© 2010 IBM Corporation IBM Power Systems Power your planet. Smarter Systems for a Smarter Planet...

© 2010 IBM Corporation

IBM Power Systems

Power your planet.Smarter Systems for a Smarter Planet

IBM iDoctor for i OverviewDawn [email protected]


IBM Power Systems

3

IBM iDoctor for i

Job Watcher Disk WatcherJob Watcher Disk Watcher PEX AnalyzerPEX AnalyzerCollection Services InvestigatorCollection Services Investigator Heap AnalyzerHeap Analyzer

Tools and services for

performance investigation


IBM Power Systems

5

iDoctor for i

Introduction•Overview and a brief history•Where these tools fit

The iDoctor GUI•Unique Features

Components•Collection Services Investigator•Disk Watcher•Job Watcher•PEX-Analyzer

the presentation flow


IBM Power Systems

7

iDoctor for i - Introduction

Overview

iDoctor is suite of dynamic tools developed by the Rochester Support Center used to collect, investigate and analyze performance data on System i. The tools are used to monitor overall system health at a high

"overview" level or to drill down to the performance details within job(s), disk unit(s) and/or programs over data collected during performance situations.

Frequently used by IBM, along with customers and consultants to help solve complex performance issues quickly.

Feedback from iDoctor users continues to help shape the enhancements incorporated into new iDoctor components and features.


IBM Power Systems

PRODUCTIVITY

Broaden the user base for Performance Investigation enable Operators, Programmers, IS Management as well as Performance Specialists, Consultants

Simplify and automate processes

Provide quick, immediate access to collected data

Provide more analysis options

Reduce the dependency on PEX traces

Goals



IBM Power Systems

11

a historical progression - the tools developed and used by service

SMTRACE TPST SAMI M P I

PEX Service Analysis ToolsPrograms - Commands - Queries

Instrumentation

TRACE STATS PROFILE

iDoctor for iSeriesPEX Analyzer - Job Watcher - Heap Analysis Tools

SimplifiedCollection

Data ViewerData Grapher

EnhancedAnalysis

2000 ------- 1999

IBM iDoctor for IBM iPEX Analyzer - Job Watcher - Collection Services

Investigator - Disk Watcher - Heap Analyzer2008 ------- 2007



IBM Power Systems

13

Level of detail by performance tool


Function Performance Tool

Capacity planning PM for Power Systems

IBM Workload Estimator

High-level system/job monitoring

iDoctor – Collection Services Investigator

IBM i Performance Tools (PT1)

Management Central Monitors

Medium-level system/job monitoring

iDoctor – Job Watcher

WRKSYSACT

IBM i Job Watcher (PT1) / STRJW

Low-level system/job tracing, stats, profiling

iDoctor – PEX Analyzer

IBM i Performance EXplorer / PRTPEXRPT

Disk stats/tracing iDoctor – Disk Watcher

iDoctor – Collection Services Investigator

IBM i Disk Watcher (PT1) / STRDW

“Classic” JVM Heap Analysis

iDoctor – Heap Analyzer

DMPJVM


IBM Power Systems

iDoctor for i – The iDoctor GUI

iDoctor vs. 6.1 Performance Tools LPP (Performance Data Investigator)

Windows Client

iDoctor GUI

Fee

Frequent updates

Prior release support (5.3-5.4)

iDoctor

Web Browser

Systems Director Navigator

Fee/Free

Updated once per release

Requires conversion

6.1 Performance Tools

Both GUIs use the same OS commands and produce the same raw data.

Different approaches applied to summarization, analysis techniques


IBM Power Systems

17


Windows client offers flexibility and functionality not yet in web version

All components offer a similar user experience

The GUI provides access for iDoctor components installed on servers running IBM i V5R3 or higher.

Requirements:System i access for WindowsLicense keys for Job Watcher (includes DW, CSI) and PEX Analyzer component usage.

Overview


IBM Power Systems

19

iDoctor for i – New Collection Services Investigator Features for 7.1

Graphs added for External Storage DS8K/DS6K statistics (cache, links, ranks)

Disk graphs for read/write response time buckets.

Launch Workload Estimator option for upgrade sizings

Collection Search option (job name, user, etc)

Situational Analysis (7 situations defined)

Hypervisor data graphs (entire system/Virtual Real Memory)


IBM Power Systems

21

iDoctor for i – New Job Watcher Features for 7.1

Collection search over multiple collections

Create job summary function

Configurable thresholds in Situational Analysis

Concurrent write support and journal cache situations added to Situational Analysis

STRIDRSUM command for batch summarizations For all collections in a library

J9 JVM (heap) graphs


IBM Power Systems

23

iDoctor for i – New PEX Analyzer Features for 7.1

All analyses rewritten to SQL stored procedures.

