Performance Reporter OS/390 Distributed Systems...

Performance Reporter OS/390Distributed Systems Performance FeatureGuide and ReferenceRelease 5 SH19-4018-04

Performance Reporter OS/390Distributed Systems Performance FeatureGuide and ReferenceRelease 5

Fifth Edition (August 2000)

This major revision obsoletes and replaces SH19-4018-03. The major changes are described in “Summary of changes”on page xxi. Changes or additions to the text and illustrations are indicated by a vertical line to the left of the change.

This edition applies to Release 5 Modification Level 0 of Tivoli Performance Reporter for OS/390, Program Number5695-101, and to all subsequent releases and modifications until otherwise indicated in new editions. Make sure youare using the correct edition for the level of the product. This product was previously known as Enterprise PerformanceData Manager/MVS (EPDM).

Order publications through your IBM representative or the IBM branch office serving your locality. Publications are notstocked at the address given below.

Tivoli welcomes your comments. Please address your comments to:

Information DevelopmentRome Tivoli LaboratoryIBM Italy S.p.A.Via Sciangai, 5300144 RomeItaly

You can also send your comments by fax +39 06 5966 2077 or electronic mail:Internet: [email protected]

When you send information to Tivoli, you grant Tivoli a nonexclusive right to use or distribute the information in anyway it believes appropriate without incurring any obligation to you.

Copyright © 1994, 2000 by IBM Corporation, including this documentation and all software. All rights reserved.May only be used pursuant to a Tivoli Systems Software License Agreement, an IBM Software License Agreement, orAddendum for Tivoli Products to IBM Customer or License Agreement. No part of this publication may be reproduced,transmitted, transcribed, stored in a retrieval system, or translated into any computer language, in any form or by anymeans, electronic, mechanical, magnetic, optical, chemical, manual, or otherwise, without prior written permission ofIBM Corporation. IBM Corporation grants you limited permission to make hardcopy or other reproductions of anymachine-readable documentation for your own use, provided that each such reproduction shall carry the IBMCorporation copyright notice. No other rights under copyright are granted without prior written permission of IBMCorporation. The document is not intended for production and is furnished “as is” without warranty of any kind. Allwarranties on this document are hereby disclaimed including the warranties of merchantability and fitness for aparticular purpose.

Note to U.S. Government Users -- Documentation related to restricted rights -- Use, duplication or disclosure is subjectto restrictions set forth in GSA ADP Schedule Contract with IBM Corp.

Note

Before using this information and the product it supports, be sure to read the general information under“Notices” on page xiii.

ISO 9001 Certification

This product was developed using an ISO 9001 certified quality system.

Certification has been awarded by Bureau Veritas Quality International (BVQI)

(Certification No. BVQI-92053).

BVQI is a world leader in quality certification and is currently recognized by more than 20 accreditation bodies.

Contents

Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Programming interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Trademarks and service marks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

About this book. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xixWho should use this book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix

How to use this book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix

Summary of changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

Changes for this edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

Changes for the previous edition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

Contacting customer support. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

Where to find more information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxii

Performance Reporter library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxii

Part I. UNIX Guide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Chapter 1. Introducing the UNIX Performance component. . . . . . . . . . . . . . . . 3Gathering and collecting performance data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Report Groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Chapter 2. Installing and configuring the UNIX Performancecomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Planning the implementation process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Considering which subcomponents to install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Installing the UNIX Performance component on your UNIX nodes . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Step 1: Check UNIX requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Step 2: Transfer Performance Reporter feature code to UNIX . . . . . . . . . . . . . . . . . . . . . . . . . 10

Capturing UNIX Performance component data and transmitting to OS/390 . . . . . . . . . . . . . . . . . . . . 14

Step 1: Set up UNIX accounting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Step 2: Create log files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Step 3: Define the receiver of UNIX Performance component data . . . . . . . . . . . . . . . . . . . . . 18

Step 4: Customize UNIX Performance component parameters . . . . . . . . . . . . . . . . . . . . . . . . . 18

Step 5: Start the drlperfd daemon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Step 6: Invoke drlperfd at boot time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Description of daily tasks performed by UNIX Performance component . . . . . . . . . . . . . . . . . . 21

iiiDistributed Systems Performance Feature

||

||

Operating the UNIX Performance component on UNIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Description of UNIX Performance component files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

AIX commands to use if UNIX Performance component installed . . . . . . . . . . . . . . . . . . . . . . 23

Installing UNIX Performance component on the OS/390 system. . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Updating the lookup tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Updating XACCT_CONNECT_TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Updating X_NODE_NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Updating XACCT_PRINT_TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Collecting data under Performance Reporter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Creating your Generation Data Groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Renaming files sets sent from UNIX for use with a GDG . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

General collect job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Collect job with specified Performance Reporter tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Backup and delete of generation data sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

Testing the installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

Putting the feature into production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

Part II. UNIX Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Chapter 3. Data flow and Performance Reporter Objects . . . . . . . . . . . . . . . . 37UNIX Performance component general data flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Description of log files, record and log definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

UNIX Accounting subcomponent data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

Where to look for further information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

UNIX Configuration subcomponent data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43


UNIX Error subcomponent data flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45


UNIX Performance subcomponent data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47


Chapter 4. Data, Lookup, and Control Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Naming standard for tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Table descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

Tables in the accounting subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

XACCT_COMMAND_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

XACCT_DISK_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

XACCT_PRINT_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

XACCT_CONNECT_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

iv Release 5

XACCT_DISK_MV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Tables in the configuration subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

XCONFIG_HW. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

XCONFIG_SW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

Tables in the error subcomponent. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

XERROR_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Tables in the performance subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

XPERF_CPU_H, _D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

XPERF_DISK_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

XPERF_DISKIO_H, _D, _M. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

XPERF_PAGING_H, _D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Lookup tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

XACCT_CONNECT_TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

X_NODE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

XACCT_PRINT_TYPE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Chapter 5. Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Report format and general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Report ID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Report group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Source. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Reports in the accounting subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

UNIX Acct Commands by User, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

UNIX Acct Users by Command, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

UNIX Acct Cmd Resource Consumption, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

UNIX Acct User Resource Usage, Monthly Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

UNIX Acct Disk Blocks by User, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

UNIX Acct Disk Blocks in 1000s, Monthly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

UNIX Acct Users and Connects, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

UNIX Acct Printed Pages by User, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

UNIX Acct Printed Pages by System, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Reports in the configuration subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

UNIX Configuration of HW for a System, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

UNIX Configuration of HW for Device Class, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

UNIX Configuration of SW for a System, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

UNIX Configuration of SW for Object, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Reports in the error subcomponent. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

vDistributed Systems Performance Feature

UNIX Error by ID, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

UNIX Error by Type, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

UNIX Error by Class, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

UNIX Error by Resource, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

UNIX Error by Resource, Monthly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

Reports in the performance subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

UNIX Perf CPU Utilization by System, Hourly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

UNIX Perf CPU Utilization by System, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

UNIX Perf Statistics by System, Hourly Trend. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

UNIX Perf Statistics all Systems, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

UNIX Perf Vol Group and File Syst, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

UNIX Perf Disk I/O for a Disk, Hourly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

UNIX Perf Disk I/O for System, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

UNIX Perf Page Space Utilization, Hourly Trend. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

Chapter 6. Log Record Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109ACCT_COMMAND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

ACCT_CONNECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

ACCT_DISK. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

ACCT_PRINT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

CONFIG_HW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

CONFIG_SW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Record section SUBTYPE_H. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Record section SUBTYPE_I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Record section SUBTYPE_L . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Record section FILE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Record section PATH_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

Record section PRODUCT_DESC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

ERROR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

PERF_CPU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

PERF_DISK_FS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

vi Release 5

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Record section FILE_SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Record section DEVICE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Record section SPACE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

PERF_DISK_VG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

Record section VOLUME_GROUP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

Record section SPACE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

PERF_DISK_IO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

PERF_PAGING. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Chapter 7. Migrating from SP6000 feature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123Migrating the host part of SP6000 feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Replacing agent programs on the nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

Part III. ARM Guide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Chapter 8. Introducing the ARM component . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

Chapter 9. Installing and Configuring the ARM Component. . . . . . . . . . . . 129Installing the ARM component on OS/390 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

Updating the lookup tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

Updating X_NODE_NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

Testing the installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

Putting the ARM component into production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

Part IV. ARM Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Chapter 10. Data Flow and Performance Reporter Objects. . . . . . . . . . . . . 133ARM Agents general data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

Chapter 11. Data Tables, Lookup Tables, and Control tables. . . . . . . . . . . 137Naming standard for tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

Table descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

ARM_TABLE_H,_D,_M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

Lookup Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

X_NODE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Chapter 12. Reports. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

viiDistributed Systems Performance Feature

ARM report definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

ARM reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

UNIX Application Monitoring Hourly Response Times trend report . . . . . . . . . . . . . . . . . . . . 142

UNIX Application Monitoring Hourly Service Level Summary report. . . . . . . . . . . . . . . . . . . 143

Chapter 13. Log Record Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145ARM log collector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

ARM log definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

ARM record definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

Part V. Distributed Monitoring Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

Chapter 14. Introducing Distributed Monitoring. . . . . . . . . . . . . . . . . . . . . . . . . 151

Chapter 15. Configuring Distributed Monitoring . . . . . . . . . . . . . . . . . . . . . . . . 153Processing Distributed Monitoring data and sending it to OS/390 . . . . . . . . . . . . . . . . . . . . . . . . . . 153

Create the task from a UNIX system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

Create the task from an OS/390 UNIX system service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

Create the task from a Windows NT system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

Execute the task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

Updating the lookup tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Updating X_NODE_NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Updating SENTRY_MONITOR_NT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Updating SENTRY_MONITOR_UN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Testing the installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Putting the Distributed Monitoring component into production . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Part VI. Distributed Monitoring Reference . . . . . . . . . . . . . . . . . . . . . . . . . . 161

Chapter 16. Data Flow and Performance Reporter Objects. . . . . . . . . . . . . 163Tivoli Distributed Monitoring general data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

Chapter 17. Data Tables, Lookup Tables, and Control Tables . . . . . . . . . . 167Naming standard for tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

Table descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

SENTRY_N_NT_H,_D,_W . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

SENTRY_N_UN_H,_D,_W . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Lookup Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

X_NODE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

SENTRY_MONITOR_NT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

viii Release 5

||

SENTRY_MONITOR_UN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Chapter 18. Reports. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173Distributed Monitoring report definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

DM Service Level Details Hourly (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

DM Service Level Details Daily (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176

DM Service Level Details Weekly (UNIX). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178

DM System Resources Monitors Daily (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180

DM Network Monitors Daily (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

DM Disk Resource Monitors Daily (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

DM Service Level Details Hourly (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

DM Service Level Details Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

DM Service Level Details Weekly (NT). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

DM TCP Monitors Daily (NT). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192

DM System Monitors Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193

DM Server Monitors Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

DM Processor Monitors Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

DM Memory Details Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

Chapter 19. Log Record Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197Distributed Monitoring log collector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

Distributed Monitoring log definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

Distributed Monitoring record definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

Part VII. Appendixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199

List of abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

ixDistributed Systems Performance Feature

x Release 5

Figures

1. Organizing and presenting system performance data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42. Implementation process for the UNIX Performance component. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83. Performance Reporter Administration window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254. Components window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255. Installation Options window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266. General COLLECT job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297. Sample COLLECT job with an INCLUDE clause. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308. Daily steps involved in using Performance Reporter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319. General UNIX Performance component data flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

10. UNIX Accounting component data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4111. UNIX Configuration subcomponent data flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4312. UNIX Error subcomponent data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4513. UNIX Performance subcomponent data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4714. Example of UNIX Acct Commands by User, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7015. Example of UNIX Acct Users by Command, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7216. Example of UNIX Acct Cmd Resource Consumption, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . 7417. Example of UNIX Acct User Resource Usage, Monthly Overview . . . . . . . . . . . . . . . . . . . . . . . . . . 7618. Example of UNIX Acct Disk Blocks by User, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . 7719. Example of UNIX Acct Disk Blocks in 1000s, Monthly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7820. Example of UNIX Acct Users and Connects, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7921. Example of UNIX Acct Printed Pages by User, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . . . 8022. Example of UNIX Acct Printed Pages by System, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . 8123. Example of UNIX Configuration of HW for a System, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 8324. Example of UNIX Configuration of HW for Device Class, Overview . . . . . . . . . . . . . . . . . . . . . . . . 8525. Example of UNIX Configuration of SW for a System, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 8626. Example of UNIX Configuration of SW for Object, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8827. Example of UNIX Error by ID, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9028. Example of UNIX Error by Type, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9129. Example of UNIX Error by Class, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9230. Example of UNIX Error by Resource, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9331. Example of UNIX Error by Resource, Monthly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9432. Example of UNIX Perf CPU Utilization by System, Hourly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . 9633. Example of UNIX Perf CPU Utilization by System, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . 9734. Example of UNIX Perf Statistics by System, Hourly Trend. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9835. Example of UNIX Perf Statistics all Systems, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10036. Example of UNIX Perf Vol Group and File Syst, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . 10237. Example of UNIX Perf Disk I/O for a Disk, Hourly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10338. Example of UNIX Perf Disk I/O for System, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10539. Example of UNIX Perf Page Space Utilization, Hourly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10640. Daily steps involved in using Performance Reporter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13041. General ARM data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13442. X_NODE_NAME lookup table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13943. UNIX Application Monitoring Hourly Response Times trend report . . . . . . . . . . . . . . . . . . . . . . . . 14244. UNIX Application Monitoring Hourly Service Level Summary report, partial view . . . . . . . . . . . . . 14345. Daily steps involved in using Performance Reporter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16046. General Distributed Monitoring data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16447. X_NODE_NAME lookup table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17048. SENTRY_MONITOR_NT lookup table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17149. SENTRY_MONITOR_UN lookup table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

xiDistributed Systems Performance Feature

50. DM Service Level Details Hourly Report (UNIX). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17451. DM Service Level Details Daily Report (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17652. DM Service Level Details Weekly Report (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17853. DM System Resources Monitors Daily Report (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18054. DM Network Monitors Daily Report (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18155. DM Disk Resource Monitors Daily Report (UNIX). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18256. DM Service Level Details Hourly Report (UNIX). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18357. DM Service Level Details Daily Report (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18658. DM Service Level Details Weekly Report (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18959. DM TCP Monitors Daily Report (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19260. DM System Monitors Daily Report (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19361. DM Server Monitor Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19462. DM Processor Monitors Daily Report (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19563. DM Memory details Daily Report (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

xii Release 5

Notices

References in this publication to Tivoli Systems or IBM products, programs, or services donot imply that they will be available in all countries in which Tivoli Systems or IBMoperates. Any reference to these products, programs, or services is not intended to imply thatonly Tivoli Systems or IBM products, programs, or services can be used. Subject to TivoliSystems’s or IBM’s valid intellectual property or other legally protectable right, anyfunctionally equivalent product, program, or service can be used instead of the referencedproduct, program, or service. The evaluation and verification of operation in conjunctionwith other products, except those expressly designated by Tivoli Systems or IBM, are theresponsibility of the user.

Tivoli Systems or IBM may have patents or pending patent applications covering subjectmatter in this document. The furnishing of this document does not give you any license tothese patents. You can send license inquiries, in writing, to:

IBM Director of LicensingIBM CorporationNorth Castle DriveArmonk, NY 10504 - 1785USA

Licensees of this program who wish to have information about it for the purpose ofenabling: (i) the exchange of information between independently created programs and otherprograms (including this one) and (ii) the mutual use of the information which has beenexchanged, should contact:

IBM CorporationP.O. Box 121953039 CornwallisResearch Triangle Park, NC 27709-2195USA

Such information may be available, subject to appropriate terms and conditions, including insome cases, payment of a fee.

For information about the trademarks of Tivoli Systems, the IBM Corporation, and othercompanies that appear in the Performance Reporter for OS/390 library, see “Trademarks andservice marks” on page xvii.

xiiiDistributed Systems Performance Feature

xiv Release 5

Programming interfaces

This book is intended to help users monitor their workstations. This book documentsinformation which is diagnosis, modification, or tuning information provided by PerformanceReporter for OS/390.

Attention: Do not use this diagnosis, modification, or tuning information as aprogramming interface.

xvDistributed Systems Performance Feature

xvi Release 5

Trademarks and service marks

The following terms are trademarks of International Business Machines Corporation orTivoli Systems Inc, in the United States, other countries, or both. In Denmark, Tivoli is atrademark licensed from Kjøbenhavns Sommer - Tivoli A/S.

IBM TivoliAIX Database 2DB2 DB2MVS NetViewOS/390 RISC System/6000System/36 System/390

UNIX is a registered trademark of The Open Group in the United States and other countries.

Microsoft, Windows, Windows NT, and the Windows logo are registered trademarks ofMicrosoft Corporation.

Other company, product, and service names may be trademarks or service marks of others.

xviiDistributed Systems Performance Feature

xviii Release 5

About this book

This book describes how to use Tivoli Performance Reporter for OS/390 to collect andreport performance data generated by operating systems and applications running on aworkstation. This book:

¶ Describes performance issues and how they affect the level of services you can offerusers

¶ Guides you through the component-selection, installation, and implementation processes

¶ Explores performance characteristics shown in Performance Reporter reports so that youcan analyze the characteristics of your system

The terms MVS and OS/390 are used interchangeably throughout this book.

Notes:

1. Release 1 Modification Level 1 of this feature had the title RISC System/6000 SystemPerformance Feature.

2. Release 2 of this feature had the title UNIX Performance feature.

3. Release 4 of this feature had the title Workstation Performance Feature.

4. Release 5 of this feature has the title Distributed Systems Performance Feature, whichreflects the additional functionality that has been incorporated.

5. The term UNIX used in this book covers AIX, HP-UX, and Sun Solaris operatingsystems. When AIX is used, this refers only to the IBM operating system (and not toHP-UX or Sun Solaris).

Who should use this bookThe Workstation Performance Feature Guide and Reference is for:

¶ Anyone who analyzes or monitors UNIX performance

¶ Anyone responsible for establishing or meeting service-level objectives.

¶ Performance Reporter administrators (primarily as a guide to feature installation and as areference to table and report definitions)

¶ Users with various backgrounds who are interested in analyzing workstationperformance data and improving workstation performance.

