft partitioning)
With the increasing number of cores per system, improvements
in power management and scheduling are key to control costs
in physical and virtualized environments
Local storage: maintain existing capabilities (e.g. XFS) and
introduce btrfs as a supported solution, to improve manageability
and give customers a maximum of flexibility
Expand network filesystem capabilities (NFSv4.x/pNFS), to
improve performance, reliability and security of filesystems
across datacenter networks

25 SUSE, All rights reserved.

 SUSE Linux Enterprise 11 Service Pack 1

Reliability Today
Reliability Availability Serviceability (RAS)
Strong cooperation with IBM on providing a Linux OS optimized for
mission critical workloads on System z
Support hardware RAS features bringing AMD64/Intel64 systems
on par with traditional RISC systems
Large blade centers benefit from swap over NFS capabilities to
centralize swap space and improve availability of the datacenter
Cost savings by using built-in device-mapper MultiPath I/O (MPIO)
replacing expensive commercial solutions.
Increased redundancy through support of RAID 6, RAID 10
Scheduler optimizations and support for new floating point
features improve performance and save costs

26 SUSE, All rights reserved.

 SUSE Linux Enterprise

Reliability Tomorrow
SUSE Linux Enterprise 11 Service Pack 2:
Improve support for hardware based RAS capabilities on
all architectures, specifically
Intel SandyBridge
IBM System z

Support for the btrfs filesystem

Reduce cost of storage management by providing an integration
of logical volume management and filesystem
Checksums on data and metadata ensure data integrity
LTTng ( Linux Trace Toolkit Next Generation )
SUSE Labs supporting LTTng technology upstream
Inclusion of LTTng base functionality in SLE 11 SP2

27 SUSE, All rights reserved.

 SUSE Linux Enterprise Server is a Component of
Intel Cluster Ready Program
The Intel Cluster Ready program is designed
to simplify purchasing, deployment, and
management of HPC clusters.
SUSE participates in the Intel Cluster Ready
program
SUSE Linux Enterprise Server is Intel Cluster Ready
and powers many of the certified Intel Cluster
Ready systems.
There are Intel Cluster Ready recipes with SUSE
Linux Enterprise Server, which are reference
designs provided to help hardware vendors,
platform integrators, and system integrators design
and build certified Intel Cluster Ready systems.
28 SUSE, All rights reserved.
 Top 500 Super Computers
by Operating System Family*
500

450

400

350

300 Linux
Windows
250 Unix
BSD Based
Mixed
200
Mac OS

150

100

50

0
2003 2004 2005 2006 2007 2008 2009 2010

Linux is increasing from 27% to 91%

UNIX is decreasing from 70% to 4.4%

29
* Numbers are based on the ISC (June) list of each year
SUSE, All rights reserved.
 Linux Operating Systems
in Top 500 Super Computers
Linux (unspecified)
SUSE (incl. CNL)
Red Hat (incl. CentOS)

500

450

400

350

300

250

200

150

100

50

0
2005 2006 2007 2008 2009 2010

SUSE Linux Enterprise is the strongest force *

*Numbers indicate minimum quantity of systems as most systems are

unspecified Linux, the actual figure might be much higher

30 SUSE, All rights reserved.

 Your Objection
We are Happy with Our Existing HPC Solution
Ask yourself:
Is your solution a maintained (Linux) environment?
Whom do you contact in case of a support incident?
Are your applications certified for the underlying
operating system?
Does your solution offer you the best TCO available
in the market today?
Are you independent of your hardware provider?