Physical disk I/Os analysis offers detailed disk analysis with by object, by unit, ASP, path, etc.

New call stacks, hot sectors and save/restore analyses added.

Collection properties – System tab has new info from QAYPERUNI file.

PEX monitors (24x7 collection but retaining only the most recent set of data)


IBM Power Systems

25


Update History Automatic checking for new PTFs / client updates SQL Editor Monitors Graph compare modeTime range toggling Alternate views Dynamic graph legend Super CollectionsSituational Analysis

Unique Features


IBM Power Systems

27


Update History Window

Keeps you informed of the latest updates and enhancements


IBM Power Systems

The iDoctor client provides automatic checks for new builds Checking performed at startup and can to be turned off in Preferences. PC firewall must allow iDoctor.exe access to the Internet.


Client update notification


IBM Power Systems

The iDoctor client checks for new Job Watcher and PEX PTFs New PTF numbers identified on the iDoctor website are sent to the PC. PC firewall must allow iDoctor.exe access to the Internet.


Automatic PTF notification


IBM Power Systems

30


SQL Editor

Available above every graph/table in iDoctorEdit SQL/rerun reports as desired


IBM Power Systems

32


Monitors

Allows 24x7 collection of Job Watcher, PEX Analyzer or Disk Watcher data.

Retain only the desired amount of historical data.When a problem occurs you will (hopefully) have the data you need to solve it.


IBM Power Systems

Dashed lines indicate a time break/different collection.


Graphing multiple collections


IBM Power Systems

Compare ASPs, disk units, jobs and more.


Graph compare mode


IBM Power Systems

Allows for either broad or detailed views of time based graphs

Time range interval toggle (clock icon) provides customization of interval size

Bar selection allows for easier drill down into desired time periods

Graph compare mode allows different time groupings to be analyzed.

These features apply to all time-based graphs in iDoctor.


Time range graphs/toggle


IBM Power Systems


Time range graphs toggle example


IBM Power Systems

Allows quick comparisons between different graphs Utilizes the report data already stored in memory to give instant

display Button shown below enables this option


Alternate views


IBM Power Systems

iDoctor for i – The iDoctor GUI Alternate Views Example


IBM Power Systems

45


Dynamic graph legend

Graph legend allows drag and drop or context menu usage to customize the data shown on the graph


IBM Power Systems

47


Super Collections

Provides the capability to capture multiple types of collections simultaneously.

Job Watcher is always collected in a Super Collection while CS, PEX and/or Disk Watcher are optional.


IBM Power Systems

49


Situational Analysis

• Identifies any potential problem jobs found in the collection.

•Built on the vast knowledge of IBM performance experts and past experiences.

• Included with Collection Services Investigator and Job Watcher.

•Different background colors/flyovers on the overview graphs indicate jobs of interest for occurring time periods.


IBM Power Systems


Situational Analysis


IBM Power Systems

52


Synchronized Table View for Graphs

Provides easy access to the “real data” behind the graph with synchronized scrolling and selection.


IBM Power Systems

54

iDoctor for i – Collection Services Investigator

Overview

General Information

Using Collection Services Investigator For wait analysis For system monitoring


IBM Power Systems

GUI client provides dozens of graphs for the high level job based data collected by Collection Services. Data is typically collected 24x7 at 15 minute intervals. Metrics shown include waits, I/Os, CPU, IFS and more.

Works against the Collection Services database files CRTPFRDTA must be ran to produce the database files (GUI option for that also available). Typically 1 collection = 1 24 hour period.

Included with a Job Watcher license.


General Information


IBM Power Systems

Graph Types Available Collection wide by time interval

By default these offer drill down into Rankings graphs for the selected time period. Rankings

Thread, job, job user, generic job, current user, pool, priority, subsystem. By default these offer drill down into the next graph type for the selected objects.

Selected “objects” by time interval Objects = thread, job, generic job, ASP, disk pool, etc.


Using Collection Services Investigator


IBM Power Systems

Wait Analysis to Locate Bottlenecks Provides support to graph waits and CPU utilization with drill down to jobs. At V5R3/V5R4 correctly displays the wait bucket graphs based on the in use mapping for that collection. At V6R1, CS bucket mapping matches the Job Watcher mapping.