How to use this bookThis book explains how to collect UNIX-generated performance data, and to create anddisplay Performance Reporter reports to both monitor and understand UNIX performance.

This book contains the following parts:

¶ Use Part I for developing and implementing your performance and service-level strategy:

v Introducing the UNIX Performance component explains basic concepts of systemsmanagement and offers suggestions on establishing performance objectives andservice-level agreements. It also describes the UNIX Performance feature role in thePerformance Reporter environment.

xixDistributed Systems Performance Feature

|

||

v Installing and configuring the UNIX Performance component contains a task-orienteddescription of how to plan for and set up the UNIX Performance feature so thatuseful reports and decision-support information are available immediately.

¶ Use Part II for a detailed description of the UNIX Performance component:

v “Data flow and Performance Reporter Objects” on page 37 describes the flow of datafrom log record to report, showing Performance Reporter log and record definitions,tables, and reports.

v “Data, Lookup, and Control Tables” on page 49 describes the supplied data tables andupdates, including columns and expressions. It contains similar descriptions of lookuptables and control tables.

v “Reports” on page 67 contains examples of the UNIX Performance componentreports and shows where the data comes from.

v “Log Record Definitions” on page 109 describes the UNIX Performance componentrecord definitions and lists the record types you can select with each of them.

v “Migrating from SP6000 feature” on page 123 explains what you must do to performthe migration to the UNIX Performance component of Tivoli Performance Reporterfor OS/390.

¶ Use Part III for a description of the ARM component and an explanation of how toinstall and use it.

v “Introducing the ARM component” on page 127 gives an overview of the ARMcomponent.

v “Installing and Configuring the ARM Component” on page 129 explains how toinstall and operate ARM.

¶ Use Part IV for a detailed description of the ARM component:

v “Data Flow and Performance Reporter Objects” on page 133 explains the flow ofdata from log report to record.

v “Data Tables, Lookup Tables, and Control tables” on page 137 describes the datatables that are supplied with the product.

v “Reports” on page 141 explains the types of reports and where the data contained inthem comes from.

v “Data Tables, Lookup Tables, and Control tables” on page 137 contains examples ofthe reports that the ARM component produces.

¶ Use Part V for developing and implementing your performance and service-levelstrategy:

v “Introducing Distributed Monitoring” on page 151 explains the gathering andcollecting of Tivoli Distributed Monitoring data.

v “Configuring Distributed Monitoring” on page 153 explains how to plan for and setupTivoli Distributed Monitoring data collection.

¶ Use Part VI for a detailed description of the Tivoli Distributed Monitoring component:

v “Data Flow and Performance Reporter Objects” on page 163 explains the flow ofdata from log record to report.

xx Release 5

v “Data Tables, Lookup Tables, and Control Tables” on page 167 describes the datatables that are supplied with the product.

v “Reports” on page 173 explains the types of reports and where the data contained inthem comes from.

v “Log Record Definitions” on page 197 contains examples of the reports that theTivoli Distributed Monitoring component produces.

Summary of changesThis section describes the changes that have been made for this edition and for previouseditions of the book.

Changes for this editionThis edition is an updated version that replaces the previous edition of the same book.Except for editorial changes, updates to this edition are marked with a vertical bar to the leftof the change.

The name of this Performance Reporter for OS/390 feature was changed to the DistributedSystems Performance Feature. In the previous release it was called the WorkstationPerformance feature.

Information has been added for support of OS/390 UNIX system service for TivoliDistributed Monitoring.

Changes for the previous editionThe following changes were introduced in the previous edition:

The name of the base product was changed to Tivoli Performance Reporter for OS/390.

The name of the Tivoli/Sentry Distributed Monitoring component was changed to theDistributed Monitoring Component (Sentry) and was added to the book.

The ARM component was added.

Contacting customer supportIf you encounter difficulties with any Tivoli products, you can enterhttp://www.support.tivoli.com to view the Tivoli Support home page. After you link to andsubmit the customer registration form, you will be able to access many customer supportservices on the Web.

Use the following phone numbers to contact customer support in the United States: theTivoli number is 1-800-848-6548 (1-800-TIVOLI8) and the IBM number is 1-800-237-5511(press or say 8 after you reach this number). Both of these numbers direct your call to theTivoli Customer Support Call Center.

We are very interested in hearing from you about your experience with Tivoli products anddocumentation. We welcome your suggestions for improvements. If you have comments orsuggestions about this documentation, please send e-mail to [email protected].

xxiDistributed Systems Performance Feature

|

|||

|||

||

|

|

|

||

|

Where to find more informationInformation that will help you as you use the Workstation Performance feature is availablein the Performance Reporter for OS/390 library and in related libraries.

Guide to the Reporting Dialog describes how to use the Performance Reporter reportingdialog to print reports. Also, Performance Reporter provides extensive online help wheneveryou use Performance Reporter dialogs from the host or from a workstation; this help givesyou access to the online versions of books in the Performance Reporter library.

Whenever you need programming interface information or more information about specificfeatures or functions of the supported workstation products, refer to books in the productlibrary for the related product.

Performance Reporter libraryTivoli customers can find the Tivoli Performance Reporter for OS/390 Release library athttp://www.support.tivoli.com/Prodman/html/AB.html. Choose Availability, thenPerformance Reporter for OS/390.

BookManager softcopy versions of the Performance Reporter library will be available onCD-ROM in the following collection kits:¶ OS/390 Collection , order number SK2T-6700.¶ Networking Systems Collection , order number SK2T-6012.

While using any Performance Reporter dialog, you can use the help pull-down to startBookManager to view any of the following Performance Reporter books:

Administration

Administration Guide SH19-6816

Reporting dialogs

Guide to the Reporting Dialog SH19-6842

Programming

Language Guide and Reference SH19-6817

User interface

Viewer Guide SH19-4517

Problem determination

Messages and Problem Determination SH19-6902

xxii Release 5

Performance analysis

Accounting Feature for the Host SH19-4495

Accounting Feature for the Workstation SH19-4516

AS/400 System Performance Feature Guide andReference

SH19-4019

Capacity Planner Feature Guide and Reference SH19-4021

CICS Performance Feature Guide and Reference SH19-6820

Distributed Systems Performance Feature Guide andReference

SH19-4018

IMS Performance Feature Guide and Reference SH19-6825

Network Performance Feature Installation andAdministration

SH19-6901

Network Performance Feature Reference SH19-6822

Network Performance Feature Reports SH19-6821

System Performance Feature Guide SH19-6818

System Performance Feature Reference Volume I SH19-6819

System Performance Feature Reference Volume II SH19-4494

xxiiiDistributed Systems Performance Feature

xxiv Release 5

Part I — UNIX GuideChapter 1. Introducing the UNIX Performance component. . . . . . . . . . . . . . . . 3Gathering and collecting performance data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Report Groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Chapter 2. Installing and configuring the UNIX Performancecomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Planning the implementation process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Considering which subcomponents to install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Installing the UNIX Performance component on your UNIX nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Step 1: Check UNIX requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Disk space requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Software requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Step 2: Transfer Performance Reporter feature code to UNIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Capturing UNIX Performance component data and transmitting to OS/390 . . . . . . . . . . . . . . . . . . . . . . . . 14

Step 1: Set up UNIX accounting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Setting up accounting for AIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Setting up accounting for HP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Setting up accounting for Sun Solaris . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Step 2: Create log files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Step 3: Define the receiver of UNIX Performance component data . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Step 4: Customize UNIX Performance component parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Parameter file values for all subcomponents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Parameter file values for performance component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Step 5: Start the drlperfd daemon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Step 6: Invoke drlperfd at boot time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Description of daily tasks performed by UNIX Performance component . . . . . . . . . . . . . . . . . . . . . . . . 21

Operating the UNIX Performance component on UNIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Description of UNIX Performance component files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22AIX commands to use if UNIX Performance component installed . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

drlpsperfd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24drlactperfd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24drldeactperfd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24drlsend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Installing UNIX Performance component on the OS/390 system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Updating the lookup tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Updating XACCT_CONNECT_TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Updating X_NODE_NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Updating XACCT_PRINT_TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Collecting data under Performance Reporter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Creating your Generation Data Groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Renaming files sets sent from UNIX for use with a GDG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Example of renaming UNIX files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28General collect job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Collect job with specified Performance Reporter tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Backup and delete of generation data sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

Testing the installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Putting the feature into production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

1Distributed Systems Performance Feature

2 Release 5

Introducing the UNIX Performancecomponent

Tivoli Performance Reporter for OS/390 (hereafter referred to as Performance Reporter) is areporting system that collects performance data logged by computer systems, summarizes thedata, and presents it in a variety of forms for use in systems management. PerformanceReporter consists of a base product and several optional features.

The UNIX Performance component is a solution for environments where OS/390 is used,and where AIX, HP-UX, or Sun Solaris operating systems are installed on nodes within yournetwork. The performance information from the various nodes is transferred to a centralOS/390 site, where reporting and analysis are performed.

This chapter describes how information is gathered and collected using the UNIXPerformance component.

Gathering and collecting performance dataAt the UNIX nodes where UNIX Performance component agents have been installed andfrom which data is to be used, the basic information used by theUNIX Performancecomponent is gathered by the agents into log files containing accounting, configuration,error, and performance data. The source data used by the UNIX Performance component tocreate the log files and a description of the log files, is given here:

Source data Description of created log files

UNIX accounting Log files containing accounting data. To create these logfiles, you must activate accounting in each node that is to beincluded in the performance analysis. However, activatingaccounting can place up to 20 percent load on the system.

UNIX configuration A log file containing configuration data (for hardware andsoftware).

UNIX error A log file containing error data. Error logging is normallyactive in all nodes where the UNIX Performance componentis installed and, therefore, no action by you should berequired to create these log files.

UNIX performance and disk Log files containing processor use, I/O rate, and pagingspace data. These are continuously updated by a programsupplied with the UNIX Performance component.

1


1.In

trod

ucin

gth

eU

NIX

Perfo

rman

ceco

mp

on

ent

The information contained in the log files must be transmitted to OS/390 for use as data setsin the Performance Reporter COLLECT procedure. The most convenient way is to use aproduct such as TCP/IP. The ftp function of TCP/IP can be used to automate logon and filetransfer.

After the transmission of log files is completed, the information contained in log data sets atthe OS/390 host is collected into Performance Reporter tables.

The collected data is combined with more data (called environment data) and is finallypresented in reports.

The process of entering and maintaining environment data is called administration.Performance Reporter provides an administration dialog for maintaining resourceinformation. Refer to the Administration Guide for information on how to use theadministration dialog.

Figure 1 illustrates how data is organized for presentation in Performance Reporter reports.(A more detailed description is provided in Figure 9 on page 38.)

Report GroupsThe reports produced by the UNIX Performance component are grouped in the followingreport groups:

UNIX Nodes

PerformanceReporterUNIXPerformanceFeature

UNIX log files

P.R.records

System data

Gather datato log files

Send log files toOS/390, then collect

required data

Environment datasummarized in theP.R. database

Required datapresented inreport format

P.R.tables

User-supplieddata

ReportReport Report

Figure 1. Organizing and presenting system performance data

4 Release 5

XACCT Accounting reports

XCONFIG Configuration reports

XERROR Error reports

XPERF Performance reports

The reports cover a wide range of needs in a data processing center, and reporting can bedone online or in batch. They are accessible from the repPerformance Reporterorting dialog.

Finally, the key to successful implementation of Performance Reporter is knowing:

¶ The information and resources on which you want to report and how to performcustomization to select them

¶ The way you want to organize, set objectives for, and process the data (used later todefine the environment)


1.In

trod

ucin

gth

eU

NIX

Perfo

rman

ceco

mp

on

ent

6 Release 5

Installing and configuring the UNIXPerformance component

This supplements the procedure in the Administration Guide for installing a component, withinformation specific to the UNIX Performance component.

This chapter describes how to:

¶ Plan the UNIX Performance component implementation process.

¶ Decide which UNIX Performance subcomponent to install.

¶ Install the UNIX Performance component on the UNIX nodes.

¶ Capture UNIX Performance component data and transmit this data to PerformanceReporter on the OS/390 system.

¶ Operate UNIX Performance component on UNIX.

¶ Install UNIX Performance subcomponent on the OS/390 system.

¶ Update the Performance Reporter lookup tables.

¶ Collect data into Performance Reporter tables from log files containing data that hasbeen transmitted from the UNIX nodes.

¶ Test the UNIX Performance component installation.

¶ Put the UNIX Performance component into production.

Planning the implementation process

Note: If you have RISC System/6000 Performance feature already installed, you must firstmigrate to the UNIX Performance component (see “Migrating from SP6000 feature”on page 123 for details).

Before installing the UNIX Performance component, you should follow these steps to planthe implementation process:

1. Describe user tasks. Then determine what data the UNIX Performance component mustgather to help users accomplish those tasks.

2. Determine which UNIX Performance subcomponent you must install to meet the userneeds.

2


2.In

stalling

the

UN

IXC

om

po

nen

t

3. Determine the administration tasks you must perform for the selected subcomponents,and make any decisions required by these tasks. These tasks help you customizePerformance Reporter and the UNIX Performance component to work efficiently andeffectively with your computer system.

4. Determine (for each selected subcomponent) the tasks you must perform to customizethe supported products to work with Performance Reporter and with the UNIXPerformance component.

If this is your first exercise in implementation planning, follow all these steps to ensure thatthe UNIX Performance component’s implementation is consistent. If you are reading thischapter in preparation for modifying your system, you might not need to perform all of thesetasks.

Use the planning process to prepare for these main customization tasks:

¶ Customizing UNIX to generate the data required by the subcomponents you install.

¶ Defining environment data, which is all the information (in addition to the input data)that the UNIX Performance component needs to create reports. Environment datacontrols the data-collection process and provides more information in the reports.

Figure 2 illustrates the process for implementing the UNIX Performance component.

Installing P.R.

Installing the UNIXPerformance component

(and its components)

Administering P.R. and theUNIX Performance component

Testing the component

Evaluating the feature'simplementation

Putting the componentinto production

Planning to use the UNIXPerformance component(implementation andmodifications)

Customizing UNIX

Modification requirements

Maintenance requirements

Figure 2. Implementation process for the UNIX Performance component

8 Release 5

Considering which subcomponents to installYour most critical planning task is determining what information users need from the UNIXPerformance component. For example, users may be interested only in error conditions or inprocessor capacity. Installing only those subcomponents needed to meet user requirementsensures that the feature benefits users while it minimizes the performance impact caused bydata collection and interpretation activities.

The UNIX Performance component is divided into four subcomponents:¶ Accounting¶ Configuration (AIX only)¶ Error (AIX only)¶ Performance

Subcomponents are groups of Performance Reporter objects(for example, predefined updatedefinitions, data tables, and reports). If you find that you need reports from a subcomponentthat you have not installed, you must install that subcomponent and then wait several daysor weeks until enough data is collected to create reports. However, if you install moresubcomponents than you need, Performance Reporter collects needless data, which takes updisk space and uses processor time.

Note: You should be aware that using the process accounting part of of the accountingsubcomponent (log type PACCT) will degrade machine performance. However, usingthe other parts of the accounting subcomponent (log types QACCT, DTMP, and WTMP)will not degrade machine performance.

At this point, you might find it helpful to examine the predefined reports for eachsubcomponent, by turning to “Reports” on page 67.

Installing the UNIX Performance component on your UNIX nodesThis section gives you practical step-by-step information on how to install PerformanceReporter on your UNIX nodes.

Step 1: Check UNIX requirements

Disk space requirementsYou need the following approximate disk space on each UNIX node:

Table 1. Disk space requirements on each UNIX nodeInstallation task Disk space

(approximate)

Loading code in the file system /usr/lpp: 400KB

Creating work files for the volume group rootvg file system. For eachnode from which you are collecting data, and for all subcomponents, yourequire:

43MB for 10 users (seetable note)


2.In

stalling

the

UN

IXC

om

po

nen

t

Table 1. Disk space requirements on each UNIX node (continued)Installation task Disk space

(approximate)

Note: The 43MB consists of:

Component Space requiredAccounting 42.25MB (of which process accounting uses 42MB)Performance 0.4MBConfiguration 180KBError 3KB

Software requirementsThe UNIX Performance component requires one of the following programs, or later levels:

¶ AIX 3.2.5 for RISC System/6000 (5756-030) with:

v (For the transfer of log files to OS/390) Either Host Support File Transfer ProgramVersion 1 (5665-311), and AIX SNA Services/6000 (5601-287), and AIX HostConnection Program/6000 Version 1 Release 3 (5601-260) or TCP/IP Version 2(5735-HAL)

v The PTF (for AIX) U413086 for correct printer accounting

¶ AIX 4.1.1 (5765-393) with:

v Accounting Services

v Software Error Logging and Dump Service Aids

v AIX 3.2 to 4.1 Compatibility Links

¶ HP-UX 9.3

¶ Sun Solaris 2.3

¶ Sun Solaris 1.1

Note: For AIX 4.1.1, HP, and Sun, you also require the corresponding communicationprograms to transfer log files to OS/390

Step 2: Transfer Performance Reporter feature code to UNIXThe UNIX part of the UNIX Performance component is distributed in these SMP targetlibrary members:¶ DRL150.SDRLWS(DRLAIX) for AIX¶ DRL150.SDRLWS(DRLHPUX) for HP-UX¶ DRL150.SDRLWS(DRLSUNOS) for Sun Solaris 1.1¶ DRL150.SDRLWS(DRLSOLAR) for Sun Solaris 2.3

Download these tar files using these steps:

1. Log in as root user on a UNIX node.

2. Enter the commands:mkdir /usr/lpp/epdmcd /usr/lpp/epdm

10 Release 5

|||||

3. Start an ftp session to your OS/390 host, and receive the file from the OS/390 userwhere you installed Performance Reporter. In the following example, the highlightedentries show the commands you must enter. Note that:

yourmvs is the IP address/name of the OS/390 system where PerformanceReporter was installed and where the tar file is located.

epdmunix is the TSO userid that has read access to the tar file.