If you answer one question with NO, it is time to

consider SUSE Linux Enterprise Server for HPC!

http://www.novell.com/hpc

31 SUSE, All rights reserved.

 Customer ChallengesOur Solution
The performance of my systems needs to be increased while
IT budgets and staff are being reduced.
Choose SUSE Linux Enterprise Server for HPC. Systems and
clusters are easy to set-up/maintain and offer better TCO.
HPC is business critical, but in my company it managed by a
separate department with a scientific flavour.
HPC is an integral component of today's enterprise computing
infrastructure. A maintained system such as SUSE Linux Enterprise
Server gives you access to updates and support.
With my current systems I am dependent on one
hardware vendor.
Linux clusters built on SUSE Linux Enterprise Server offer hardware
independence.
There is no support for the full stack available.
SUSE partners closely with all stack component vendors

32 SUSE, All rights reserved.

 Super Computing
SUSE Linux Enterprise

SUSE Linux Enterprise based operating systems are

used on six of the world's top 10 super computers*
OS is used in integrated solutions from
SGI/LinuxNetworx, Cluster Resources, HP, IBM and
Cray
Large installed base in Research, Semiconductor,
Aerospace, Oil and Gas:
Examples:
Oakridge National Labs
NASA Ames Research Center
BMW
Total Exploration
CESGA
Top 10 Super Computers

*top500.org 11/2008

33 SUSE, All rights reserved.

 Ecosystem
IT Consulting
A team of highly skilled consultants, along with our
partners, is specialized in SUSE Enterprise Linux
technologies.
With real-world knowledge of how our products work in
production data centers, our IT Consulting helps
customers design and implement the best possible
data center solutions.
Complex data center solutions require skills and
expertise; no one has more experience in deploying
our products than SUSE and its partners.

34 SUSE, All rights reserved.

 Ecosystem
Technical Support
SUSE's Technical Support is a world-class support
organization, providing customers with 9 global
support centers and an expansive field team around
the world.
Your SUSE Linux Enterprise Server systems are
backed by
More than 25 years experience supporting global enterprises
Industry leading support for heterogeneous environments
Industry and customer recognition for quality

35 SUSE, All rights reserved.

 Value Proposition
Linux is the basis of approximately 80 percent of the HPC market.
Linux is the driver for server consolidation and simultaneous
implementation of DB and business application (e.g. SAP) clusters

To help customers reduce the complexity and risk associated with buying
an HPC cluster solution, SUSE participates in the Intel Cluster Ready program.
Developed in conjunction with HW and SW vendors, the Intel Cluster Ready
program is designed to simplify purchasing, deployment, and management
of HPC clusters. Rely on SUSE Linux Enterprise Server powering
many of the certified Intel Cluster Ready systems.

SLES stands for proven performance benefits and real-time capabilities. It

supports theoretically unlimited node clusters, and comes with cluster
management software options. It is the only real cross platform solution for
HPC. It is certified for many major clustering software products as the
underlying operating system. Certification and maintenance is the basis for
support. There is a HPC specific pricing model for SUSE Linux Enterprise
Server via Partners.

36 SUSE, All rights reserved.

 Next Steps

1 2 3
Learn Download Talk
more evaluation to us

www.suse.com/products/server/hpc.html www.suse.com/products/server/eval.html
www.suse.com/products/realtime/eval.html

37 SUSE, All rights reserved.

SUSE LINUX Enterprise
Technology Update
 The SUSE Build Service* Advantage

Open Source Package Build Service Quality Testing Enterprise-

Projects Selection and Class
Integration x86 Softw are
Feature
Test
x86_64
Manual
Regression
Itanium
SUSE Linux
Automated
Enterprise
Regression
POWER
Novell

System z System Test

Contribution

Reduces production problems

Consolidates IT skills across disparate systems
Delivers critical updates in hours not days or weeks

* SUSE Build Service is the internal entity of the open(SUSE ) BuildService infrastructure

40 SUSE, All rights reserved.

 SUSE Linux Enterprise

Common Code Base

SUSE Linux Enterprise platform

JeOS &
SLES SLED SLE SDK SLE HA
Appliances

Binary Code Base

Common (Open Source) Code Base

Foundation for SUSE Linux Enterprise products

Fully supported core system: L3 support
commitment

41 SUSE, All rights reserved.

 SUSE Linux Enterprise
Standard Platform Lifecycle
General Support Extended Support

Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 9 Year 10

GA

SP1 Long Term Service Pack Support

SP2 Long Term Service Pack Support

SP3 Long Term Service Pack Support

SP4 Long Term Service Pack Support

10-year lifecycle (7 years General Support, 3 years Extended

Support)
Major releases every ~4years, service packs every 16-18 months
Six month upgrade window
Long Term Service Pack Support extend upgrade window or
extend major release lifecycle

42 SUSE, All rights reserved.

 SUSE Linux Enterprise
Current Platform Lifecycle

2009 2010 2011 2012 2013 2014 Notes

CR: SLE 9 SP4 (2007)

SLE 9 EOGS: Q3 2011
EOSS: Q3 2014

CR: SLE 10 SP3 (2009)

SLE10 x
SP3 SP4
x EOGS: Q3 2013
EOSS: Q3 2016

CR: SLE 11 SP1 (2010)

SLE 11 GA x
SP1 x
SP2 SP3 EOGS: Q2 2016
EOSS: Q2 2019

SLE 12 GA

SUSE announces Service Pack releases and development

and product schedules to customers and partners
Dependable release timing
Predictability for planning rollouts and migrations

43 SUSE, All rights reserved.

 SUSE Linux Enterprise 11
Maintenance Delivery Evolution

SLES 9 GA SP1 SP2 SP3 Flexibility

SLES 10 GA Control

SP1

SP2

SP3
6 months overlap

Flexibility +
SLES 11 SP1 General Updates (GU)
Control
SP2
SP3

44 SUSE, All rights reserved.

 SUSE Linux Enterprise 11 post Service Pack 1

Enhancements via Maintenance

Common Code Base:
Firefox
SUSE Manager integration
Ruby on Rails 2.3.8
PHP 5.2.14
Unattended Upgrade from SLE 10 to SLE 11
Intel Sandybridge enablement (see next slide)

Server
OFED 1.5.2
PureFTPd (Netware compatibility)

Desktop
LibreOffice

45 SUSE, All rights reserved.

 SUSE Linux Enterprise 11 post Service Pack 1

Intel Sandybridge enablement

SLE 11 SP1 (Common Code Base)
PCI IDs and other minor enhancements delivered via Kernel
maintenance update in March 2011
Challenge: new drivers required for recent machines
(Patsburg chipset)
Bootstrapping
kISO (bootable ISO with recent kernel) via PLDP
http://drivers.suse.com/novell/SLE-SANDYBRIDGE/sle11-
sp1/install-readme.html
Graphics stack (X11)
intrusive patches
not suitable for general availability
only used for controlled environments (desktop preloads)
Full enablement for Server and Desktop for SLE 11 SP2

46 SUSE, All rights reserved.

 SUSE Linux Enterprise High Availability Extension

Enhancements via Maintenance

Load-Balancing
Support for IPv6 based networks

Partnering
Cluster Health Check Promotion with Linbit
Formation of HA working group at Linux Foundation

Tools
Granular Access Control Lists

Cluster Communication
Unicast support

Fencing
Support for multiple SBD devices

47 SUSE, All rights reserved.

 SUSE Linux Enterprise 11 SP2
Hardware enablement / incl. RAS features
Compete with Solaris
Filesystem level: btrfs with Copy on Write, data
checksums and snapshotting
LXC delivers container support based on Control
Groups
Enhance tracing capabilities, by delivering LTTng /
Linux Trace Toolkit (separate Extension)
Snapshot / Rollback for package updates:
YaST2 + ZYPP + btrfs
New Maintenance Model:
more flexibility for customers while retaining full control
SLE HA: Geo-Cluster, automated and pre-configuration

Expected: Q1 2012
48 SUSE, All rights reserved.
 SUSE Linux Enterprise 11 Service Pack 1