Monitor System Performance Use the high level graphs to monitor performance based on the desired metrics CPU, I/Os, transactions, etc.

Using Collection Services Investigator



IBM Power Systems

59


Collection Overview showing object lock contention


IBM Power Systems

This graph is showing time spent in journal, object lock contention.

iDoctor for i – Collection Services InvestigatorCollection Overview “Waits Analysis” Example


IBM Power Systems

Drill Down Example Showing All 3 Graph Types



IBM Power SystemsiDoctor for i – Collection Services InvestigatorCPU consumption by thread example


IBM Power SystemsiDoctor for i – Collection Services Investigator

Physical disk I/O totals example


IBM Power Systems

• Graphical user interface over Disk Watcher datadrill-down capabilitiesstatistical and trace data

• Included with the iDoctor Job Watcher component

• Includes a trace summary function to provide graphing with drill down of I/O rates, average response times, I/O time buckets and more.

iDoctor for i – Disk Watcher

Overview


IBM Power Systems

OS function offering the ability to trace and summarize I/O activity utilizing a wrapping flight recorder.

Provides near real-time diagnosis of disk performance issues. Collects operational details used to determine the hardware, job,

program, and object associated with each I/O. Provides the data needed to analyze disk performance problems

through the use of I/O statistical reports and views. Surfaces data beyond that provided by such tools as:

– WRKDSKSTS– WRKSYSSTS– WRKSYSACT

V5R3M0, V5R3M5, V5R4M0, V5R4M5 available via PTF V6R1M0 available as part of the base OS

What is Disk Watcher?


IBM Power Systems

Evaluate the performance of I/O operations to: – Internal disks– External disks (LUNs)

▫ Multi-path

Evaluate the performance of I/O queuing

Determine how performance may be improved by re-spreading data across units

Determine the optimal placement of devices, IOAs, or buses

With Disk Watcher you can …

Trace every I/O performed

Collect statistical summary data

AND/OR


IBM Power Systems Disk Watcher - Goals

Reduce disk analysis complexity

Provide quick, immediate access to collected data– Near ‘real time’ analysis

Provide for “after the fact” analysis of the collected data

Reduce the dependency on PEX traces

Provide conditional data collection – To limit the size of a collection


IBM Power Systems

Trace– TDE resolution– Program resolution– Object resolution– Conditional data collection

Stats (summarization)– Fixed interval time or dynamic

Disk Watcher – Collection Options


IBM Power Systems

Disk Watcher – A user can define:

What to collect– Disk unit, ASP, Pool, I/O type– Resolution level (TDE, program, object)

How long to collect– Time, size, intervals

Conditional data collection– Service time, deferred queue time, response time


IBM Power Systems

Dumping should lead new entries. If new entries overtake the dumping or if the selected time interval is too long then we wrap the flight recorder and have missed data.

Dynamically start dumping data when the flight recorder is 80% full

Disk Watcher – Wrapping Flight Recorder

Or select a specific time interval


IBM Power Systems

Launching iDoctor Disk Watcher


IBM Power Systems

Disk Watcher – Select and Launch


IBM Power Systems

Disk Watcher - Start


IBM Power Systems

Disk Watcher - Start Wizard will guide you


IBM Power Systems

Basic– Collection name and location– Collection type (Summary/Trace/Both)– Variable/dynamic interval length

Advanced– Data availability– ASP threshold overrides (system/user)– Optional hardware information

Disk Selection– Storage pools (ASPs)– Disk units– Memory pools

Termination– Disk space– Time – Intervals

Disk Watcher – Collection Options


IBM Power Systems

Disk Watcher – Graphing


IBM Power Systems

Disk Watcher – Graph Selections

Trace graph choices

Statistical graph choices


IBM Power Systems


I/O counts and times


IBM Power Systems


Operation/second with response time by disk unit


IBM Power Systems

Disk Watcher – Report Selections

Graphs can be used to drill down for additional report information


IBM Power Systems

Disk Watcher – Data files Related data files are easily accessed

Custom queries and graphing


IBM Power Systems

83

iDoctor for i

Job Watcher

provides real-time, non-intrusive, detailed and summarized views of job/thread/task performance data - some of which, until now, were unavailable on a System i server

it's the step to take to help avoid a system wide trace or to help ensure that a trace will yield useful data

it's a super WRKSYSACT that displays both "running" and "waiting" components for a job

includes the Collection Services Investigator and Disk Watcher analysis tools

summary


IBM Power Systems

85

iDoctor for i

PEX Analyzer

simplifies the collection and enhances the analysis of PROFILE, STATS and Trace data

provides the details necessary for the low-level analysis of processor utilization, DASD operations, file space usage, waits, file opens and much more

it's the system performance "X-ray"

summary


IBM Power Systems

87

iDoctor for i - Website

http://www-912.ibm.com/i_dir/iDoctor.nsf

Documentation

Downloads

FAQ

Consulting Services

Education


IBM Power Systems

88

The Redbook was published in early 2005 but still can be very useful in understanding the fundamentals of Job Watcher.http://www.redbooks.ibm.com/abstracts/sg246474.html?Open