# ftp yourmvsConnected to yourmvs.yourcorp.com.220-FTPSERVE at YOURMVS.SE.IBM.COM, 11:05:24 on 03/24/93220 Connection will close if idle for more than 5 minutes.Name (yourmvs:root): epdmunix331 Send password please.Password: secret230 EPDMUNIX is logged on.ftp> binary200 Representation type is IMAGE.ftp> get 'drl150.sdrlws(drlaix)' epdmaix.tarftp> get 'drl150.sdrlws(drlhpux)' epdmhp.tarftp> get 'drl150.sdrlws(drlsunos)' epdmsunos.tarftp> get 'drl150.sdrlws(drlsolar)' epdmsolaris.tar200 Port request OK.125 Sending data set DRL150.SDRLWS(DRLAIX) FIXrecfm 128125 Sending data set DRL150.SDRLWS(DRLHPUX) FIXrecfm 128125 Sending data set DRL150.SDRLWS(DRLSUNOS) FIXrecfm 128125 Sending data set DRL150.SDRLWS(DRLSOLAR) FIXrecfm 128250 Transfer completed successfully.389120 bytes received in 25.2 seconds (6.349 Kbytes/s)ftp> quit221 Quit command received. Goodbye.#

4. When the UNIX Performance component code has been received by one UNIX node,send the tar files to all other UNIX nodes where the UNIX Performance component is tobe installed.

5. On each of the UNIX nodes to which the tar files have been sent, enter the following tarcommand to unpack the epdmxxxx tar files (where xxxx is one of aix, hp, sunos, orsolaris) and create the files in the directories:tar -xvf epdmxxxx.tar

6. (Optional) After completing the previous step you can remove the tar files using thiscommand:rm epdmxxxx.tar

(If you need the tar files in the future, they still exist on and caOS/390n be used again).

7. Create file system or directory for log files.

¶ For the AIX operating system, the command drlcrfs is provided to create a filesystem with the name /var/epdm/, in volume group rootvg. The size of the filesystem is determined by an argument used with the drlcrfs. The UNIXPerformance component uses the file system /var/epdm/ to store work and data filesuntil they have been sent to OS/390.

The drlcrfs command is also used to mount the file system using /var/epdm as themount point.


||||||||||||||||||||||||

2.In

stalling

the

UN

IXC

om

po

nen

t

By having the /var/epdm/ directory in a separate file system, there is no risk ofobtaining a “file system full” condition on file system /var (which would stop UNIXprocessing).

For example, to create and mount a new file system with size 50MB, enter:/usr/lpp/epdm/bin/drlcrfs 50

Confirm that the new file system has been created, by entering the followingcommand:df /var/epdm

¶ For the HP and Sun operating systems, the command drlcrfs is provided tocreate a directory with the name /var/epdm. The UNIX Performance component usesthe directory /var/epdm to store work and data files until they have been sent toOS/390.

For example, to create a directory, enter:/usr/lpp/epdm/bin/drlcrfs

Confirm that the new directory has been created, by entering the followingcommand:ls /var/epdm

8. Create symbolic links.

The following programs require symbolic links from the /usr/bin/ directory:drlsenddrlpsperfddrlactperfddrldeactperfdzccmsgwritedrlparmget

To create these symbolic links, enter the following command:/usr/lpp/epdm/bin/drlmklinks

Confirm that the symbolic links are created, using the following command:ls -l /usr/bin/drl* /usr/bin/zcc*

9. The drl.cfg parameter file is distributed as /usr/lpp/epdm/etc/drl.cfg. Copy the parameterfile to /etc/drl.cfg by entering:cp /usr/lpp/epdm/etc/drl.cfg /etc/drl.cfg

12 Release 5

Table 2 shows the files created by Steps 2 to 9.

Table 2. Files created on each system by Steps 2 to 9Directory Description Created files:

/usr/lpp/epdm/bin/ This directory was created when the tarcommand was run. It contains executableUNIX Performance componentcommands.

drlpacctdrlwtmpdrldtmpdrlqacct (see Note 1)drlerrpt (see Note 1)drlconfhw (see Note 1)drlconfsw (see Note 1)drlgatherdrlmklinksdrlrmlinksdrlperfddrliostatdrlvmstatdrlcrfsdrlrmfsdrlsend (see Note 2)drlpsperfd (see Note 2)drlactperfd (see Note 2)drldeactperfd (see Note 2)drlydaydrlparmget (see Note 2)zccmsgwrite (see Note 2)

/usr/lpp/epdm/etc/ This directory was created when the tarcommand was run. It contains variousfiles used by the UNIX Performancecomponent.

drlcrondrl.cfg (see Note 3)rc.epdmunix

/usr/lpp/epdm/lib/ This directory was created when the tarcommand was run. It contains librariesused by UNIX Performance component.

libDrl.a

/usr/lpp/epdm/lib/nls/msg/$LANG/

This directory was created when the tarcommand was run. It contains themessage catalog used by UNIXPerformance component.

drl.cat

Notes:

1. AIX only

2. A symbolic link has been created from /usr/bin/ to this program.

3. This file has been copied to /etc/.


2.In

stalling

the

UN

IXC

om

po

nen

t

Capturing UNIX Performance component data and transmitting toOS/390

This section describes:

¶ The steps you must perform to set up the UNIX Performance component system tocapture performance data, and to transmit the data to the OS/390 system:

Step 1: Set up UNIX accounting (accounting subcomponent).Step 2: Create log filesStep 3: Define the receiver of UNIX Performance component dataStep 4: Customize UNIX Performance component parametersStep 5: Start the drlperfd daemon (performance subcomponent)Step 6: Invoke drlperfd at boot time (performance subcomponent)

Note: Where a step is for a specific subcomponent only, the subcomponent name isgiven in parentheses in the heading for the step. Otherwise, you must carry outthe step for all the subcomponents (accounting, configuration, error, andperformance).

¶ The daily tasks that the UNIX Performance component will perform on your UNIXnode, if you use the default options when setting up the UNIX Performance componentsystem to capture performance data (page21).

Step 1: Set up UNIX accounting(accounting subcomponent only)

This section describes how to set up accounting for:

1. AIX

2. HP (on page 16)

3. Sun (on page 16)

Setting up accounting for AIXThis information can also be found in the section on setting up an accounting system in theAIX Version 3 System Management Guide.

1. Log in as root user.

2. Enter the nulladm command to ensure that the files have the correct accesspermissions, as follows:/usr/sbin/acct/nulladm wtmp pacct

3. Update the /usr/lib/acct/holidays file. The UNIX Performance component does not usethis file, because the accounting data is collected every hour, every day of the year.However, the accounting routines must have a valid file with a valid year in order torun correctly.

4. Start process accounting at boot time, by adding the following line to the /etc/rc file orby deleting (if it exists) the &numsign (comment) in front of the line./usr/bin/su - adm -c /usr/sbin/acct/startup

5. Identify each file system that you want to include in disk accounting by adding thefollowing line to the stanza for the file system in the /etc/filesystems file:account = true

14 Release 5

Example:

/home:dev = /dev/hd1vfs = jfslog = /dev/hd8mount = truecheck = truevol = /homefree = falseaccount = true

6. Specify the data files to be used for printer accounting data by adding one separate lineto each queue stanza in the /etc/qconfig file. The UNIX Performance componentseparates the print queue reports by using different print queue account file names.

Example:

dlp0:device = lp0acctfile = /var/adm/qacct.prquename <== your print queue name

lp0:header = alwaystrailer = neverfile = /dev/lp0backend = /usr/lpd/piobe

7. Create three new accounting directories if they do not already exist:cd /var/adm/acctmkdir nite fiscal sum

8. Make the daily accounting routines run automatically.

The following statement, contained in /usr/lpp/epdm/etc/drlcron, show you how togather all log types. Add this statement to the cron file.0 2 * * * /usr/lpp/epdm/bin/drlgather acct perf >/dev/null 2>&1

Update the root cron file using this command:crontab -e

9. Create a file named siteacct in the /var/adm directory, and insert these statements init. If the file already exists and is a script, insert the statements in the existing file. Thestatements are:date=)date +"%m%d")cat /var/adm/Spacct*.$date >/var/adm/acct/nite/opacct

Set the file permissions and ownership:chmod 550 siteacctchown adm:adm siteacct

If the /var/adm/siteacct file already exists and is a binary file (an executable file),use one of these methods instead: Either: Update /usr/sbin/acct/runacct to alsocall prsiteacct after siteacct, and create a file as above but name it prsiteacct.or: Rename the existing siteacct file to siteacct, and create file siteacct as above,and add after the two statements a call to siteacct as follows:/var/adm/siteacct


2.In

stalling

the

UN

IXC

om

po

nen

t

10. Start process accounting (if accounting is not already running on the AIX system). Thiswill be done at boot time as a result of the above change to the /etc/rc file. However,to start immediate process accounting, issue this command:nohup /usr/sbin/acct/runacct 2> /var/adm/acct/nite/accterr &

Setting up accounting for HPSet up accounting on HP nodes by following the instructions in the appropriate manual. Inaddition, follow these steps:

1. Create a file named runacct.local in the /usr/lib/acct directory, and insert the followingstatements in it. If the file already exists and is a script, insert the statements in theexisting file. The statements to insert are:date=)date +"%m%d")cat /usr/adm/Spacct*.$date >/usr/adm/acct/nite/opacct

Set the file permissions and ownership:chmod 550 runacct.localchown adm:adm runacct.local

2. To run runacct.local, edit the file named /usr/lib/acct/runacct. If necessary, insert thefollowing statement:[ -s /usr/lib/acct/runacct.local ] && /usr/lib/acct/runacct.local

3. Start process accounting (if accounting is not already running):/bin/su - adm -c /usr/lib/acct/startup

You can also add this statement to the /etc/rc file to enable it to be executed at boottime.

Setting up accounting for Sun SolarisSet up accounting on Sun nodes by following the instructions in the appropriate manual. Inaddition, follow these steps:

1. Create a file named runacct.local in the /usr/lib/acct directory, and insert the followingstatements in it. If the file already exists and is a script, insert the statements in theexisting file. The statements to insert are:date=)date +"%m%d")cat /var/adm/Spacct*.$date >/var/adm/acct/nite/opacct

Set the file permissions and ownership:chmod 550 runacct.localchown adm runacct.localchgrp adm runacct.local

2. To run runacct.local, edit the file named /usr/lib/acct/runacct. If necessary, insert thefollowing statement after USEREXIT:[ -s /usr/lib/acct/runacct.local ] && /usr/lib/acct/runacct.local

3. Start process accounting (if accounting is not already running):/usr/bin/su - adm -c /usr/lib/acct/startup

You can also add this statement to the /etc/rc file to enable it to be executed at boottime.

16 Release 5

Step 2: Create log filesBy including drlgather commands in the cron file, you can schedule the creation of logfiles, by entering the following command:/usr/lpp/epdm/bin/drlgather -c conf_day component_list

where:

conf_day is used with the configuration component, and specifies the day on whichthe configuration log file will be created. The creation of the log file occurson (default) the first day of each calendar month. In the following example,the configuration log file will be created every fifth day:-c"1 5 10 15 20 25"

component_listcan be one or more of the following:

¶ acct (accounting subcomponent)

¶ conf (configuration subcomponent)

¶ error (error subcomponent)

¶ perf (performance subcomponent)

1. For AIX, edit the root cron file using the following command:crontab -e

For HP, edit the root cron file using the following commands:cd /var/spool/cron/crontabsvi rootcrontab root (to update the cron daemon)

For Sun Solaris, edit the root cron file using the following commands:cd /usr/spool/cron/crontabsvi rootcrontab root (to update the cron daemon)

For the accounting, error, and performance subcomponents, the creation of the log filesoccurs at (default) 02.00 hours each day.

To verify that the updates were successful, enter:crontab -l

2. Select the subcomponents for which you wish to gather log files.

The following example shows you how to automatically schedule log files for allsubcomponents:0 2 * * * /usr/lpp/epdm/bin/drlgather -c1 acct error perf conf

The next example shows you how to automatically schedule only the log files for theaccounting and performance subcomponents:0 2 * * * /usr/lpp/epdm/bin/drlgather acct perf

(This statement can also be found in /usr/lpp/epdm/etc/drlcon)


2.In

stalling

the

UN

IXC

om

po

nen

t

Step 3: Define the receiver of UNIX Performance component data1. Insert in the home directory .netrc file, the the following entries:

machine mvssystem login epdmuser password secret

2. Change the mvssystem, epdmuser, and secret to the values corresponding to the OS/390host user that will receive the data files. Ftp uses this file when performing the filetransfer, each night.

Note: The value for password must be the same as the current password used on OS/390

3. After editing the .netrc file, check that the file has private read/write permission only, byentering the following command:chmod 600 .netrc

4. Use the command ping -c1 mvssystem, to check that communication between theUNIX Performance component and OS/390 is possible. You can break from thiscommand by typing Ctrl+C. If the ping command fails, you should add a host definitionto the file /etc/hosts, as shown in the following example:9.99.99.99 mvssystem

where ’9.99.99.99’ is the IP address, and ’mvssystem’ is the name of the systemOS/390.

Step 4: Customize UNIX Performance component parametersEdit the parameter file /etc/drl.cfg to change the default values supplied by the UNIXPerformance component to those for your installation. If you are using the parameter filedefault values, you only need to change the value of epdmHost. All the subcomponents usethese parameter file values.

Parameter file values for all subcomponentslogPath = /var/epdmexePath = /usr/lpp/epdm/bin (Do not change)epdmHost = yourmvsepdmPrefix = EPDMUNIXminFree = 2000site = bl=10204 lr=255 rec=vb cy pri=1 sec=5maxRetries = 72retInterval = 60

Parameter Description

logPath The name of directory where all the data log files will be stored.

exePath The name of the directory where all commands are.

msgCatalog The name of the file where all messages are stored.

epdmHost Host name (TCP nodename or TCP internet address) of the OS/390 host toreceive the UNIX Performance component data files.

epdmPrefix Log data set prefix, which you set according to your own requirements.

minFree Minimum free disk space in KB on /var/epdm before running the dailydrlgather.

site OS/390 data set characteristics for the data files sent using TCP/IP FTP toOS/390.

maxRetries When the UNIX Performance component data files are to be automaticallytransferred to OS/390 each night, the communication might be broken for

18 Release 5

some reason, or the OS/390 host might be temporarily unavailable. If that isthe case, the routine drlgather will re-attempt the file transfer of the datafiles for a number of times given by the parameter ’maxRetries’. Using thedefault value for maxRetries, and with the retInterval set at its default of 60minutes, the file transfer retry period will, therefore, cover 3 days.

retInterval Interval in seconds between ftp retries (default=30).

Notes:

1. Do not change the first two parameters logPath, and exPath as they are for internal useonly.

2. You must specify the node name in the epdmHost parameter and the data set prefix inthe epdmPrefix parameter. This is the name of the host and user to receive the data viaTCP/IP FTP. These names must be the same as the values of node name and user ID inthe .netrc file.

Parameter file values for performance componentThe parameter file values used by the performance component (only), are as follows:perfInterval = 60perfCount = 10diskInterval = 60

Parameter Description

perfInterval The program drlperfd issues vmstat and iostat commands, which containthe perfInterval value as the interval parameter. The default interval betweensamples is 60 seconds.

perfCount The program drlperfd samples vmstat and iostat data, at the above intervalvalues. This data is saved in the file /var/epdm/node_name.perf.dyymmdd.n,as follows:

¶ The minimum, maximum and average values during the perfCountinterval

¶ The number of users and processes

¶ The usage of paging space

The default interval between samples is 10 minutes.

diskInterval The program drlperfd issues the commands df and lsvg, and save file spaceand volume group usage in the file /var/epdm/node_name.disk.dyymmdd.nThe default interval between samples is 60 minutes.

Table 3 on page 20 shows the directory created by the drlcrfs command explained in item7 of “Step 2: Transfer Performance Reporter feature code to UNIX” on page 10.


2.In

stalling

the

UN

IXC

om

po

nen

t

Table 3. Directory created by drlcrfs command, and files it will containDirectory Description Files to be created when UNIX Performance

component running

/var/epdm/ This directory is created by thecommand drlcrfs.

The directory resides in a newseparate file system, generatedand mounted over /var/epdm byan installation procedure. Nofiles are loaded in this directoryat installation. The directorywill be used to hold work files,data files and log files. The datafiles will be deleted as soon asthey are successfullytransmitted.

work files: (see Note 1)drlerrpt.time (see Note 1)node_name.perf.worknode_name.disk.work (see Note 1)node_name.perfio.worknode_name.perfvm.work

.sk log data files: (see Notes 2, 3, 4)node_name.pacct.dyymmdd.1node_name.wtmp.dyymmdd.1node_name.dtmp.dyymmdd.1node_name.qacct.dyymmdd.1 (see Note 1)node_name.errpt.dyymmdd.1 (see Note 1)node_name.perf.dyymmdd.1node_name.disk.dyymmdd.1 (see Note 1)node_name.conf.dyymmdd.1 (see Note 1)node_name.epdmunix.dyymmdd.1 (see Note6)

.sk message log files: (see Notes 2, 3)node_name.drlperfd.errorlognode_name.log.dyymmddnode_name.errorlog.dyymmdd

.sk ftp work files: (see Notes 2, 3, 5)node_name.ftpwork.dyymmddnode_name.ftpmsg.dyymmddnode_name.ftperr.dyymmdd

Notes:

1. AIX only

2. ’node_name’ is the name of the node.

3. ’yymmdd’ is the current date.

4. These files will be removed after they have been transferred to OS/390.

5. These files will be removed after the log data files have been transferred to OS/390. If thetransfer to OS/390 is not succesful, these files contain details of the failure.

6. This file contains the 8 files listed above it (node_name.pacct.dyymmdd.1,node_name.wtmp.dyymmdd.1, and so on) and is the file to be transmitted.

Step 5: Start the drlperfd daemon(performance component only)

The processes that will collect performance and disk space data during the day, are initiatedby program drlperfd.

You should now:

1. Start the 4 background processes that gather performance and disk space data, byentering the following command:drlactperfd

20 Release 5

2. Enter the command drlpsperfd to display the status of the UNIX Performancecomponent processes. The four processes that should be listed, are shown in “drlpsperfd”on page 24

Performance data will now be entered in the following files:

¶ work files

v node_name.perf.work

v node_name.disk.work

v node_name.perfio.work

v node_name.perfvm.work

¶ log data files

v node_name.perf.dyymmdd

v node_name.disk.dyymmdd

The file node_name.drlperfd.errorlog contains possible error messages. This file remainsempty providing the drlactperfd command runs successfully.