Filesystems
Feature Ext 3 reiserfs XFS OCFS 2 btrfs
Data/Metadata Journaling / / / / N/A [3]
Journal internal/external / / / / N/A
Offline extend/shrink / / / / /
Online extend/shrink / / / / /
Inode-Allocation-Map table u. B*-tree B+-tree table B-tree
Sparse Files
Tail Packing
Defrag
ExtAttr / ACLs / / / / /
Quotas
Dump/Restore
Blocksize default 4KiB
max. Filesystemsize [1] 16 TiB 16 TiB 8 EiB 4 PiB 16 EiB
max. Filesize [1] 2 TiB 1 EiB 8 EiB 4 PiB 16 EiB
Support Status SLES SLES SLES SLE HA Technology
Preview
SUSE Linux Enterprise was the first enterprise Linux distribution to support journaling filesystems and logical volume managers back in 2000. Today, we have customers running XFS and ReiserFS with more than
8TiB in one filesystem, and the SUSE Linux Enterprise engineering team is using our 3 major Linux journaling filesystems for all their servers. We are excited to add the OCFS2 cluster filesystem to the range of
supported filesystems in SUSE Linux Enterprise. For large-scale filesystems, for example for file serving (e.g., with with Samba, NFS, etc.), we recommend using XFS. (In this table "+" means "available/supported"; "-"
is "unsupported")
[1] The maximum file size above can be larger than the filesystem's actual size due to usage of sparse blocks. It should also be noted that unless a filesystem comes with large file support (LFS), the maximum file
size on a 32-bit system is 2 GB (2 31 bytes). Currently all of our standard filesystems (including ext3 and ReiserFS) have LFS, which gives a maximum file size of 2 63 bytes in theory. The numbers given in the
above tables assume that the filesystems are using 4 KiB block size. When using different block sizes, the results are different, but 4 KiB reflects the most common standard.
[2] 1024 Bytes = 1 KiB; 1024 KiB = 1 MiB; 1024 MiB = 1 GiB; 1024 GiB = 1 TiB; 1024 TiB = 1 PiB; 1024 PiB = 1 EiB (see also http://physics.nist.gov/cuu/Units/binary.html )
[3] Btrfs is a copy-on-write logging-style file system, so rather than needing to journal changes before writing them in-place, it writes them in a new location, and then links it in. Until the last write,
the new changes are not committed.
[4] Btrfs quotas will operate differently than traditional quotas. The quotas will be per-subvolume rather than operating on the entire filesystem at the user/group level. They can be made
functionally equivalent by creating a subvolume per- user or group.

49 SUSE, All rights reserved.

 Corporate Headquarters +49 911 740 53 0 (Worldwide) Join us on:
Maxfeldstrasse 5 +www.suse.com www.opensuse.org
90409 Nuremberg
Germany

50 SUSE, All rights reserved.

Unpublished Work of SUSE. All Rights Reserved.
This work is an unpublished work and contains confidential, proprietary and trade secret information of SUSE.
Access to this work is restricted to SUSE employees who have a need to know to perform tasks within the scope of
their assignments. No part of this work may be practiced, performed, copied, distributed, revised, modified, translated,
abridged, condensed, expanded, collected, or adapted without the prior written consent of SUSE. Any use or
exploitation of this work without authorization could subject the perpetrator to criminal and civil liability.

General Disclaimer
This document is not to be construed as a promise by any participating company to develop, deliver, or market a
product. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making
purchasing decisions. SUSE makes no representations or warranties with respect to the contents of this document,
and specifically disclaims any express or implied warranties of merchantability or fitness for any particular purpose.
The development, release, and timing of features or functionality described for SUSE products remains at the sole
discretion of SUSE. Further, SUSE reserves the right to revise this document and to make changes to its content, at
any time, without obligation to notify any person or entity of such revisions or changes. All SUSE marks referenced in
this presentation are trademarks or registered trademarks of Novell, Inc. in the United States and other countries.
All third-party trademarks are the property of their respective owners.