IBM Power Systems

89

PEX Stats Redbookhttp://www.redbooks.ibm.com/abstracts/sg247457.html?Open


IBM Power Systems

91

iDoctor for i

PEX Analyzer PROFILE - program instructions STATS - program calls TRACE - timed events

Job Watcher snapshot current status snapshot deltas

the tandem used to expose run/wait components

batch run time

running waiting

Job/Thread/Task Run/Wait Components

idleblocked

transaction response time

DASDCPUq


IBM Power Systems

93

iDoctor for i

PEX (Performance Explorer) - Trace (Method: record all events) (PRO: exact timing, complete event detail) (CON: takes time and resource to dump and analyze)

- STATS (Method: summarize all calls) (PRO: details the time spent in program modules) (CON: overhead adds/alters code path length of what your investigating

Job Watcher (Method: record snapshots, compute interval deltas) (PRO: quick, non-intrusive and relatively cheap to collect data) (CON: a bit sketchy, could miss jobs initiating/terminating within intervals)

JOBxxx

snapshot snapshot

intervaldelta

some PEX Analyzer and Job Watcher comparisons


IBM Power Systems

95

iDoctor for i

the challenge

trace data is preferred because of the detail

but a trace takes time to dump and analyze

and busy, larger, faster systems can generate even more trace data within shorter trace windows

and often, until a trace is dumped and analyzed, you

don't know if the trace captured what you needed

why Job Watcher was developed


IBM Power Systems

97

iDoctor for i

Job Watcher as a trace alternative

its rapid snapshot, interval analysis, and data harvesting options can reduce the need for trace as a first step - or ensure that a trace underway has captured worthwhile data

it's a non-intrusive collector (like WRKSYSACT)

its trigger wizard allows the setup of "and/or" rules based on the data collected which can be used to initiate a call to a user program

why Job Watcher was developed...


IBM Power Systems

99

iDoctor for i

Data Harvested

most of the counters captured by WRKSYSACT, WRKACTJOB, and Collection Services for Jobs, including threads and tasks

the program stack (50 deep, 1000 starting in V5R3), activation group data, active SQL data, wait buckets, Java statistics

the current wait, its duration, the object waited on, the holder if available

the Job Watcher snapshot


IBM Power Systems

101

iDoctor for i

Same Interval Analysis as Collection Services

cumulative counter deltas

Plus wait status at snapshot

Job Watcher factors the current wait duration into the interval statistics

intervals as small as 100 msecs are possible and may provide enough information to avoid a trace

the Job Watcher snapshot interval


IBM Power Systems

103

iDoctor for i

JOBxxxx

snapshot intervalsnapshot #1 snapshot #2

otherwaits cpu current wait

the Job Watcher snapshot interval and the current wait


IBM Power Systems

105

iDoctor for i

snapshot #1 snapshot #2

Eye Wait Wait WaitENUM Catcher Bucket Description Object Holder

158 SRd 10 DASD Read Request XYZfile na

7 QTQ 27 TreeQueue XYZqueue na

100 Rex 11 Seize Read exclusive XYZindex XYZjob

current wait

current wait examples


IBM Power Systems

107

iDoctor for i


• OK if idle wait (like Key/Think time)• But if a long DASD request or a block? (and the

current wait has not ended yet)

current wait

current wait can span snapshot intervals


IBM Power Systems

109

JW – the snapshot approach


Job 1000

Job 0001t i m e

JW snapshot analysis begins

(satisfied rule can call user program)

(Interval statistics written to DB)

snapshot #3

data is

available for

viewing/analysis


IBM Power Systems

111

JW – the run/wait summary for a snapshot interval

running waiting

01 02 … 05 06 … 16 17 18 … 32

wait buckets

CPUq Syn DASD LockCPU

Job

the system maintains counts and accumulated time for all jobs/threads/tasks


IBM Power Systems

113

JW – the last current wait at snapshot time

Job

• ~1000 instrumented wait points

• ~200 unique wait identifiers (ENUMs)