The file node_name.drlperfd.log contains information messages which inform you, forexample, of the files that have been transferred to OS/390.

Step 6: Invoke drlperfd at boot time(performance component only)

To automatically start the drlperfd program at IPL (boot) time, perform these steps:

1. For AIX, enter this command:mkitab epdm:2:wait:/etc/rc.epdmunix

For HP and Sun operating systems, edit /etc/inittab and add the epdm line at the end asbelow.# 1.28 com/cfg/etc/inittab, bos, bos320 10/3/91 10:46:51# COMPONENT_NAME: CFGETC## ORIGINS: 3, 27...epdm:2:wait:/etc/rc.epdmunix

2. Copy rc.epdmunix to /etc/ using the following command:cp /usr/lpp/epdm/etc/rc.epdmunix /etc/

Description of daily tasks performed by UNIX Performance componentDuring the day, UNIX Performance component has four processes running which are startedby one program. UNIX is set up so that these four processes are autostarted at each IPL(boot-time), as described in “Step 6: Invoke drlperfd at boot time”. The program that startsthe four processes is drlperfd, which periodically logs:

¶ Various performance data to the daily file /var/epdm/node_name.perf.dyymmdd.n (thedefault is every 10 minutes).

¶ Disk space data to the daily file /var/epdm/node_name.disk.dyymmdd.n (the default isevery 60 minutes).


2.In

stalling

the

UN

IXC

om

po

nen

t

Note: An explanation of node_name, yymmdd, and ’.n’ is given in Table 3 on page 20.

Each night, UNIX accounting routines and the UNIX Performance component routines arerun. These routines are started from the UNIX cron daemon at default times (which can becustomized by you), in the root’s cron file.

Time Action

01:00 UNIX starts the dodisk command via cron.

01:30 UNIX starts the runacct command via cron.

02:00 UNIX starts the drlgather command via cron.

Command drlgather calls the following routines:

¶ drlpacct, which reads yesterday’s process accounting file and producesan ASCII file suitable for sending to OS/390.

¶ drlwtmp, which reads yesterday’s login/logout accounting file andproduces an ASCII file suitable for sending to OS/390.

¶ drldtmp, which reads yesterday’s disk accounting file and produces anASCII file suitable for sending to OS/390.

¶ drlqacct, which reads yesterday’s print queue accounting files andproduces an ASCII file suitable for sending to OS/390.

¶ drlerrpt, which prints the AIX error log since previous print andproduces an ASCII file suitable for sending to OS/390.

¶ drlconf, which produces a hardware and software configuration file inASCII format for sending to OS/390. This is performed once per month.

¶ drlsend, which transmits a log file containing the contatenated ASCIIfiles given above, together with the performance and disk space filescreated by drlperfd, to OS/390 (using TCP/IP ftp). If the link to OS/390is temporarily down, or if ftp receives a time out during the connectfrom UNIX to OS/390, the file will be transmitted (by default) one hourlater and this will be repeated (by default) 72 times.

02:30 UNIX starts the monacct command via cron daemon.

Operating the UNIX Performance component on UNIXThis section describes:

¶ The UNIX Performance component logs

¶ Four commands that you use to operate the UNIX Performance component on UNIX (“AIX commands to use if UNIX Performance component installed” on page 23).

Description of UNIX Performance component filesExamine the /var/epdm directory to see how the nightly file transfer routines haveperformed. Look for the files node_name.log.dyymmdd, where ’node_name’ is the name ofthe node and ’yymmdd’ is current date. There should be one information log file per day.

By looking into a particular information log file you can see:

¶ What files were sent that day.

22 Release 5

¶ How big they were.

¶ If and why the ftp file transfer did not complete.

This is an example of the /var/epdm directory:ls -l /var/epdmtotal 1544-rw-r--r-- 1 root system 770 Nov 24 19:12 drl.cfg-rw-r--r-- 1 root system 7847 Feb 07 10:43 node1.disk.work-rw-r--r-- 1 root system 0 Jan 02 19:25 node1.drlperfd.errorlog-rw-r--r-- 1 root system 618581 Jan 22 02:11 node1.epdmunix.d950127.1-rw-r--r-- 1 root system 93 Jan 22 02:12 node1.ftpwork.d950127-rw-r--r-- 1 root system 436 Jan 22 02:12 node1.ftpmsg.d950127-rw-r--r-- 1 root system 224 Jan 22 02:12 node1.ftperr.d950127-rw-r--r-- 1 root system 0 Jan 22 02:00 node1.errorlog.d950122-rw-r--r-- 1 root system 0 Jan 23 02:00 node1.errorlog.d950123-rw-r--r-- 1 root system 0 Jan 24 02:00 node1.errorlog.d950124-rw-r--r-- 1 root system 0 Jan 25 02:00 node1.errorlog.d950125-rw-r--r-- 1 root system 0 Jan 26 02:00 node1.errorlog.d950126-rw-r--r-- 1 root system 0 Jan 27 02:00 node1.errorlog.d950127-rw-r--r-- 1 root system 0 Jan 28 02:00 node1.errorlog.d950128-rw-r--r-- 1 root system 0 Jan 29 02:00 node1.errorlog.d950129-rw-r--r-- 1 root system 0 Jan 22 02:00 node1.errorlog.d950122-rw-r--r-- 1 root system 168 Feb 01 02:01 node1.errorlog.d950201-rw-r--r-- 1 root system 0 Feb 02 02:00 node1.errorlog.d950202-rw-r--r-- 1 root system 0 Feb 03 02:00 node1.errorlog.d950203-rw-r--r-- 1 root system 11 Feb 07 19:10 node1.errpt.time-rw-r--r-- 1 root system 2307 Jan 22 02:02 node1.log.d950122-rw-r--r-- 1 root system 2321 Jan 23 02:02 node1.log.d950123-rw-r--r-- 1 root system 2302 Jan 24 02:03 node1.log.d950124-rw-r--r-- 1 root system 2409 Jan 25 02:04 node1.log.d950125-rw-r--r-- 1 root system 2043 Jan 26 02:02 node1.log.d950126-rw-r--r-- 1 root system 2312 Jan 27 02:03 node1.log.d950127-rw-r--r-- 1 root system 2351 Jan 28 02:03 node1.log.d950128-rw-r--r-- 1 root system 2234 Jan 29 02:02 node1.log.d950129-rw-r--r-- 1 root system 2287 Jan 30 02:04 node1.log.d950130-rw-r--r-- 1 root system 2357 Jan 31 02:03 node1.log.d950131-rw-r--r-- 1 root system 2308 Feb 01 02:02 node1.log.d950201-rw-r--r-- 1 root system 2319 Feb 02 02:04 node1.log.d950202-rw-r--r-- 1 root system 2401 Feb 03 02:02 node1.log.d950203-rw-r--r-- 1 root system 35293 Feb 07 11:08 node1.perf.work-rw-r--r-- 1 root system 111 Feb 07 11:07 node1.perfio.work-rw-r--r-- 1 root system 116 Feb 07 11:07 node1.perfvm.work

In this example there is a communication problem on January 27, so the files for this dayhave not been sent. You could check the contents of the files node1.ftpwork.d950127,node1.ftpmsg.d950127, and node1.ftperr.d950127 for details of the problem. The merged logdata file for this date has been retained by the system. When the problem has been solved,you could then manually send the log data file (node1.epdmunix.d950127.1), using thecommand drlsend.

You can also see that an error has occurred on February 01, because the filenode1.errorlog.d950201 is not empty. You could check the contents of this file for details ofthe problem.

Obsolete log files must be deleted manually.

AIX commands to use if UNIX Performance component installedUse the following four commands when operating the UNIX Performance component onUNIX.


2.In

stalling

the

UN

IXC

om

po

nen

t

drlpsperfdThis command displays the status of the UNIX Performance component processes in UNIXdrlpsperfdThe following drlperfd processes are running

USER COMMAND PID PPIDroot drlperfd 10200 1root drlperfd 11735 10200 (Not for HP or Sun)root drlvmstat 17153 11735root drliostat 19970 11735

drlactperfdUse this command to start the performance component processes in UNIX. It starts thedrlperfd daemon. This daemon will create performance log files. Command drlactperfd canalso be used to restart the drlperfd daemon.

drldeactperfdUse this command to stop the performance component processes in UNIX.

drlsendIf there is a communication problem, the UNIX Performance component routinesautomatically attempt to resend the files. However, for the following reasons you might needto manually resend the data files to OS/390:

¶ If AIX has been re-booted and was “down” at the time of a retry event, the event willnever be re-scheduled.

¶ If the ftp retry period has been exceeded. The ftp retry period is defined using theparameters maxRetries and retInterval in the parameter file /etc/drl.cfg.

All data files for those days in which the data files were not sent, remain on the /var/epdmdirectory. After the communication problems has been corrected, the data files for aparticular day can be sent using the drlsend command. The drlsend command deletes thefiles after they have been successfully transferred. Issue this command with the dateparameter, as follows:drlsend yymmdd

where ’yymmdd’ can be found in the corresponding file names, in the /var/epdm directory.

Installing UNIX Performance component on the OS/390 systemWhen you install UNIX Performance component subcomponents, Performance Reporter willinstall the required log and record definitions, record procedures, and update definitions toPerformance Reporter system tables. Performance Reporter will also install the predefinedtables (described in “Data, Lookup, and Control Tables” on page 49) and reports (describedin “Reports” on page 67). To install the UNIX subcomponents, use the Administrationdialog.

Perform steps 1 to 4, as follows:

1. From the Performance Reporter Administration window ( Figure 3 on page 25), select 2,Components and press Enter.

24 Release 5

The Components window is displayed, (as shown in Figure 4).

2. From the Components window, select the components to install (here, the AIXaccounting component) and press F6.

3. The Installation Options window is displayed, (as shown in Figure 5 on page 26).

Other Utilities Help--------------------------------------------------------------------------

Performance Reporter Administration

Select one of the following. Then press Enter.

2_ 1. System2. Components3. Logs4. Tables5. Reports

Command ===> ______________________________________________________________F1=Help F2=Split F3=Exit F9=Swap F10=Actions F12=Cancel

Figure 3. Performance Reporter Administration window

Components Other Help------------------------------------------------------------------------------

Components ROW 1 TO 13 OF 47

Select one or more components. Then press Enter to Open component.

/ Components Status Date_ Network Line Utilization Component Installed 00-03-19_ Network NetView FTP Component Installed 00-03-19_ Network NPM Internal Utilization Component Installed 00-03-19_ Network NV/SM Internal Utilization Component Installed 00-03-19_ Network Problem Component Installed 00-03-19_ Network RTM Response Time Component Installed 00-03-19_ Network Service Component Installed 00-03-19_ DB2 Component Installed 00-03-17_ Information/Management Component Installed 00-03-17/ UNIX Accounting Component_ UNIX Configuration Component_ UNIX Error Component_ UNIX Performance Component

Command ===> _________________________________________________________________F1=Help F2=Split F3=Exit F5=New F6=Install F7=BkwdF8=Fwd F9=Swap F10=Actions F12=Cancel

Figure 4. Components window


2.In

stalling

the

UN

IXC

om

po

nen

t

4. Using the component-installation procedure in the Administration Guide specify whetherthe component is to be installed online, or in batch mode.

Batch mode installation results in less output than in online. Furthermore, for onlineinstallation your terminal will be blocked for the duration of the installation. Therefore, itis recommended that you install components in batch.

Updating the lookup tablesAll the subcomponents of the UNIX Performance component include lookup tables that youcan customize to specify the groupings you want reflected in your reports.

If you specify online installation, Performance Reporter displays the Lookup Tables window.To edit a lookup table using ISPF edit, select a table and press Enter.

If you specify batch mode installation, you can edit the lookup tables using the ISPF editor,after the component is installed. To enter an ISPF edit from Performance Reporter, youshould:

1. Select ’Performance Reporter Administration’ from the Performance Reporter PrimaryMenu

2. Select ’4. Tables’

3. Select the lookup table that you wish to edit, select the ’Edit’ pull-down, and press Enter

4. Select ’3. ISPF Editor’ from the Edit pull-down.

With the UNIX Performance component, customization involves updating theXACCT_CONNECT_TYPE, X_NODE_NAME, and XACCT_PRINT_TYPE lookup tables,described in the following sections.

Components Other Help.-----------------------------------------------------------------------------.| ROW 1 TO 7 OF 7 || .--------------------------------------------------------------. || S| Installation Options | Enter. || | | || /| Select one of the following. Then press Enter. |Date || /| | || _| __ 1. Online | || _| 2. Batch | || _| | || _| F1=Help F2=Split F6=Objects F9=Swap F12=Cancel | || _|--------------------------------------------------------------| || _ || ***************************** BOTTOM OF DATA ****************************** || || || || || || Command ===> ______________________________________________________________ || F1=Help F2=Split F7=Bkwd F8=Fwd F9=Swap F12=Cancel |'-----------------------------------------------------------------------------'F18=Fwd F9=Swap F10=Actions F12=Cancel

Figure 5. Installation Options window

26 Release 5

Updating XACCT_CONNECT_TYPEThis lookup table has information about connect types. It is used during reporting, totranslate hexadecimal connect codes to a descriptive text.

A sample of the lookup table contents is given in “XACCT_CONNECT_TYPE” on page 63.

Updating X_NODE_NAMEThis lookup table contains UNIX node names, accounting groups, Internet addresses, andperiod plans. It is used during the collect, to translate the node name to its accounting groupand period plan.

If you are installing online, you will be given the opportunity to edit the contents using theISPF editor. A sample of the lookup table contents is given in “X_NODE_NAME” onpage 64. Add the host names of all your UNIX nodes for which you are going to doreporting.

Note: If you have many UNIX nodes, you can obtain a listing of the node names, byentering the following command on one of your AIX nodes:hostent -S

A sample of the lookup table contents is given in “X_NODE_NAME” on page 64.

Updating XACCT_PRINT_TYPENode names and print queue names are grouped together into print types. This lookup tablecontains each print type.

A sample of the lookup table contents is given in “XACCT_PRINT_TYPE” on page 65.

After installation is complete, Performance Reporter returns you to the Components window,and the Status field indicates that the component is installed.

Collecting data under Performance ReporterThere are various methods you can define yourself, to set up the JCL for the COLLECTprocedure. This section, however, describes a method that uses a GDG (generation data set).It explains:¶ How a Generation Data Group (GDG) is created.¶ How the files received from UNIX are renamed, so they can be used with a GDG.¶ The JCL for a general collect job, which uses a GDG to collect data into Performance

Reporter tables.¶ The JCL for a collect job with specified Performance Reporter tables to be updated.

Creating your Generation Data GroupsYou are recommended to use a generation data group (GDG) to avoid the large amount ofJCL renaming work that would otherwise be required each time a collect was run. This isespecially true if you work with a system containing many nodes.

The GDG must be created before you run the first collect. Thereafter, the same GDG can beused with each collect. The following job is provided to create the GDG:DRL150.SDRLCNTL(DRLJXGDG)


|

2.In

stalling

the

UN

IXC

om

po

nen

t

Renaming files sets sent from UNIX for use with a GDGIn a large UNIX network, there will be many files sent from UNIX nodes to the hostOS/390 each night. Each UNIX file transferred to OS/390 is given a unique data set nameaccording to this naming convention:prefix.node_name.EPDMUNIX.Dyymmdd

where

prefix This is the high-level qualifier (for example, the user ID), for receiving filesfrom UNIX.

node_name First qualifier from the UNIX node name.

yymmdd Creation date of the file in UNIX.

This naming convention ensures that no data is lost by overlaying existing data sets, andmakes it easy to identify the origin of the data.

Before each collect of data, you must rename the files sent from UNIX so they become ageneration data set within a GDG. The GDG is then the input to the collect job. Thefollowing job is provided to do this:DRL150.SDRLCNTL(DRLJXPCO)

This job uses the REXX program DRLEXRNM, which renames each UNIX file to ageneration data set within the GDG. The GDG has the following naming convention:prefix.COLLECT.EPDMUNIX

Note: You should customize this job to meet your own naming conventions. Commentedwithin the job DRL120.SDRLCNTL(DRLJXPCO) are detailed instructions on how todo this.

Example of renaming UNIX filesThis example shows how files from the nodes ADAM, BAKER, and CESAR, are renamed togeneration data sets, within the GDG with the name USER1.COLLECT.EPDMUNIX.GDG USER1.COLLECT.EPDMUNIX emptyFollowing datasets found for USER1.*.EPDMUNIX.D* and renamed to USER1.COLLECT.EPDMAIXUSER1.ADAM.EPDMUNIX.D921214 G0001V00USER1.ADAM.EPDMUNIX.D921215 G0002V00USER1.BAKER.EPDMUNIX.D921214 G0003V00USER1.CESAR.EPDMUNIX.D921213 G0004V00USER1.CESAR.EPDMUNIX.D921214 G0005V00USER1.CESAR.EPDMUNIX.D921215 G0006V00

Number of datasets:6

Notes:

1. The program DRLEXRNM used the search criteria ’USER1.*.EPDMUNIX.D*’.

2. The generation numbers are given in the right column (G0001V00, G0002V00, and soon).

3. Generation data sets are given the next sequential generation number (existinggenerations are retained). For example, if the generation data set with generation numberG0006V00 currently exists in the GDG, new data sets will be given the generationnumbers G0007V000, G0008V000, and so on.

28 Release 5

|

General collect jobThe member DRL150.SDRLCNTL(DRLJCOLL) contains sample JCL for a general collect.

Another way to obtain the COLLECT JCL, is to:

1. Select ’Performance Reporter Administration’ from the Performance Reporter PrimaryMenu.

2. Select ’3. LOGS’.

3. Select ’UNIX’ from the list of logs, select the ’Utilities’ pull-down, and press Enter.

4. Select ’1. Collect’.

5. Type your installation-specific information in this window, and press Enter.

The JCL given in Figure 6 is the general collect job which uses (in the DRLLOG statement)the example of a GDG for the nodes ADAM, BAKER, and CESAR.

Collect job with specified Performance Reporter tablesThe Administration Guide describes how to optimize Performance Reporter data collectionby specifying only the needed tables in the INCLUDE (or EXCLUDE) clauses of theCOLLECT statement.

The example in Figure 7 on page 30 shows how to selectively include two of the accountingfeature data tables.