• mapped to ~30 wait buckets

• the unique wait identifier (ENUM)

• the wait time (so far)

• the object waited on

• the holder of the object (when available)

Long ”bad waits” can be spotted right here (blocks, contention)


IBM Power Systems

115

utilization10090% -80% -70% -60% -50% -40% -30% -20% -10% - t i m e

avg Tnx rsp- 2.50- 2.25- 2.00- 1.75- 1.50- 1.25- 1.00- 0.75- 0.50- 0.25

Collected Data - Intervalized, Summarized, Averaged, Expanded

t i m e

utilization

40% -30% -20% -10% -

avg Tnx rsp

- 1.00- 0.75- 0.50- 0.25


IBM Power Systems

117

iDoctor for i

long intervals less recording overhead and space required depend on averages, rates and percentages and the changes in same (bursts may not be apparent)

good for strategic planning - capacity and operations - for identifying when peak loading occurs

JOBxxxx

snapshot intervalsnapshot #1 snapshot #2

current waitcpuotherwaits

interval duration - the pros and cons of long intervals


IBM Power Systems

119

iDoctor for i

snapshot snapshot snapshot snapshot

short intervals more recording overhead and file space required bursts become more apparent more frequent current wait and stack info (avoid traces) good for performance investigation

JOBxxxx

reported current waits

1 2 3

4

otherwaits

otherwaits

cpu cpu current wait

interval duration - the pros and cons of short intervals


IBM Power Systems

121

iDoctor for i

Starting a watch

select active jobs / threads / tasks (even job queue entries) from a GUI list

set interval duration, watch duration, watch name, library and description, and what details to collect

optionally set a trigger to control when the watch should start or what program to call if specified conditions are met during a watch (a trigger wizard is provided.)

the "watch"


IBM Power Systems

123

iDoctor for i

IBM i 6.1 include the Job Watcher Performance Data Collector

Commands to define Job Watcher collections and to start and end Job Watcher are now included with 6.1 ADDJWDFN, RMVJWDFN STRJW, ENDJW

Job Watcher data can be collected without the iDoctor product

Job Watcher GUI included with the IBM i Performance Tools LPP (5761-PT1, Option 3)

the "watch"


IBM Power Systems

125

iDoctor for i

Summary of all intervals (% of time by component)

The Run/Wait Signature (% of time by component by interval)

JOBxxxx

JOBxxxx

cpuDASDwaits

blockwaits

idle wait

viewing the "watch" - during the watch or after a completed watch


IBM Power Systems

127


IBM Power Systems

129

iDoctor for i

Drill for details about the watch

JOBxxxx

current currentinterval current wait wait cpu

interval usecs wait usecs % interval % interval ...........12345

(select fields, sort columns, query records and more)

cpuDASDwaits

blockwaits

idle wait

viewing the "watch" - during the watch or after a completed watch


IBM Power Systems

131

iDoctor for i – Job Watcher

JOBxxxx

Drill for details about an interval

Program Stack Run/Wait Component Counts and Times Current Wait, Wait Object, Holder Basic Statistics, Transitions, etc. SQL Activation Group Comm Java

viewing the “collection" - during or after completion


IBM Power Systems

133

iDoctor for i

...PEX Analyzer events could show (if collected)

every physical DASD request, its size and duration, the IOP and unit, and the object name, type and address

file opens and closes

MI entry and exit

and more if required

PEX Analyzer surfaces even more detail


IBM Power Systems

135

iDoctor for i


IBM Power Systems

137

iDoctor for i


IBM Power Systems

139

iDoctor for i


IBM Power Systems

141

iDoctor for i


IBM Power Systems

143

iDoctor for i

%Time Representation

Job Signature

Tnx Signature

running waiting

CPU CPUq DASD Blocks COMM Other Idle

CPU CPUq DASD Blocks COMM Other

more about Run/Wait Components


IBM Power Systems

145

iDoctor for i

Job Watcher snapshots and snapshot intervals groups waits into wait buckets accurate total times and counts non-intrusive to the Job's code path length quick "real time" and post collection investigation possible

PEX Analyzer time stamped task switch type events each individual wait exact timing of every individual wait adds a little to the Job's code path length must dump and analyze the trace data

both Job Watcher and PEX Analyzer expose Run/Wait components


IBM Power Systems

147

iDoctor for i

...PEX Analyzer events could show (if collected)

every physical DASD request, its size and duration, the IOP and unit, and the object name, type and address

file opens and closes

MI entry and exit

and more if required

PEX Analyzer surfaces even more detail


IBM Power Systems

149

This document was developed for IBM offerings in the United States as of the date of publication. IBM may not make these offerings available in other countries, and the information is subject to change without notice. Consult your local IBM business contact for information on the IBM offerings available in your area.