//JOBCARD//* *//* Notes: *//* Before you submit the job: *//* - Check the Performance Reporter and DB2 data set names. *//* - Check the DB2 subsystem name (default is DSN) *//* and P.R. system table prefix (default is DRLSYS). *//* - Insert the correct collect statement in DRLIN *//* (as described above). *//* - Specify the name of the log data set (or GDG) in *//* DRLLOG. In this example, all existing generations *//* are to be collected for files from UNIX *//* *//***************************************************************//COLLECT EXEC PGM=DRLPLC,PARM=('SYSTEM=DB2A SYSPREFIX=DRLSYS')//STEPLIB DD DISP=SHR,DSN=DRL150.SDRLLOAD// DD DISP=SHR,DSN=DSN230.DSNLOAD//DRLIN DD *

COLLECT UNIX;

//DRLLOG DD DISP=SHR,DSN=USER1.COLLECT.EPDMUNIX//DRLOUT DD SYSOUT=*,DCB=(RECFM=F,LRECL=80)//DRLDUMP DD SYSOUT=*,DCB=(RECFM=VB,LRECL=32756)/*

Figure 6. General COLLECT job


|

2.In

stalling

the

UN

IXC

om

po

nen

t

Backup and delete of generation data setsAfter a successful collect, you should back up and delete all generation data sets containedin the GDG. There are 2 reasons why you should do this:

1. It will enable you to use the same procedure without any JCL changes, when running thenext collect job.

2. Unless you delete the existing generation data sets, the next collect job will reuse oldgeneration data sets.

Testing the installationBefore starting the daily use of the UNIX Performance component, run a few tests to checkthat:

1. The installation was successful.

¶ Performance Reporter is collecting the correct data

¶ The data is being stored correctly

¶ The correct data is being used for the creation of reports

2. The lookup tables contain appropriate values.

Refer to Administration Guide for detail about the steps involved in testing componentinstallation.

Putting the feature into productionAfter you run the tests and verify that the installation is successful, you can put the UNIXPerformance component and its subcomponents into production.

//JOBCARD//* *//* Note: *//* - instead of entering each of the *//* 2 table names (below), you can use: *//* INCLUDE LIKE 'DRL.XACCT_D%' *//* *//***************************************************************//COLLECT EXEC PGM=DRLPLC,PARM=('SYSTEM=DB2A SYSPREFIX=DRLSYS')//STEPLIB DD DISP=SHR,DSN=DRL150.SDRLLOAD// DD DISP=SHR,DSN=DSN230.DSNLOAD//DRLIN DD *

COLLECT UNIXINCLUDE DRL.XACCT_DISK_D,

DRL.XACCT_DISK_M,;

//DRLLOG DD DISP=SHR,DSN=USER1.COLLECT.EPDMUNIX//DRLOUT DD SYSOUT=*,DCB=(RECFM=F,LRECL=80)//DRLDUMP DD SYSOUT=*,DCB=(RECFM=VB,LRECL=32756)/*

Figure 7. Sample COLLECT job with an INCLUDE clause

30 Release 5

Figure 8 shows the daily steps involved in using Performance Reporter.

You can run reports in batch, after setting batch parameters for each report using theadministration dialog.

For detailed information about these steps, refer to the Administration Guide.

Check that thefile transferprogram hastransferred data.

RunCOLLECTjob

Performdatabase

maintenance

Createreports

Figure 8. Daily steps involved in using Performance Reporter


2.In

stalling

the

UN

IXC

om

po

nen

t

32 Release 5

Part II — UNIX ReferenceChapter 3. Data flow and Performance Reporter Objects . . . . . . . . . . . . . . . . 37UNIX Performance component general data flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Description of log files, record and log definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40UNIX Accounting subcomponent data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

Where to look for further information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42UNIX Configuration subcomponent data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Where to look for further information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44UNIX Error subcomponent data flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

Where to look for further information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46UNIX Performance subcomponent data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Where to look for further information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Chapter 4. Data, Lookup, and Control Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Naming standard for tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Table descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50Tables in the accounting subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

XACCT_COMMAND_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51XACCT_DISK_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52XACCT_PRINT_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53XACCT_CONNECT_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54XACCT_DISK_MV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Tables in the configuration subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56XCONFIG_HW. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56XCONFIG_SW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

Tables in the error subcomponent. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58XERROR_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Tables in the performance subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59XPERF_CPU_H, _D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59XPERF_DISK_D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60XPERF_DISKIO_H, _D, _M. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61XPERF_PAGING_H, _D, _M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Lookup tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63XACCT_CONNECT_TYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Example of table contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63X_NODE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Example of table contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64XACCT_PRINT_TYPE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Example of table contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Chapter 5. Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Report format and general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Report ID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Report group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Source. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Reports in the accounting subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69UNIX Acct Commands by User, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70


UNIX Acct Users by Command, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72UNIX Acct Cmd Resource Consumption, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74UNIX Acct User Resource Usage, Monthly Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76UNIX Acct Disk Blocks by User, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77UNIX Acct Disk Blocks in 1000s, Monthly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78UNIX Acct Users and Connects, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79UNIX Acct Printed Pages by User, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80UNIX Acct Printed Pages by System, Monthly Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Reports in the configuration subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82UNIX Configuration of HW for a System, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83UNIX Configuration of HW for Device Class, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85UNIX Configuration of SW for a System, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86UNIX Configuration of SW for Object, Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Reports in the error subcomponent. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89UNIX Error by ID, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90UNIX Error by Type, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91UNIX Error by Class, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92UNIX Error by Resource, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93UNIX Error by Resource, Monthly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

Reports in the performance subcomponent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95UNIX Perf CPU Utilization by System, Hourly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96UNIX Perf CPU Utilization by System, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97UNIX Perf Statistics by System, Hourly Trend. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98UNIX Perf Statistics all Systems, Daily Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100UNIX Perf Vol Group and File Syst, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102UNIX Perf Disk I/O for a Disk, Hourly Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103UNIX Perf Disk I/O for System, Daily Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105UNIX Perf Page Space Utilization, Hourly Trend. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

Chapter 6. Log Record Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109ACCT_COMMAND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110ACCT_CONNECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111ACCT_DISK. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112ACCT_PRINT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113CONFIG_HW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114CONFIG_SW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115Record section SUBTYPE_H. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115Record section SUBTYPE_I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115Record section SUBTYPE_L . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115Record section FILE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115Record section PATH_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116Record section PRODUCT_DESC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

ERROR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

PERF_CPU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

34 Release 5

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118PERF_DISK_FS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119Record section FILE_SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119Record section DEVICE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119Record section SPACE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

PERF_DISK_VG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120Record section VOLUME_GROUP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120Record section SPACE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

PERF_DISK_IO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

PERF_PAGING. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122Main section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Chapter 7. Migrating from SP6000 feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123Migrating the host part of SP6000 feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123Replacing agent programs on the nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124


UN

IXR

eference

36 Release 5

Data flow and Performance ReporterObjects

This chapter describes:

¶ The general data flow, starting with the gathering of data at the UNIX nodes into logfiles, and ending with the production of Performance Reporter reports.

¶ The record definitions and log types used for mapping data

¶ The data flow for these UNIX Performance subcomponents, including the names ofUNIX log files, and Performance Reporter records, tables, and reports:

v Accounting subcomponent

v Configuration subcomponent

v Error subcomponent

v Performance subcomponent

3


3.D

ataan

dR

epo

rterO

bjects

UNIX Performance component general data flow

The processing steps shown in Figure 9 are:

�1� Agents gather UNIX data into log files

�2� Transmit log file to Performance Reporter.

�3� Collect log data set information into Performance Reporter tables.

�4� Create reports.

The following sections discuss these steps in more detail.

1. Gather UNIX data, Step �1�

The basic data used by the UNIX Performance component is gathered by agents into logfiles. Agents either issue UNIX (or AIX-only) commands or process files, to gather thedata. For each UNIX Performance component, the method of gathering data into log filesis:

Component How data is gathered

Accounting Agents process UNIX files pacct, wtmp, dtmp, qacct

Configuration Agents issue AIX commands lsdev, lslpp

Error Agents issue AIX command errpt

UNIX Nodes

Log files

Log data sets

Log definitionsRecord definitions

Tables

Reports

Control tablesLookup tables

Performance Reporter System(OS/390)

P.R.“Collect”

Lookup tables

1

2

3

4

Figure 9. General UNIX Performance component data flow

38 Release 5

Performance Agents issue AIX commands df, lsvg, lsps, and UNIXcommands iostat, vmstat

Log files and their use in creating Performance Reporter records, are described in Table 4on page 40.

2. Transmit log file to Performance Reporter, Step �2�

To transmit the log file to Performance Reporter, TCP/IP is used.

3. Collect log data set information to Performance Reporter tables, Step �3�

In processing each record contained in the log data set, the collect procedure:

a. Uses a log collector to update the Performance Reporter table with the record. To dothis, the log collector:

¶ Uses the log definition and record definitions to update the Performance Reportertable with the reformatted record.

¶ Uses an update definition to decide which reformatted record fields are to beincluded in which Performance Reporter table, including further summarizing intoother tables (for example, updating the monthly table XPERF_DISK_M from theinformation used for updating the daily table XPERF_DISK_D).

¶ Takes information from control tables (for example the UNIX Performancecomponent may determine the period in which the measurements were made bylooking up the day type information in the SPECIAL_DAY or DAY_OF_WEEKtables).

¶ Uses lookup tables (which contain user-defined information that defines anorganization’s operating environment) to add user-defined data to thePerformance Reporter table record.

A description of the collect procedure is provided in Language Guide and Reference.

A description of the use of control tables is provided in Administration Guide.

4. Create reports, Step �4�

A description of how to create new reports is provided in Guide to the Reporting Dialog.

The reports that can be created when using the UNIX Performance component aredescribed in “Reports” on page 67.


3.D

ataan

dR

epo

rterO

bjects

Description of log files, record and log definitionsTable 4 gives for each UNIX Performance component subcomponent, the UNIX node logfile (and record sub-types, if any), the Performance Reporter record definition, and thePerformance Reporter log definition.

Table 4. Record and log definitions used by the UNIX Performance componentUNIX Performancecomponentsubcomponent

Log file (and recordsub-types, when used)

Performance Reporter recorddefinition and description (see tablenote 2)

PerformanceReporter logdefinition

Accounting PACCT (commandaccounting)

XACCT_COMMAND (commandaccounting data)

UNIX

Accounting WTMP (connect timeaccounting)

XACCT_CONNECT (user accountingdata)

UNIX

Accounting DTMP (disk accounting) XACCT_DISK (disk usage data) UNIX

Accounting QACCT (print accounting) XACCT_PRINT (print accountingdata)

UNIX

Configuration CONF (A,D,S) (configurationhardware)

XCONFIG_HW (configuration ofhardware data)

UNIX

Configuration CONF (H,I,L) (configurationsoftware)

XCONFIG_SW (configuration ofsoftware data)

UNIX

Error ERRPT (error) XERROR (error log data) UNIX

Performance PERF (VM) (processorperformance)

XPERF_CPU (processor usage data) UNIX

Performance DISK (disk performance) XPERF_DISK (disk space usage data) UNIX

Performance PERF (IO) (disk I/Operformance)

XPERF_DISKIO (disk I/O data) UNIX

Performance PERF (PS) (pagingperformance)

XPERF_PAGING (paging space usagedata)

UNIX

Notes::

¶ Log file names also contain the name of the UNIX node, and the date on which the log file was created.

¶ The fields contained in each Performance Reporter record definition are described in “Log Record Definitions” onpage 109. You can also display these fields, when you do the following:

1. Select 2 (Performance Reporter Administration), from the Performance Reporter Primary Menu.

2. Select 3 (Logs).

3. Select the log definition UNIX.

4. Select the record definition you require, and the fields will be displayed.

The UNIX Performance component does not use Performance Reporter record procedures.

40 Release 5

UNIX Accounting subcomponent data flow

Figure 10 describes the data flow for the accounting component.

The processing steps shown in Figure 10 on page 41 are:

�1� Agents process UNIX files pacct, wtmp, dtmp, qacct and gather UNIX accountingdata in log files PACCT, WTMP, DTMP, and QACCT. Each log file name also contains thedate when the log file data was created, and the name of the node.

�2� Transmit log file to Performance Reporter.

�3� Collect log data set information into Performance Reporter tables, using informationfrom the log definition, record definitions, and control tables.

�4� Create reports, using lookup table information.

1

2

2

4

UNIX Nodes

Log filesPACCTWTMPDTMPQACCT

Performance Reporter system(OS/390)

Log data sets:prefix.node_name.EPDMUNIX.Dyymmdd

UNIXLog Defintion:

Record Definitions:XACCT_COMMANDXACCT_CONNECTXACCT_DISKXACCT_PRINT

XACCT_COMMAND_DXACCT_COMMAND_MXACCT_CONNECT_DXACCT_CONNECT_MXACCT_DISK_DXACCT_DISK_MXACCT_PRINT_DXACCT_PRINT_M

Control Tables:SPECIAL_DAYDAY_OF_WEEKPERIOD_PLAN

Lookup Tables:XACCT_PRINT_TYPEX_NODE_NAME:

Tables:

P.R."Collect"

Reports

3

Lookup table:XACCT_CINNECT_TYPE

Figure 10. UNIX Accounting component data flow


3.D

ataan

dR

epo

rterO

bjects

Where to look for further information

Table 5. Further information for accounting component data flowFor details of: Turn to:

Steps �1�, �2�, �3�, and �4� page 38

The relationship between log files, log definition, record definitions Table 4 on page 40

A description of record definitions XACCT_COMMAND, XACCT_CONNECT,XACCT_DISK, and XACCT_PRINT

page 109

XACCT_COMMAND_D and XACCT_COMMAND_M tables page 51

XACCT_CONNECT_D and XACCT_CONNECT_M tables page 54

XACCT_DISK_D and XACCT_DISK_M tables page 52

XACCT_PRINT_D and XACCT_PRINT_M tables page 53

XACCT_CONNECT_TYPE lookup table page 63

XACCT_PRINT_TYPE lookup table page 65

X_NODE_NAME lookup table page 64

Accounting component reports page69

Note: Control tables are explained in the Administration Guide.

42 Release 5

UNIX Configuration subcomponent data flow

Figure 11 describes the data flow for the configuration component (which supports AIXsystems only).


�1� Agents issue commands lsdev and lslpp and gather configuration data in log filesCONF. Each log file name also contains the date when the log file data was created,and the name of the node. Log files have the following record sub-types:

¶ A,D,S for hardware information (HW)

¶ H,I,L for software information (SW)

�2� Transmit log files to Performance Reporter.


�4� Create reports (also using lookup table information).

UNIX Nodes

Log files:CONF


Log definition:UNIXRecord definitions:XCONFIG_HWXCONFIG_SW

Tables:XCONFIG_HWXCONFIG_SW

Reports


P.R.“Collect”

1

2

3

4

Figure 11. UNIX Configuration subcomponent data flow


3.D

ataan

dR

epo

rterO

bjects


Table 6. Further information for configuration component data flowFor details of: Turn to:

Steps �1�, �2�, �3�, and �4� page 38

The relationship between log files, log definition, and record definitions Table 4 on page 40

A description of record definitions XCONFIG_HW and XCONFIG_SW page 109

XCONFIG_HW table page 56

XCONFIG_SW table page 57

Configuration component reports page 82

44 Release 5

UNIX Error subcomponent data flow

Figure 12 describes the data flow for the error component (which supports AIX systemsonly).


�1� Agents issue the command errpt and gather error data in log files ERRPT. Each logfile name also contains the date when the log file data was created, and the name ofthe node.


�3� Collect log data set information into Performance Reporter tables (also using controltable and lookup table information).


UNIX Nodes

Log files:ERRPT


Log definition:UNIXRecord definitions:XERROR

Tables:XERROR_DXERROR_M

CONTROL Tables:SPECIAL_DAYDAY_OF_WEEKPERIOD_PLANLookup Table:X_NODE_NAME

Reports


P.R.“Collect”

1

2

3

4

Figure 12. UNIX Error subcomponent data flow


3.D

ataan

dR

epo

rterO

bjects


Table 7. Further information for error component data flowFor details of: Turn to:

Steps �1�, �2�, �3�, and �4� page 38

The relationship between log file, log definition, and record definition Table 4 on page 40

A description of record definition XERROR page 109

XERROR_D and XERROR_M tables page 58


Error component reports page 89

Note: Control tables are explained in Administration Guide.

46 Release 5

UNIX Performance subcomponent data flow

Figure 13 describes the data flow for the performance component.


�1� Agents issue AIX commands df, lsvg, lsps, and UNIX commands iostat, vmstat, andgather performance data in log files PERF and DISK. Each log file name alsocontains the date when the log file data was created, and the name of the node. Logfiles have the following record sub-types:

¶ IO,VM,PS for PERF

¶ FS,VG for DISK (AIX only)

1

2

3

4

UNIX Nodes

Log files:PERFDISK



Log Defintion:UNIXRecord Definitions:XPERF_CPUXPERF_DISK_FSXPERF_DISK_VGXPERF_DISKIDXPERF_PAGING

XPERF_CPU_HXPERF_CPU_DXPERF_CPU_MXPERF_DISK_DXPERF_DISK_MXPERF_DISKID_HXPERF_DISKID_DXPERF_DISKID_MXPERF_PAGING_HXPERF_PAGING_DXPERF_PAGING_M

Control Tables:SPECIAL_DAYDAY_OF_WEEKPERIOD_PLAN

Lookup Table:X_NODE_NAME:

Tables:

P.R."Collect"

Reports

Figure 13. UNIX Performance subcomponent data flow


3.D

ataan

dR

epo

rterO

bjects


�3� Collect log data set information into Performance Reporter tables (also using controltable information).

�4� Create reports (also using lookup table information).


Table 8. Further information for performance component data flowFor details of: Turn to:

Steps �1�, �2�, �3�, and �4� page 38

The relationship between log files, log definition, and record definitions Table 4 on page 40

A description of record definitions XPERF_CPU, XPERF_DISK_FS,XPERF_DISK_VG, XPERF_DISKIO, and XPERF_PAGING

page 109

XPERF_CPU_H, XPERF_CPU_D, and XPERF_CPU_M tables page 59

XPERF_DISK_D and XPERF_DISK_M tables page 60

XPERF_DISKIO_H, XPERF_DISKIO_D, and XPERF_DISKIO_M tables page 61

XPERF_PAGING_H, XPERF_PAGING_D, and XPERF_PAGING_M tables page 62


Performance component reports page 95

Note: Control tables are explained in the Administration Guide.

48 Release 5

Data, Lookup, and Control Tables

The Performance Reporter database is a collection of DB2 tables, where each table containsa fixed number of columns. The number of rows in each table varies with time, because ofrows added by the collect function and because of database maintenance.


¶ The format that is used for defining UNIX Performance component table names andviews

¶ The layout of the tables that are used in this chapter to describe each tableUNIXPerformance component

¶ The tables used by the accounting subcomponent

¶ The table used by the configuration subcomponent

¶ The tables used by the error subcomponent

¶ The tables used by the performance subcomponent

¶ The lookup tables used by some of the subcomponents

Note: For descriptions of common data tables used by theUNIX Performance componentand other Performance Reporter features, refer to the Administration Guide

Naming standard for tablesThe names of UNIX Performance component tables use this format:

Xsubcomponent_content_suffix

where:

¶ component identifies the component (for example, ACCT for the accountingcomponent).

¶ content is a description (for example, XACCT_COMMAND for the accountingsubcomponent command statistics).

¶ suffix indicates the summarization level of the data in the table (for example,XACCT_COMMAND_D for command statistics summarized by day). Table names forthe configuration subcomponent do not contain suffixes.

A table name can have these summarization-level suffixes:_H The table holds data summarized by hour(hourly data)._D The table holds data summarized by day(daily data)._M The table holds data summarized by month(monthly data).

4


4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

Lookup tables and control tables do not have a suffix; control tables also do not have aprefix.

The names of UNIX Performance component views use the format consisting of the tablename from which the view is taken, followed by a V. For example, the viewXACCT_DISK_MV is taken from the table XACCT_DISK_M.

Table descriptionsEach table description includes information about the table, a description of each of the keycolumns, and a description of each of the data columns:

¶ Key columns are marked with a: k. They are sorted in the sequence they appear in thetable.

¶ Data columns follow the last key column and are sorted in alphabetical order with theunderscore ignored.

The descriptions of most key columns and data columns contain references to the fieldsfrom which they are derived in the record (for example, “From AC_UID”). For anexplanation of such fields, refer to the applicable product documentation.

For each subcomponent, the tables appear in alphabetical order, with underscores andsuffixes ignored.

Tables with similar contents (that is, tables with the same name but with different suffixes)are described under one heading. For example, the heading “XACCT_COMMAND_D, _M”covers two similar tables: XACCT_COMMAND_D and XACCT_COMMAND_M. Exceptfor the DATE column, the contents of these tables are identical. Differences that exist in thecontents of similar tables are explained in the column descriptions.

50 Release 5

Tables in the accounting subcomponentThis section describes the following accounting component tables:

¶ XACCT_COMMAND_D, _M.

¶ “XACCT_DISK_D, _M” on page 52.

¶ “XACCT_PRINT_D, _M” on page 53.

¶ “XACCT_CONNECT_D, _M” on page 54.

¶ “XACCT_DISK_MV” on page 55.

XACCT_COMMAND_D, _MThese tables provide daily and monthly statistics on UNIX commands, and resourceconsumption. They contain command accounting data from UNIX records with record type’PACCT’.

The default retention periods are:30 days for XACCT_COMMAND_D765 days for XACCT_COMMAND_M

Column name Data type Description

DATE k DATE Date when the records were written. For XACCT_COMMAND_M, this isthe first day of the month. From DTE.

NODE_NAME k CHAR(16) Node name. From NODE_NAME.

PERIOD_NAME k CHAR(8) Name of the period. This is derived using fields NODE_NAME, DTE andTME from the record as parameters in the PERIOD function.

USER_GROUP k CHAR(8) Group name. From AC_GID.

NODE_GROUP k CHAR(16) Node group. From ACCOUNTING_GROUP in the X_NODE_NAMElookup table. This is derived using field NODE_NAME from the recordas key. If no match is found, this column is set to ’?’.

USER_NAME k CHAR(8) User name. From AC_UID.

TTY k CHAR(8) Control typewriter. From AC_TTY.

COMMAND k CHAR(8) Command name. From AC_COMM.

COMMANDS INTEGER Total number of commands. This is the count of records.

ELAPSED_SEC FLOAT Total elapsed time, in seconds. Calculated as the sum of AC_ETIME/64.

IO_CHARS INTEGER Total characters transferred. This is the sum of AC_IO.

MEMORY_BYTES FLOAT Total memory usage, in bytes. This value should be divided byCOMMANDS to get the average memory usage. This is the sum ofAC_MEM.

MEMORY_KB_SEC FLOAT Total memory usage, in kilobyte-seconds. Calculated as the sum of(AC_MEM/1024)*(AC_ETIME/64).

RW_BLOCKS INTEGER Total blocks read/written. This is the sum of AC_RW.

SYSTEM_SEC FLOAT Total system time, in seconds. Calculated as the sum of AC_STIME/64.

USER_SEC FLOAT Total user time, in seconds. Calculated as the sum of AC_UTIME/64.


4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

XACCT_DISK_D, _MThese tables provide daily and monthly statistics on UNIX disk usage. They contain diskaccounting data from UNIX records with record type ’DTMP’.

The default retention periods are:30 days for XACCT_DISK_D765 days for XACCT_DISK_M


DATE k DATE Date when the records were written. For XACCT_DISK_M, this is thefirst day of the month. From DTE.


USER_NAME k CHAR(8) User name. From USER_NAME.

NODE_GROUP k CHAR(16) Node group. From ACCOUNTING_GROUP in the X_NODE_NAMElookup table. This is derived using field NODE_NAME from therecord as key. If no match is found, this column is set to ’?’.

DISK_BLOCKS_AVG FLOAT Average number of disk blocks. This is the average of DISK_BLOCKSover RECORDS_COLLECTED.

DISK_BLOCKS_SUM FLOAT Total number of disk blocks. This is the sum of DISK_BLOCKS.Before using this column, divide it by the value in theRECORDS_COLLECTED column.

RECORDS_COLLECTED INTEGER Number of records. This is the count of records.

52 Release 5

XACCT_PRINT_D, _MThese tables provide daily and monthly statistics on UNIX printing. They contain printaccounting data from UNIX records with record type ’QACCT’.

The default retention periods are:30 days for XACCT_PRINT_D765 days for XACCT_PRINT_M


DATE k DATE Date when the records were written. For XACCT_PRINT_M, this is the firstday of the month. From DTE.


PERIOD_NAME k CHAR(8) Name of the period. This is derived using fields NODE_NAME, DTE andTME from the record as parameters in the PERIOD function.

USER_NAME k CHAR(8) User name. From the first word of PR_USER.

PRINT_QUEUE k CHAR(8) Print queue name. From PRINT_QUEUE.

PRINT_TYPE k CHAR(8) Print queue type. From PRINT_TYPE in the XACCT_PRINT_TYPE lookuptable. This is derived using field PRINT_QUEUE from the record as key. If nomatch is found, this column is set to the value in field PRINT_QUEUE.

NODE_GROUP k CHAR(16) Node group. From ACCOUNTING_GROUP in the X_NODE_NAME lookuptable. This is derived using field NODE_NAME from the record as key. If nomatch is found, this column is set to ’?’.

REQUEST_NODE k CHAR(32) Node name of requesting node. From the second word of PR_USER. If novalue is found the printout is local and this column is set to NODE_NAME.

PAGES INTEGER Number of pages. This is the sum of PAGES.

PRINTOUTS INTEGER Number of printouts. This is the count of records.


4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

XACCT_CONNECT_D, _MThese tables provide daily and monthly statistics on UNIX users and their connections. Theycontain connect time accounting data from UNIX records with record type ’WTMP’.

The default retention periods are:30 days for XACCT_CONNECT_D765 days for XACCT_CONNECT_M


DATE k DATE Date when the records were written. For XACCT_CONNECT_M, this is the firstday of the month.


PERIOD_NAME k CHAR(8) Name of the period. This is derived using fields NODE_NAME, DTE and TMEfrom the record as parameters in the PERIOD function.

USER_NAME k CHAR(8) User login name. From UT_USER.

CONNECT_TYPE k CHAR(2) Connect type code. From UT_TYPE.

CONNECTS INTEGER Total number of connections. This is the count of records.

54 Release 5

XACCT_DISK_MVThis view provides monthly UNIX disk accounting statistics for the ’root’ user. It is basedon the XACCT_DISK_M table.


DATE DATE Date when the record was written. From DTE. This is the date of thefirst day of the month.

NODE_NAME CHAR(16) Node name. From NODE_NAME.



4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

Tables in the configuration subcomponentThis section describes the following configuration component tables:

¶ XCONFIG_HW

¶ XCONFIG_SW

XCONFIG_HWThis table provides AIX hardware configuration data. It contains hardware configuration datafrom AIX records with record type ’CONF’ and subtype ’A’, ’D’ or ’S’.

The default retention period for this table is 765 days.


DATE k DATE Date when the record was written. From DTE.

TIME k TIME Time when the record was written. From TME.


DEVICE_CLASS k CHAR(16) Device class. From DEVICE_CLASS.

DEVICE_SUBCLASS k CHAR(16) Device subclass. From DEVICE_SUBC.

DEVICE_NAME k CHAR(16) Device name. From DEVICE_NAME.

DEVICE_TYPE k CHAR(16) Device type. From DEVICE_TYPE.

STATUS k CHAR(2) Device status. From RECORD_SUBTYPE.

56 Release 5

XCONFIG_SWThis table provides AIX software configuration data. It contains software configuration datafrom AIX records with record type ’CONF’ and subtype ’H’, ’I’ or ’L’.

The default retention period for this table is 765 days.


DATE k DATE Date when the record was written. From DTE.

TIME k TIME Time when the record was written. From TME.


SOFTWARE_OBJECT k CHAR(32) Software object name. From FILE_NAME.

PATH k CHAR(32) Path name for object. From FILE_NAME.

DESCRIPTION VARCHAR(80) Product description. From FILE_NAME.

FEATURE CHAR(4) Feature ID. From FEATURE_ID.

INSTALL_DATE CHAR(8) Product installation date. From INST_DATE.

INSTALL_STATUS CHAR(10) Product installation status. From STATUS.

PRODUCT_ID CHAR(10) Product ID. From PRODUCT_ID.

PRODUCT_NAME CHAR(20) Product name. From PRODUCT_NAME.

PRODUCT_STATE CHAR(1) Product state. From STATE.

RELEASE CHAR(15) Product release number. From RELEASE.


4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

Tables in the error subcomponentThis section describes the following error component table.

XERROR_D, _MThese tables provide daily and monthly statistics on AIX error log. They contain error datafrom AIX records with record type ’ERRPT’.

The default retention periods are:30 days for XERROR_D765 days for XERROR_M


DATE k DATE Date when the error occurred. From DTE. For XERROR_M, this is thefirst day of the month.


PERIOD_NAME k CHAR(8) Name of the period. This is derived using fields NODE_NAME, DTEand TME from the record as parameters in the PERIOD function.

ERROR_ID k CHAR(8) Error identification. From ERROR_ID.

ERROR_TYPE k CHAR(1) Error type. From ERROR_TYPE.

ERROR_CLASS k CHAR(1) Error class. From ERROR_CLASS.

RESOURCE_NAME k CHAR(14) Resource name. From RESOURCE_NAME.

DESCRIPTION VARCHAR(40) Error description. From DESCRIPTION.

ERRORS INTEGER Number of errors. This is the count of records.

58 Release 5

Tables in the performance subcomponentThis section describes the following performance subcomponent tables:

¶ XPERF_CPU_H, _D, _M

¶ XPERF_DISK_D, _M

¶ XPERF_DISKIO_H, _D, _M

¶ XPERF_PAGING_H, _D, _M

XPERF_CPU_H, _D, _MThese tables provide hourly, daily, and monthly statistics on UNIX utilization. They containCPU and memory usage, paging rate and number of users and processes from UNIX recordswith record type ’PERF’ and subtype ’VM’.

The default retention periods are:7 days for XPERF_CPU_H30 days for XPERF_CPU_D765 days for XPERF_CPU_M


DATE k DATE Date when the records were written. For XPERF_CPU_M, this is thefirst day of the month. From DTE.

TIME k TIME Time (rounded down to the nearest hour) when the record was written.From TME. This field is present only in XPERF_CPU_H.



CPU_USAGE_PCT_AVG FLOAT Average CPU usage, in percent. This is the average of CPU_AVG.

CPU_USAGE_PCT_MAX INTEGER Maximum CPU usage, in percent. This is the maximum of CPU_MAX.

CPU_USAGE_PCT_MIN INTEGER Minimum CPU usage, in percent. This is the minimum of CPU_MIN.

MEASURED_SEC INTEGER Total measured time, in seconds. This is the sum of INTERVAL*60.

MEM_FREE_PAGES_AVG FLOAT Average number of free pages. This is the average of MEMORY_AVG.

MEM_FREE_PAGES_MAX INTEGER Maximum number of free pages. This is the maximum ofMEMORY_MAX.

MEM_FREE_PAGES_MIN INTEGER Minimum number of free pages. This is the minimum ofMEMORY_MIN.

PAGING_RATE_AVG FLOAT Average number of pages paged IN/OUT per second. This is theaverage of PAGING_AVG.

PAGING_RATE_MAX INTEGER Maximum number of pages paged IN/OUT per second. This is themaximum of PAGING_MAX.

PAGING_RATE_MIN INTEGER Minimum number of pages paged IN/OUT per second. This is theminimum of PAGING_MIN.

PROCESSES_AVG FLOAT Average number of processes. This is the average of PROCESSES.

PROCESSES_MAX INTEGER Maximum number of processes. This is the maxium of PROCESSES.

PROCESSES_MIN INTEGER Minimum number of processes. This is the minium of PROCESSES.

RECORDS_COLLECTED INTEGER Total number of records. This is the count of records.

USERS_AVG FLOAT Average number of users. This is the average of USERS.

USERS_MAX INTEGER Maximum number of users. This is the maximum of USERS.

USERS_MIN INTEGER Minimum number of users. This is the minimum of USERS.


4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

XPERF_DISK_D, _MThese tables provide daily and monthly statistics on AIX disk usage. They contain diskperformance data from AIX records with record type ’DISK’ and subtype ’FS’ or ’VG’.

The default retention periods are:30 days for XPERF_DISK_D765 days for XPERF_DISK_M


DATE k DATE Date when the records were written. For XPERF_DISK_M, this is thefirst day of the month. From DTE.



SPACE_TYPE k CHAR(2) Space type, FS or VG. From RECORD_SUBTYPE.

VOLUME k CHAR(16) Volume. From FILE_SYSTEM for space type FS and fromVOLUME_GROUP for space type VG.

FILE_SYSTEM k CHAR(16) File system. From FILE_SYSTEM for space type FS and set to blanksfor space type VG.

RECORDS_COLLECTED INTEGER Number of samples. This is the count of records.

SPACE_FREE_MB FLOAT Free space, in MB, accumulated for all samples. This is the sum ofFREE_SPACE_MB for space type VG and calculated as the sum ofFREE_SPACE_KB/1024.0 for space type FS. This value should bedivided by RECORDS_COLLECTED to get the average free space.

SPACE_SIZE_MB FLOAT Total size of space, in MB, accumulated for all samples. This is the sumof TOTAL_SPACE_MB for space type VG and calculated as the sum ofTOTAL_SPACE_KB/1024.0 for space type FS. Divide this value byRECORDS_COLLECTED to get the average value.

SPACE_USED_PCT FLOAT Used space, in percent, accumulated for all samples. Calculated as thesum of 100*USED_SPACE_MB/TOTAL_SPACE_MB space type VGand as the sum of USED_SPACE_PCT for space type FS. Divide thisvalue by RECORDS_COLLECTED to get the average value.

60 Release 5

XPERF_DISKIO_H, _D, _MThese tables provide hourly, daily, and monthly statistics on UNIX disk I/O. They containdisk I/O performance data from UNIX records with record type ’PERF’ and subtype ’IO’.

The default retention periods are:7 days for XPERF_DISKIO_H30 days for XPERF_DISKIO_D765 days for XPERF_DISKIO_M


DATE k DATE Date when the records were written. For XPERF_DISKIO_M, this isthe first day of the month. From DTE.

TIME k TIME Time (rounded down to the nearest hour) when the record was written.From TME. This field is present only for XPERF_DISKIO_H.



DISK k CHAR(10) Physical disk name. From DISK.

DISK_BUSY_PCT_AVG FLOAT Average disk busy, in percent. This is the average of BUSY.

DISK_BUSY_PCT_MAX INTEGER Maximum disk busy, in percent. This is the maximum of BUSY.

MEASURED_SEC INTEGER Total measurement time, in seconds. Calculated as the sum ofINTERVAL*60.

READ_KB_TOTAL INTEGER Amount of data read, in kilobytes. This is the sum of TOTREAD.

READ_MAXRATE FLOAT Maximum amount of data read, in kilobytes per second. Calculated asthe maximum of TOTREAD/(INTERVAL*60.0).


WRITE_KB_TOTAL INTEGER Amount of data written, in kilobytes. This is the sum of TOTWRITE.

WRITE_MAXRATE FLOAT Maximum amount of data written, in kilobytes per second. Calculatedas the maximum of TOTWRITE/(INTERVAL*60.0).

RW_KB_TOTAL FLOAT Amount of data written, in kilobytes. Calculated as the sum of TOTRW.

RW_MAXRATE INTEGER Maximum amount of data written, in kilobytes per second. Calculatedas the sum of TOTRW/(INTERVAL*60.0).


4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

XPERF_PAGING_H, _D, _MThese tables provide hourly, daily, and monthly statistics on UNIX page space. They containpage space performance data from UNIX records with record type ’PERF’ and subtype ’PS’.

The default retention periods are:7 days for XPERF_PAGING_H30 days for XPERF_PAGING_D765 days for XPERF_PAGING_M


DATE k DATE Date when the records were written. For XPERF_PAGING_M, this isthe first day of the month. From DTE.

TIME k TIME Time (rounded down to the nearest hour) when the record was written.From TME. This field is present only for XPERF_PAGING_H.

NODE_NAME k CHAR(16) Node name. From NODE_NAME


DISK k CHAR(10) Physical disk name. From DISK.