Information in this document concerning non-IBM products was obtained from the suppliers of these products or other public sources. Questions on the capabilities of non-IBM products should be addressed to the suppliers of those products.

IBM may have patents or pending patent applications covering subject matter in this document. The furnishing of this document does not give you any license to these patents. Send license inquires, in writing, to IBM Director of Licensing, IBM Corporation, New Castle Drive, Armonk, NY 10504-1785 USA.

All statements regarding IBM future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only.

The information contained in this document has not been submitted to any formal IBM test and is provided "AS IS" with no warranties or guarantees either expressed or implied.

All examples cited or described in this document are presented as illustrations of the manner in which some IBM products can be used and the results that may be achieved. Actual environmental costs and performance characteristics will vary depending on individual client configurations and conditions.

IBM Global Financing offerings are provided through IBM Credit Corporation in the United States and other IBM subsidiaries and divisions worldwide to qualified commercial and government clients. Rates are based on a client's credit rating, financing terms, offering type, equipment type and options, and may vary by country. Other restrictions may apply. Rates and offerings are subject to change, extension or withdrawal without notice.

IBM is not responsible for printing errors in this document that result in pricing or information inaccuracies.

All prices shown are IBM's United States suggested list prices and are subject to change without notice; reseller prices may vary.

IBM hardware products are manufactured from new parts, or new and serviceable used parts. Regardless, our warranty terms apply.

Any performance data contained in this document was determined in a controlled environment. Actual results may vary significantly and are dependent on many factors including system hardware configuration and software design and configuration. Some measurements quoted in this document may have been made on development-level systems. There is no guarantee these measurements will be the same on generally-available systems. Some measurements quoted in this document may have been estimated through extrapolation. Users of this document should verify the applicable data for their specific environment.

Revised September 26, 2006

Special notices


IBM Power Systems

150

IBM, the IBM logo, ibm.com AIX, AIX (logo), AIX 6 (logo), AS/400, Active Memory, BladeCenter, Blue Gene, CacheFlow, ClusterProven, DB2, ESCON, i5/OS, i5/OS (logo), IBM Business Partner (logo), IntelliStation, LoadLeveler, Lotus, Lotus Notes, Notes, Operating System/400, OS/400, PartnerLink, PartnerWorld, PowerPC, pSeries, Rational, RISC System/6000, RS/6000, THINK, Tivoli, Tivoli (logo), Tivoli Management Environment, WebSphere, xSeries, z/OS, zSeries, AIX 5L, Chiphopper, Chipkill, Cloudscape, DB2 Universal Database, DS4000, DS6000, DS8000, EnergyScale, Enterprise Workload Manager, General Purpose File System, , GPFS, HACMP, HACMP/6000, HASM, IBM Systems Director Active Energy Manager, iSeries, Micro-Partitioning, POWER, PowerExecutive, PowerVM, PowerVM (logo), PowerHA, Power Architecture, Power Everywhere, Power Family, POWER Hypervisor, Power Systems, Power Systems (logo), Power Systems Software, Power Systems Software (logo), POWER2, POWER3, POWER4, POWER4+, POWER5, POWER5+, POWER6, POWER7, pureScale, System i, System p, System p5, System Storage, System z, Tivoli Enterprise, TME 10, TurboCore, Workload Partitions Manager and X-Architecture are trademarks or registered trademarks of International Business Machines Corporation in the United States, other countries, or both. If these and other IBM trademarked terms are marked on their first occurrence in this information with a trademark symbol (® or ™), these symbols indicate U.S. registered or common law trademarks owned by IBM at the time this information was published. Such trademarks may also be registered or common law trademarks in other countries. A current list of IBM trademarks is available on the Web at "Copyright and trademark information" at www.ibm.com/legal/copytrade.shtml