PAGE_SPACE k CHAR(10) Page space name. From PAGE_SPACE.

PS_SIZE_AVG_KB FLOAT Average page space size, in KB. This is the average of PSSIZE.

PS_USED_AVG_KB FLOAT Average page space used, in KB. This is the average of PSUSE.


62 Release 5

Lookup tablesThis section describes the following lookup tables specific to the UNIX Performancecomponent:

¶ XACCT_CONNECT_TYPE

¶ XNODE_NAME

¶ XACCT_PRINT_TYPE

For descriptions of common lookup tables used by the UNIX Performance component andother Performance Reporter features, refer to the Administration Guide.

XACCT_CONNECT_TYPEThis lookup table contains descriptions of connect types.


CONNECT_TYPE k CHAR(2) Connect type code.

CONNECT_DESC CHAR(16) Connect type description.

Example of table contents

PROCESS PROCESSCODE TYPE------- ----------------00 EMPTY01 RUN_LVL02 BOOT_TIME03 OLD_TIME04 NEW_TIME05 INIT_PROCESS06 LOGIN_PROCESS07 USER_PROCESS08 DEAD_PROCESS09 ACCOUNTING


4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

X_NODE_NAMEThis lookup table contains node names and their associated accounting groups, period plansand TCP/IP address.


NODE_NAME k CHAR(16) Node name. This is an alias name for the node. This may contain globalsearch characters.

ACCOUNTING_GROUP CHAR(16) Accounting group.

INTERNET_ADDRESS CHAR(15) Internet address. This may contain global search characters.

PERIOD_PLAN CHAR(8) Period plan. This value can be used as a parameter when using thePERIOD function looking up PERIOD_NAME.


NODE ACCOUNTING PERIOD INTERNETNAME GROUP PLAN ADDRESS---------------- ---------------- -------- ---------------node1 GROUP1 BASE 9.52.50.7node2 GROUP2 PLAN1 9.52.50.8node3 GROUP2 PLAN2 9.52.50.9node5 GROUP3 PLAN3 9.52.50.11% DEFAULT DEFAULT 9.99.99.99

64 Release 5

XACCT_PRINT_TYPEThis lookup table converts print-queues to a print-type. It updates the XACCT_PRINT_Dtable.


NODE_NAME k CHAR(16) Node name. This may contain global search characters.

PRINT_QUEUE k CHAR(8) Print queue name. This may contain global search characters.

PRINT_TYPE CHAR(8) Print queue type.


NODE PRINT PRINTNAME QUEUE TYPE---------------- -------- --------node1 prt01 dep01node2 prt01 dep01node3 prt05 dep02% prt99 dep99


4.D

ata,L

oo

kup

,an

dC

on

trol

Tables

66 Release 5

Reports

The reporting function produces reports based on the data in the Performance Reporterdatabase. Reports can show data from tables or from views. You can request reports onlineor by submitting batch jobs. Typically, you use online reporting for reports that you useonce, and batch reporting for regularly required reports.


¶ The format of the names used to define each report, and how source tables, attributesand variables are used.

¶ The reports in the accounting subcomponent

¶ The reports in the configuration subcomponent

¶ The reports in the error subcomponent

¶ The reports in the performance subcomponent

Report format and general descriptionPerformance Reporter presents reports in tables and graphs. All reports have the same basicreport layout. This section describes the elements that are common among PerformanceReporter feature reports:¶ Report ID¶ Report group¶ Source¶ Attributes¶ Variables

Report IDPerformance Reporter assigns each report a unique identifier. The UNIX Performancecomponent uses this format for report IDs:

Xyxx

Where:

¶ y is:ACCT

The accounting subcomponentCONFIG

The configuration subcomponentERROR

The error subcomponentPERF The performance subcomponent

5


5.R

epo

rts

¶ xx is a sequential number identifying the report.

Examples:XACCT01XPERF02

Report groupPerformance Reporter uses several predefined report groups. For the UNIX Performancecomponent, each subcomponent has one group. The four UNIX Performance componentreport groups are given on page 4.

SourceEach report contains information from one or more source tables. The report descriptions inthis chapter list source tables. Refer to these source tables if you are interested in learningwhere certain data originates.

AttributesEach report has certain attributes associated with it. Use these attributes as keywords tosearch for specific reports in the dialogs.

You can specify any number of attributes for a report, but the area to which the reportbelongs (for example, UNIX) is always present for predefined reports.

You can also specify these attributes, when appropriate:

¶ Resource types, such as storage or processor

¶ Performance issues, such as availability or response

¶ Presentation forms, such as detail, overview, or trend

¶ Time resolutions, such as hourly, daily, or monthly

VariablesEach report has variables associated with it. You specify the values for these variables whenyou generate the report using the reporting dialog.

When you specify a date for a monthly report, specify the first day of the month. Otherwise,there is no match in the data table.

If a character variable happens to have only numeric characters, enclose it in single quotemarks, otherwise it will not match the data. For example, if you have a system ID of 1234,specify it as ’1234’ on the Variables window.

68 Release 5

Reports in the accounting subcomponentThis section describes the following accounting subcomponent reports:

¶ “UNIX Acct Commands by User, Daily Overview” on page 70.

¶ “UNIX Acct Users by Command, Daily Overview” on page 72.

¶ “UNIX Acct Cmd Resource Consumption, Daily Overview” on page 74.

¶ “UNIX Acct User Resource Usage, Monthly Overview” on page 76.

¶ “UNIX Acct Disk Blocks by User, Monthly Overview” on page 77.

¶ “UNIX Acct Disk Blocks in 1000s, Monthly Trend” on page 78.

¶ “UNIX Acct Users and Connects, Daily Overview” on page 79.

¶ “UNIX Acct Printed Pages by User, Monthly Overview” on page 80.

¶ “UNIX Acct Printed Pages by System, Monthly Overview” on page 81.

The data flow for the accounting subcomponent (including the names of log files,Performance Reporter records and tables) is given in Figure 10 on page 41.


5.R

epo

rts

UNIX Acct Commands by User, Daily OverviewFor a specific node in the network, this report (see Figure 14) provides daily overviewinformation about the resource consumption by user name and command name. The report isproduced by period name (for example, PRIME or NIGHT). You can use the report toidentify and have control of, the amount of total resources that each user is consuming.

You can easily amend the information provided in the report to suit your own requirements.For example, you can add the System time and User time values, to have a measure of totalprocessor resource.

The report can be used as a guide for charging users and departments for the systemresources they have used.

This information identifies the report:

Report ID XACCT01

Report group Accounting Subcomponent Reports

Source XACCT_COMMAND_D, (described on page 51)

Attributes UNIX, Acct, Accounting, Command, Cmd, User, Daily, Overview

Variables Date, Node name, Period name

The report contains this information:

User name The name of the user who has used this node.

Command name The name of the UNIX command entered by the user.

Commands (count) The number of times this user has entered this command.

Elapsed time (hour) The elapsed time (in hours) used by this user, entering thiscommand. This is calculated as ELAPSED_SEC/3600.

UNIX Acct Commands by User, Daily OverviewDate: 1999-12-15 Period: 'PRIME '

Node name: 'sun222 '

| Elapsed System UserUser Command | Commands time time time Cmd mem IO RW blocksname name | (count) (hour) (sec) (sec) avg (KB) (KB) (count)-------- --------| -------- -------- -------- -------- -------- -------- ---------smith aixterm | 9 0.02 1 3 0.23 7 3

awk | 33 0.09 5 5 0.55 14 5. | . . . . . . .. | . . . . . . .xset | 14 0.10 4 10 0.78 18 9

| -------- -------- -------- -------- -------- -------- ---------Total| 314 2.19 58 55 5.15 92 80

|root awk | 18 0.80 7 5 0.15 13 18

bsh | 38 0.21 8 11 0.89 22 68. | . . . . . . .. | . . . . . . .xmservd | 12 1.17 21 24 1.19 51 10

| -------- -------- -------- -------- -------- -------- ---------Total| 104 10.18 112 122 5.08 103 130

|| ======== ======== ======== ======== ======== ======== =========

Total| 418 12.37 170 177 10.23 195 210

Performance Reporter Report: XACCT01

Figure 14. Example of UNIX Acct Commands by User, Daily Overview

70 Release 5

System time (sec) The system time (in seconds) used by this user, entering thiscommand.

User time (sec) The user time (in seconds) used by this user, entering thiscommand.

Cmd mem avg (KB) The average memory (in kilobytes) used by this user,entering this command. This is calculated asMEMORY_BYTES/(1024*COMMANDS).

IO (KB) Characters (in kilobytes) that have been transferred by thisuser, using this command. This is calculated asIO_CHARS/1024.

RW blocks (count) The number of blocks that have been read/written by thisuser, using this command.


5.R

epo

rts

UNIX Acct Users by Command, Daily OverviewFor a specific node in the network, this report (see Figure 15) provides daily overviewinformation about command usage by each user; the number of times a user has used acommand, how much system and user time the command required, how much memory, andhow much I/O. The report is produced by period name (for example, PRIME or NIGHT).

You can use the report to identify and have control of, the amount of resources that eachuser is consuming, by command. The information provided in the report can be amended tosuit your own requirements. For example, by modifying the SQL routine that produces thereport, you can remove from the report users whose command usage is not relevant.



Report ID XACCT02






Command name The name of the UNIX command.


Commands (count) The number of times this user has entered this command.




Cmd mem avg (KB) The memory (in kilobytes) used by this user, entering this

UNIX Acct Users by Command, Daily OverviewDate: 1999-12-16 Period: 'PRIME '

Node name: 'hp555 '

| Elapsed System UserCommand User | Commands time time time Cmd mem IO RW blocksname name | (count) (hour) (sec) (sec) avg (KB) (KB) (count)-------- --------| -------- -------- -------- -------- -------- -------- ---------aixterm smith | 9 0.05 2 3 0.23 11 2awk smith | 7 0.07 3 2 0.07 2 4

root | 18 0.08 2 1 0.15 13 7bind smith | 28 0.10 6 4 0.44 10 3. . . . . . . . .. . . . . . . . .


Figure 15. Example of UNIX Acct Users by Command, Daily Overview

72 Release 5

command. This is calculated asMEMORY_BYTES/(1024*COMMANDS).

IO (KB) Characters (in kilobytes) that have been transferred by thisuser, using this command. This is calculated asIO_CHARS/1024.

RW blocks (count) The number of blocks that have been read/written by thisuser, using this command.


5.R

epo

rts

UNIX Acct Cmd Resource Consumption, Daily OverviewFor a specific node in the network, this report (see Figure 16) provides daily overviewinformation about command usage; the number of times a command has been used, howmuch system and user time the command required, how much memory, and how much I/O.The report is produced by period name (for example, PRIME or NIGHT). You can use thereport to identify and have control of, the amount of total resources that each command isconsuming.

You can easily amend the information provided in the report to suit your own requirements.For example, by modifying the SQL routine that produces the report, you can remove fromthe report commands that are not particularly important. Or you can compare the amount ofresources (for example, System time + User time) a command requires for different nodes. Ifcommand xyz requires more System and User time at ’nodeA’ than at ’nodeB’, you couldinvestigate, for example, if paging is too high at the first node, or if the command has toprocess more data at the first node.


Report ID XACCT03






Command name The name of the UNIX command.

Commands (count) The number of times this command has been entered.

Elapsed avg (min) The average elapsed time (in minutes) used by thiscommand, This is calculated asELAPSED_SEC/(COMMANDS*60).

System avg (sec) The average system time (in seconds) used by thiscommand. This is calculated asSYSTEM_SEC/COMMANDS.

UNIX Acct Cmd Resource Consumption, Daily OverviewDate: 1999-12-17 Period: 'PRIME '

Node name: 'aix111 '

Command | Commands Elapsed System User Cmd mem IO RWname | (count) avg (min) avg (sec) avg (sec) avg (KB) avg (KB) avg (block)--------| -------- --------- --------- --------- -------- -------- -----------aixterm | 7 0.08 0.23 0.33 0.23 17.0 5awk | 5 0.03 0.04 0.05 0.13 4.1 2bind | 8 0.02 0.09 0.64 0.44 12.5 3bsh | 26 0.02 0.10 0.03 0.07 4.4 1. | . . . . . . .. | . . . . . . .


Figure 16. Example of UNIX Acct Cmd Resource Consumption, Daily Overview

74 Release 5

User avg (sec) The average user time (in seconds) used by this command.This is calculated as USER_SEC/COMMANDS.

Cmd mem avg (KB) The average memory (in kilobytes) used by this command.This is calculated asMEMORY_BYTES/(1024*COMMANDS).

IO avg (KB) The average number of characters (in kilobytes) used by thiscommand. This is calculated as(IO_CHARS/1024)/COMMANDS.

RW avg (block) The average number of blocks read/written by thiscommand. This is calculated asRW_BLOCKS/COMMANDS.


5.R

epo

rts

UNIX Acct User Resource Usage, Monthly OverviewFor a specific node in the network, this report (see Figure 17) provides monthly overviewinformation about the resource usage by user name. The report would typically be run eachmonth in batch mode, and if the report shows that further investigation for a specific user(or users) is required, you could then use the report “UNIX Acct Commands by User, DailyOverview” on page 70 for a daily breakdown of resource usage by the user(s). The report isproduced by period name (for example, PRIME or NIGHT).



Report ID XACCT04


Source XACCT_COMMAND_M, (described on page 51)

Attributes UNIX, Acct, Accounting, Command, Cmd, User, Monthly, Overview

Variables Month, Node name, Period name



Commands (count) The total number of commands this user has entered.




IO (KB) Characters (in kilobytes) that have been transferred by thisuser, using this command. This is calculated asIO_CHAR/1024.

UNIX Acct User Resource Usage, Monthly OverviewMonth: 1999-12-01 Period: 'PRIME '


| Elapsed System UserUser | Commands time time time IOname | (count) (hour) (sec) (sec) (KB)--------| -------- -------- -------- -------- --------smith | 514 1.19 58 125 3392root | 104 0.18 4 2 317


Figure 17. Example of UNIX Acct User Resource Usage, Monthly Overview

76 Release 5

UNIX Acct Disk Blocks by User, Monthly OverviewFor a specific node in the network, this report (see Figure 18) provides monthly overviewinformation about the average disk blocks allocated by each user, at this node. The reportwould typically be run each month in batch mode.

The report can be used as a guide for charging users and departments for the disk resourcesthey have used.


Report ID XACCT05


Source XACCT_DISK_M, XACCT_DISK_MV (View), (described on page 52)

Attributes UNIX, Acct, Accounting, Disk, User, Monthly, Overview

Variables Month, Node name



Disk blocks avg The average number of blocks allocated by this user. This iscalculated as DISK_BLOCKS/(RECORDS_COLLECTEDfor USER=root).

UNIX Acct Disk Blocks by User, Monthly OverviewMonth: 1999-12-01


User | Disk blocksname | avg--------| -----------adm | 216archer | 4bin | 398644smith | 188daemon | 4guest | 4johanbe | 4loadl | 116root | 103824sys | 8uucp | 972xbill | 132

| ===========Total| 504116


Figure 18. Example of UNIX Acct Disk Blocks by User, Monthly Overview


5.R

epo

rts

UNIX Acct Disk Blocks in 1000s, Monthly TrendFor a specific node in the network, this report (see Figure 19) provides monthly trendinformation about the average disk blocks allocated by the node. The number of blocks isgiven in units of one thousand. You can use the report to anticipate potential bottlenecks indisk space capacity (where, for example, the usage shows a steadily increasing trend).

The report would typically be run each month in batch mode. If you require more detailedinformation about a node’s disk usage for a specific month, you can proceed to the report“UNIX Acct User Resource Usage, Monthly Overview” on page 76.


Report ID XACCT06


Source XACCT_DISK_M, XACCT_DISK_MV (View), (described on page 52)

Attributes UNIX, Acct, Accounting, Disk, Monthly, Trend

Variables From month, To month, Node name


Month The month start date.

Node name The name of the node in the network.

Disk blocks avg (1000s) The average number of blocks (in thousands) used by thisnode. This is calculated as(DISK_BLOCKS/1000)/(RECORDS_COLLECTED forUSER=root).

UNIX Acct Disk Blocks in 1000s, Monthly TrendMonth: '1999-10-01' to '1995-12-01'


| <----------------- Month ------------------->| 1995-10-01 1995-11-01 1995-12-01| Disk blocks Disk blocks Disk blocks

Node name | avg (1000s) avg (1000s) avg (1000s)----------------| ----------- ----------- -----------aix555 | 504 684 824


Figure 19. Example of UNIX Acct Disk Blocks in 1000s, Monthly Trend

78 Release 5

UNIX Acct Users and Connects, Daily OverviewFor a specific node in the network, this report (see Figure 20) provides daily overviewinformation about the connects that have been made at the node. The report is produced byperiod name (for example, PRIME or NIGHT). You can use the report to control the use ofthe node.


Report ID XACCT07


Source XACCT_CONNECT_D, XACCT_CONNECT_TYPE, (described on pages54 and 63 respectively)

Attributes UNIX, Acct, Accounting, Process, User, Daily, Overview




Connect description The connect type description.

Connects (count) The total number of connects for this connect type, this userhas made.

UNIX Acct Users and Connects, Daily OverviewDate: 1999-12-15 Period: 'PRIME '


User | Connect Connectsname | description (count)--------| ---------------- ---------archer | USER_PROCESS 8smith | USER_PROCESS 14cron | INIT_PROCESS 1root | LOGIN_PROCESS 1


Figure 20. Example of UNIX Acct Users and Connects, Daily Overview


5.R

epo

rts

UNIX Acct Printed Pages by User, Monthly OverviewFor a specific node in the network, this report (see Figure 21) provides monthly overviewinformation about how much of the printing resources each user has used. The report isproduced by period name (for example, PRIME or NIGHT). The report can be used as aguide for charging users and departments for the printing resources they have used.


Report ID XACCT08


Source XACCT_PRINT_M, (described on page 53)

Attributes UNIX, Acct, Accounting, Print, Page, User, Monthly, Overview

Variables Month, Node name, Period name



Print queue name The name of the print queue used by this user.

Request node The name of the node from which print was requested.When the request node is the same as the node name forwhich the report is being produced, the print request is local.

Printouts The number of printouts printed by this user, at this node.

Pages Total number of pages printed by this user, at this node.