The Power Architecture and Power.org wordmarks and the Power and Power.org logos and related marks are trademarks and service marks licensed by Power.org.UNIX is a registered trademark of The Open Group in the United States, other countries or both. Linux is a registered trademark of Linus Torvalds in the United States, other countries or both.Microsoft, Windows and the Windows logo are registered trademarks of Microsoft Corporation in the United States, other countries or both.Intel, Itanium, Pentium are registered trademarks and Xeon is a trademark of Intel Corporation or its subsidiaries in the United States, other countries or both.AMD Opteron is a trademark of Advanced Micro Devices, Inc.Java and all Java-based trademarks and logos are trademarks of Sun Microsystems, Inc. in the United States, other countries or both. TPC-C and TPC-H are trademarks of the Transaction Performance Processing Council (TPPC).SPECint, SPECfp, SPECjbb, SPECweb, SPECjAppServer, SPEC OMP, SPECviewperf, SPECapc, SPEChpc, SPECjvm, SPECmail, SPECimap and SPECsfs are trademarks of the Standard Performance Evaluation Corp (SPEC).NetBench is a registered trademark of Ziff Davis Media in the United States, other countries or both.AltiVec is a trademark of Freescale Semiconductor, Inc.Cell Broadband Engine is a trademark of Sony Computer Entertainment Inc.InfiniBand, InfiniBand Trade Association and the InfiniBand design marks are trademarks and/or service marks of the InfiniBand Trade Association. Other company, product and service names may be trademarks or service marks of others.

Revised February 9, 2010

Special notices (cont.)


IBM Power Systems

151

The IBM benchmarks results shown herein were derived using particular, well configured, development-level and generally-available computer systems. Buyers should consult other sources of information to evaluate the performance of systems they are considering buying and should consider conducting application oriented testing. For additional information about the benchmarks, values and systems tested, contact your local IBM office or IBM authorized reseller or access the Web site of the benchmark consortium or benchmark vendor.

IBM benchmark results can be found in the IBM Power Systems Performance Report at http://www.ibm.com/systems/p/hardware/system_perf.html .

All performance measurements were made with AIX or AIX 5L operating systems unless otherwise indicated to have used Linux. For new and upgraded systems, AIX Version 4.3, AIX 5L or AIX 6 were used. All other systems used previous versions of AIX. The SPEC CPU2006, SPEC2000, LINPACK, and Technical Computing benchmarks were compiled using IBM's high performance C, C++, and FORTRAN compilers for AIX 5L and Linux. For new and upgraded systems, the latest versions of these compilers were used: XL C Enterprise Edition V7.0 for AIX, XL C/C++ Enterprise Edition V7.0 for AIX, XL FORTRAN Enterprise Edition V9.1 for AIX, XL C/C++ Advanced Edition V7.0 for Linux, and XL FORTRAN Advanced Edition V9.1 for Linux. The SPEC CPU95 (retired in 2000) tests used preprocessors, KAP 3.2 for FORTRAN and KAP/C 1.4.2 from Kuck & Associates and VAST-2 v4.01X8 from Pacific-Sierra Research. The preprocessors were purchased separately from these vendors. Other software packages like IBM ESSL for AIX, MASS for AIX and Kazushige Goto’s BLAS Library for Linux were also used in some benchmarks.

For a definition/explanation of each benchmark and the full list of detailed results, visit the Web site of the benchmark consortium or benchmark vendor.

TPC http://www.tpc.org SPEC http://www.spec.org LINPACK http://www.netlib.org/benchmark/performance.pdf Pro/E http://www.proe.com GPC http://www.spec.org/gpc VolanoMark http://www.volano.com STREAM http://www.cs.virginia.edu/stream/ SAP http://www.sap.com/benchmark/ Oracle Applications http://www.oracle.com/apps_benchmark/ PeopleSoft - To get information on PeopleSoft benchmarks, contact PeopleSoft directly Siebel http://www.siebel.com/crm/performance_benchmark/index.shtm Baan http://www.ssaglobal.com Fluent http://www.fluent.com/software/fluent/index.htm TOP500 Supercomputers http://www.top500.org/ Ideas International http://www.ideasinternational.com/benchmark/bench.html Storage Performance Council http://www.storageperformance.org/results

Revised March 12, 2009

Notes on benchmarks and values


IBM Power Systems

152

Revised March 12, 2009

Notes on HPC benchmarks and valuesThe IBM benchmarks results shown herein were derived using particular, well configured, development-level and generally-available computer systems. Buyers should consult other sources of information to evaluate the performance of systems they are considering buying and should consider conducting application oriented testing. For additional information about the benchmarks, values and systems tested, contact your local IBM office or IBM authorized reseller or access the Web site of the benchmark consortium or benchmark vendor.

IBM benchmark results can be found in the IBM Power Systems Performance Report at http://www.ibm.com/systems/p/hardware/system_perf.html .

All performance measurements were made with AIX or AIX 5L operating systems unless otherwise indicated to have used Linux. For new and upgraded systems, AIX Version 4.3 or AIX 5L were used. All other systems used previous versions of AIX. The SPEC CPU2000, LINPACK, and Technical Computing benchmarks were compiled using IBM's high performance C, C++, and FORTRAN compilers for AIX 5L and Linux. For new and upgraded systems, the latest versions of these compilers were used: XL C Enterprise Edition V7.0 for AIX, XL C/C++ Enterprise Edition V7.0 for AIX, XL FORTRAN Enterprise Edition V9.1 for AIX, XL C/C++ Advanced Edition V7.0 for Linux, and XL FORTRAN Advanced Edition V9.1 for Linux. The SPEC CPU95 (retired in 2000) tests used preprocessors, KAP 3.2 for FORTRAN and KAP/C 1.4.2 from Kuck & Associates and VAST-2 v4.01X8 from Pacific-Sierra Research. The preprocessors were purchased separately from these vendors. Other software packages like IBM ESSL for AIX, MASS for AIX and Kazushige Goto’s BLAS Library for Linux were also used in some benchmarks.

For a definition/explanation of each benchmark and the full list of detailed results, visit the Web site of the benchmark consortium or benchmark vendor.SPEC http://www.spec.org LINPACK http://www.netlib.org/benchmark/performance.pdf Pro/E http://www.proe.com GPC http://www.spec.org/gpc STREAM http://www.cs.virginia.edu/stream/ Fluent http://www.fluent.com/software/fluent/index.htm TOP500 Supercomputers http://www.top500.org/ AMBER http://amber.scripps.edu/ FLUENT http://www.fluent.com/software/fluent/fl5bench/index.htm GAMESS http://www.msg.chem.iastate.edu/gamess GAUSSIAN http://www.gaussian.com ANSYS http://www.ansys.com/services/hardware-support-db.htm

Click on the "Benchmarks" icon on the left hand side frame to expand. Click on "Benchmark Results in a Table" icon for benchmark results.ABAQUS http://www.simulia.com/support/v68/v68_performance.php ECLIPSE http://www.sis.slb.com/content/software/simulation/index.asp?seg=geoquest& MM5 http://www.mmm.ucar.edu/mm5/ MSC.NASTRAN http://www.mscsoftware.com/support/prod%5Fsupport/nastran/performance/v04_sngl.cfm STAR-CD www.cd-adapco.com/products/STAR-CD/performance/320/index/html NAMD http://www.ks.uiuc.edu/Research/namd HMMER http://hmmer.janelia.org/

http://powerdev.osuosl.org/project/hmmerAltivecGen2mod


IBM Power Systems

153

Revised April 2, 2007

Notes on performance estimatesrPerf for AIX

rPerf (Relative Performance) is an estimate of commercial processing performance relative to other IBM UNIX systems. It is derived from an IBM analytical model which uses characteristics from IBM internal workloads, TPC and SPEC benchmarks. The rPerf model is not intended to represent any specific public benchmark results and should not be reasonably used in that way. The model simulates some of the system operations such as CPU, cache and memory. However, the model does not simulate disk or network I/O operations.

rPerf estimates are calculated based on systems with the latest levels of AIX and other pertinent software at the time of system announcement. Actual performance will vary based on application and configuration specifics. The IBM eServer pSeries 640 is the baseline reference system and has a value of 1.0. Although rPerf may be used to approximate relative IBM UNIX commercial processing performance, actual system performance may vary and is dependent upon many factors including system hardware configuration and software design and configuration. Note that the rPerf methodology used for the POWER6 systems is identical to that used for the POWER5 systems. Variations in incremental system performance may be observed in commercial workloads due to changes in the underlying system architecture.

All performance estimates are provided "AS IS" and no warranties or guarantees are expressed or implied by IBM. Buyers should consult other sources of information, including system benchmarks, and application sizing guides to evaluate the performance of a system they are considering buying. For additional information about rPerf, contact your local IBM office or IBM authorized reseller.

========================================================================

CPW for IBM i

Commercial Processing Workload (CPW) is a relative measure of performance of processors running the IBM i operating system. Performance in customer environments may vary. The value is based on maximum configurations. More performance information is available in the Performance Capabilities Reference at: www.ibm.com/systems/i/solutions/perfmgmt/resource.html

© 2010 IBM Corporation IBM Power Systems Power your planet. Smarter Systems for a Smarter Planet...

Documents

Transcript of © 2010 IBM Corporation IBM Power Systems Power your planet. Smarter Systems for a Smarter Planet...