UNIX Acct Printed Pages by User, Monthly OverviewMonth: 1999-01-01 Period: 'PRIME'


User Print queue Requestname name node Printouts Pages-------- ----------- -------------------------------- --------- --------smith prt0 aix555 2 2

wn5prt aix555 2 2--------- --------

Total 4 4

========= ========Total 4 4


Figure 21. Example of UNIX Acct Printed Pages by User, Monthly Overview

80 Release 5

UNIX Acct Printed Pages by System, Monthly OverviewFor a specific node in the network, this report (see Figure 22) provides monthly overviewinformation about how much of the printing resources each node has used. The report isproduced by period name (for example, PRIME or NIGHT). The report can be used tocontrol the amount of printing being performed by each node.


Report ID XACCT09


Source XACCT_PRINT_M, (described on page 53)

Attributes UNIX, Acct, Accounting, Print, Page, System, Monthly, Overview

Variables Month, Period name



Print queue name The print queue name.

Request node The node from which the print was requested. When therequest node is the same as the node name for which thereport is being produced, the print request is local.

Printouts For the specified request node, the total number of printoutsprinted.

Pages For the specified request node, the total number of pagesprinted.

UNIX Acct Printed Pages by System, Monthly OverviewMonth: 1999-01-01 Period: 'PRIME'

Print queue RequestNode name name node Printouts Pages---------------- ----------- -------- --------- --------sun555 prt0 aix222 2 12

wn5prt aix555 2 52--------- --------

Total 4 64

========= ========Total 4 64


Figure 22. Example of UNIX Acct Printed Pages by System, Monthly Overview


5.R

epo

rts

Reports in the configuration subcomponentThis section describes the following configuration subcomponent reports:

¶ “UNIX Configuration of HW for a System, Overview” on page 83.

¶ “UNIX Configuration of HW for Device Class, Overview” on page 85.

¶ “UNIX Configuration of SW for a System, Overview” on page 86.

¶ “UNIX Configuration of SW for Object, Overview” on page 88.

The data flow for the configuration subcomponent (including the names of log files,Performance Reporter records and tables) is given in Figure 11 on page 43.

82 Release 5

UNIX Configuration of HW for a System, OverviewThis report only covers nodes that use AIX. For a specific node in the network and for atime period determined by the From and To-dates, this report (see Figure 23) providesoverview information about the devices that were found when the last “collect” job was run,information about each device (status, sub-class, type, date and time of configuration) aregiven.


Report ID XCONFIG01

Report group Configuration Subcomponent Reports

Source XCONFIG_HW, (described on page 56)

Attributes AIX, Configuration, HW, Hardware, Overview

Variables From date, To date, Node name


Device class A class of devices configured at this node.

Device name The name of a device contained within the given deviceclass.

Status The status of the device contained within the given deviceclass. The possible values are:

A = ActiveD = Defined

Device subclass The subclass of the device contained within the device class.

Device type The type of the device contained within the device class.

UNIX Configuration of HW for a System, OverviewDate: '1999-11-13' to '1999-12-13'


Device class Device name | Status Device subclass Device type Date Time---------------- ----------------| ------ ---------------- ---------------- ---------- --------adapter ent0 | A sio ient_2 1995-12-13 16.45.12

fda0 | A sio fda_2 1995-12-13 16.45.12ppa0 | A sio ppa 1995-12-13 16.45.12. | . . . . .. | . . . . .. | . . . . .

aio aio0 | D Asynchronous node 1995-12-13 16.45.12bus bus0 | A sys mca 1995-12-13 16.45.12

sgabus0 | A sys sgabus 1995-12-13 16.45.12disk hdisk0 | A scsi 1000mb 1995-12-13 16.45.12diskette fd0 | A siofd fd 1995-12-13 16.45.12en en0 | D Standard EN 1995-12-13 16.45.12fpa fpa0 | A sys fpa1 1995-12-13 16.45.12

. | . . . . .

. | . . . . .

. | . . . . .

Performance Reporter Report: XCONFIG01

Figure 23. Example of UNIX Configuration of HW for a System, Overview


5.R

epo

rts

Date The date when the collect was run, and when theconfiguration data for this device was captured.

Time The time when the collect was run, and when theconfiguration data for this device was captured.

84 Release 5

UNIX Configuration of HW for Device Class, OverviewThis report only covers nodes that use AIX. For a specific device class or for all deviceclasses, this report (see Figure 24) provides overview information about the number ofdevices in the network, found when the collect job(s) was run. If no device class is entered,the number of devices for all device classes in the network, will be displayed.

The information is displayed for a time period determined by From and To dates that youenter.


Report ID XCONFIG02


Source XCONFIG_HW, (described on page 56)

Attributes AIX, Configuration, HW, Hardware, Device, Overview

Variables From date, To date, Device class


Date The date(s) when the collect job was run.


Devices The number of devices that were found by the collect job,on the date(s) the collect was run.

UNIX Configuration of HW for Device Class, OverviewDate: '1999-11-13' to '1995-12-13'

Device class: 'adapter '

<----------- Date --------->1995-11-13 1999-12-13

Node name Devices Devices---------------- ----------- -----------aix555 12 15


Figure 24. Example of UNIX Configuration of HW for Device Class, Overview


5.R

epo

rts

UNIX Configuration of SW for a System, OverviewThis report covers only nodes that use AIX. For a specific node in the network, this report(see Figure 25) provides overview information about the software configurations that havetaken place. The information is displayed for a time period determined by the From and Todates that you enter.


Report ID XCONFIG03


Source XCONFIG_SW, (described on page 57)

Attributes AIX, Configuration, SW, Software, System, Overview

Variables From date, To date, Node name


Feature code The feature code identification of the software object that is installed.

Path name The path to where the software object is located.

Software objectThe name of the software object.

Install date The date when the software object was installed.

Description A description of the software object.

Date The date of the collect run, when the information about the software objectwas obtained.

UNIX Configuration of SW for a System, OverviewDate: '1999-01-01' to '1999-07-0 1'


Feature Software Installcode Path name object date Description Date Time------- ------------------ ------------------ ------- ----------------------------- ---------- --------

/usr/lib/objrepo xmconsole.data 09/13/95 XMconsole configuration files 1996-01-17 09.51.31xmconsole.obj 09/13/95 XMconsole executables 1996-01-17 09.51.31

/etc/objrepo xlccmp.obj 09/10/95 AIX XL C Compiler/6000 1996-01-17 09.51.31/usr/lib/objrepo xlccmp.obj 09/10/95 AIX XL C Compiler/6000 1996-01-17 09.51.31

2924 /usr/lib/objrepo netmgrmEn_US.msg 09/10/95 Network Management I Messages 1996-01-17 09.51.315005 /etc/objrepo netmgr.am.obj 09/10/95 Alert Manager 1996-01-17 09.51.31

netmgr.api.obj 09/10/95 SNMP Appl. Programming Interf 1996-01-17 09.51.31netmgr.clm.obj 09/10/95 SNMP Command Line Manager 1996-01-17 09.51.31netmgr.nvdm.obj 09/10/95 NetView Dist. Manager Catcher 1996-01-17 09.51.31

/etc/objrepo nv6000.features.ob 10/29/95 AIX SystemView NetView/6000 Feat 1996-01-17 09.51.31/usr/lib/objrepo nv6000.base.obj 10/29/95 AIX SystemView NetView/6000 Base 1996-01-17 09.51.31

nv6000.features.ob 10/29/95 AIX SystemView NetView/6000 Feat 1996-01-17 09.51.310000 /usr/lib/objrepo perfmgr 09/10/95 Performance Manager 1996-01-17 09.51.310000 /etc/objrepo perfagent 09/10/95 Performance Agent 1996-01-17 09.51.31

/usr/lib/objrepo perfagent 09/10/95 Performance Agent 1996-01-17 09.51.315005 /usr/lib/objrepo aic.obj 09/10/95 AIXwindows Interface Composer 1996-01-17 09.51.312924 /usr/lib/objrepo bsmEn_US.msg 09/10/95 Base System Messages-U.S. Engl 1996-01-17 09.51.312937 /usr/lib/objrepo bsmSv_SE.msg 09/10/95 Base System Messages-Swedish 1996-01-17 09.51.315005 /etc/objrepo bos.obj 01/01/90 The Base Operating System 1996-01-17 09.51.31

bosext1.ecs.obj 09/10/95 Remote Customer Services 1996-01-17 09.51.31bosext1.mh.obj 09/10/95 Mail Handler 1996-01-17 09.51.31


Figure 25. Example of UNIX Configuration of SW for a System, Overview

86 Release 5

Time The time of the collect run, when the information about the software objectwas obtained.


5.R

epo

rts

UNIX Configuration of SW for Object, OverviewThis report only covers nodes that use AIX. For a specific software object, this report (seeFigure 26) provides overview information about the software objects that exist on thenetwork. If no software object is selected for display, the report will provide informationabout all software objects.

The information is displayed for a time period determined by From and To dates that youenter.


Report ID XCONFIG04


Source XCONFIG_SW, (described on page 57)

Attributes AIX, Configuration, SW, Software, Object, Overview

Variables From date, To date, Object


Software objectThe name of the software object that has been selected.

Node name The node within the network, where the software object is located.

Path name The path(s) to where the software object is located.

Release The release level of the software object.

Install status The status of the installation of the software object.

Install date The date when the software object was installed.

Date The date of the collect run, when the information about the software objectwas obtained.

Time The time of the collect run, when the information about the software objectwas obtained.

UNIX Configuration of SW for Object, OverviewDate: '1999-01-01' to '1999-02-10'

Object: 'bos.obj '

Software Install Installobject Node name Path name Release status date Date Time---------------------------------- ----------------- --------------- ---------- -------- ---------- --------bos.obj aix555 /etc/obj/repo 03.02.0000.0000 COMPLETE 01/01/92 1996-01-31 15.52.57

/usr/lib/objrepo 03.02.0000.0000 COMPLETE 01/01/92 1996-01-31 15.52.57


Figure 26. Example of UNIX Configuration of SW for Object, Overview

88 Release 5

Reports in the error subcomponentThis section describes the following error subcomponent reports:

¶ “UNIX Error by ID, Daily Overview” on page 90.

¶ “UNIX Error by Type, Daily Overview” on page 91.

¶ “UNIX Error by Class, Daily Overview” on page 92.

¶ “UNIX Error by Resource, Daily Overview” on page 93.

¶ “UNIX Error by Resource, Monthly Trend” on page 94.

The data flow for the error subcomponent (including the names of log files, PerformanceReporter records and tables) is given in Figure 12 on page 45.


5.R

epo

rts

UNIX Error by ID, Daily OverviewThis report only covers nodes that use AIX. For a specific node in the network, this report(see Figure 27) gives daily overview information about the error messages that have beenissued. For each error message, the report shows error ID, description, number of times themessage has occurred, and the percentage occurrence of the message.

You can use this report to identify potential problems within an application, since the errorID allows you to differentiate between errors caused by software applications, system errors,and so on.


Report ID XERROR01

Report group Error Subcomponent Reports

Source XERROR_D, (described on page 58)

Attributes AIX, Error, ID, Daily, Overview

Variables Date, Node name


Error ID The identification of the error message.

Description Description of the error message.

Errors (count) The number of times this error has occurred.

Errors (%) The percentage of the total errors, in which this message hasoccurred.

UNIX Error by ID, Daily OverviewDate: 1999-11-29


Errors ErrorsError ID| Description (count) (%)--------| ---------------------------------------- ---------- --------0F27AAE5| SOFTWARE PROGRAM ABNORMALLY TERMINATED 6 54.55E18E984F| SOFTWARE PROGRAM ERROR 3 27.271104AA28| System reset interrupt received 1 9.099DBCFDEE| Error logging turned on 1 9.09

| ========== ========| Total 11 100.00

Performance Reporter Report: XERROR01

Figure 27. Example of UNIX Error by ID, Daily Overview

90 Release 5

UNIX Error by Type, Daily OverviewThis report only covers nodes that use AIX. For a specific node in the network, this report(see Figure 28) gives daily overview information about the types of error messages that havebeen issued. For each error type, the report shows the number of times the message type hasoccurred, and the percentage occurrence of the message type.

You can use this report to identify potential problems within a system, since the messageresource type allows you to measure the number of errors that are being generated for aparticular error type.


Report ID XERROR02



Attributes AIX, Error, Type, Daily, Overview



Error type The error type, which can be the following:P = Program errorT = Token ring error

Errors (count) The number of times this error type has occurred.

Errors (%) The percentage of the total errors, that this error type hasoccurred.

UNIX Error by Type, Daily OverviewDate: 1999-11-29


| Errors ErrorsError type| (count) (%)----------| ---------- --------P | 9 81.82T | 2 18.18

| ========== ========Total| 11 100.00


Figure 28. Example of UNIX Error by Type, Daily Overview


5.R

epo

rts

UNIX Error by Class, Daily OverviewThis report only covers nodes that use AIX. For a specific node in the network, this report(see Figure 29) gives daily overview information about the class of error messages that havebeen issued. For each error class, the report shows the number of times the message hasoccurred, and the percentage occurrence of the message class.

You can use this report to identify potential problems within a system, since the messageresource type allows you to measure the number of errors that are being generated for aparticular error class.


Report ID XERROR03



Attributes AIX, Error, Class, Daily, Overview



Error class The error class, which can be the following:S = Software messageH = Hardware messageO = Error log command message

Errors (count) The number of times this error class has occurred.

Errors (%) The percentage of the total errors, that this error class hasoccurred.

UNIX Error by Class, Daily OverviewDate: 1999-11-29


Errors ErrorsError class (count) (%)----------- ---------- --------S 10 90.91O 1 9.09

========== ========Total 11 100.00


Figure 29. Example of UNIX Error by Class, Daily Overview

92 Release 5

UNIX Error by Resource, Daily OverviewThis report only covers nodes that use AIX. For a specific node in the network, this report(see Figure 30) gives daily overview information about the resource type of error messagesthat have been issued. For each resource type, the report shows the number of timesmessages of this type have occurred, and the percentage occurrence of messages of thisresource type.

You can use this report to identify potential problems within a system, since the messageresource type allows you to measure the number of errors that are being generated for aparticular resource.


Report ID XERROR04



Attributes AIX, Error, Resource, Daily, Overview



Resource nameThe name of the resource.

Errors (count)The number of times errors originating from this resource name, haveoccurred.

Errors (%) The percentage of the total errors, that have originated from this resource.

UNIX Error by Resource, Daily OverviewDate: 1999-11-29


Resource Errors Errorsname (count) (%)

-------------- ---------- ----------SYSPROC 7 63.64SRC 3 27.27errdemon 1 9.09

========== ==========Total 11 100.00


Figure 30. Example of UNIX Error by Resource, Daily Overview


5.R

epo

rts

UNIX Error by Resource, Monthly TrendThis report only covers nodes that use AIX. For a specific node in the network, this report(see Figure 31) gives monthly trend information about the resource name of error messagesthat have been issued. For each resource name, the report shows the number of timesmessages of the resource name have occurred.

You can use this report to identify potential problems within a system, since the messageresource type allows you to measure the number of errors that are being generated for aparticular resource.


Report ID XERROR05


Source XERROR_M, (described on page 58)

Attributes AIX, Error, Resource, Monthly, Trend

Variables From month, To month, Node name


Month The start date of the month.

Resource nameThe name of the resource.

Errors The number of times errors originating from this resource name, haveoccurred.

UNIX Error by Resource, Monthly TrendMonth: '1999-11-01' to '1999-12-01'


<-------- Month -------->1995-11-01 1995-12-01

Resource name Errors Errors-------------- ----------- -----------errdemon 1 1tok0 1 1SRC 3 4SYSPROC 19 18SYSVMM 3 5


Figure 31. Example of UNIX Error by Resource, Monthly Trend

94 Release 5

Reports in the performance subcomponentThis section describes the following performance subcomponent reports:

¶ “UNIX Perf CPU Utilization by System, Hourly Trend” on page 96.

¶ “UNIX Perf CPU Utilization by System, Daily Overview” on page 97.

¶ “UNIX Perf Statistics by System, Hourly Trend” on page 98.

¶ “UNIX Perf Statistics all Systems, Daily Overview” on page 100.

¶ “UNIX Perf Vol Group and File Syst, Daily Overview” on page 102.

¶ “UNIX Perf Disk I/O for a Disk, Hourly Trend” on page 103.

¶ “UNIX Perf Disk I/O for System, Daily Overview” on page 105.

¶ “UNIX Perf Page Space Utilization, Hourly Trend” on page 106.

The data flow for the performance subcomponent (including the names of log files,Performance Reporter records and tables) is given in Figure 13 on page 47.


5.R

epo

rts

UNIX Perf CPU Utilization by System, Hourly TrendFor a specific node in the network, this graphical representation (see Figure 32) shows thehourly trend of processor utilization, over a specified time period. Such information is usefulas an entry point when investigating system performance.

This information identifies the display:

Report ID XPERF01

Report group Performance Subcomponent Reports

Source XPERF_CPU_H, (described on page 59)

Attributes UNIX, Performance, CPU, Utilization, Usage, Hourly, Trend



Hour The hour of the day.

CPU avg The average processor time that has been utilized, in percent.

CPU max The maximum processor time that has been utilized, in percent.

UNIX Perf CPU Utilization by System, Hourly TrendDate: 1995-06-08

Node name: 'AIX99'

00

%

02 04 06 08 10Hour

12 14 16 18 20 22

CPU max

CPU avg

0

70

60

40

30

20

10

50

80

90

100

Figure 32. Example of UNIX Perf CPU Utilization by System, Hourly Trend

96 Release 5

UNIX Perf CPU Utilization by System, Daily OverviewFor all nodes in the network, this graphical representation (see Figure 33) shows the averagedaily processor utilization. The display is produced by period name (for example, PRIME orNIGHT). Such information is useful as an entry point when investigating systemperformance. An hourly graphical display of processor utilization for a specific node, isgiven in “UNIX Perf CPU Utilization by System, Hourly Trend” on page 96.


Report ID XPERF02


Source XPERF_CPU_D, (described on page 59)

Attributes UNIX, Performance, Cpu, Utilization, Usage, Daily, Overview

Variables Date, Period name



CPU avg The average processor time that has been utilized, in percent.

UNIX Perf CPU Utilization by System, Daily OverviewDate: 2000-06-08 Period: 'PRIME'

AIX11 AIX22 AIX33 HP111Node name

SUN11 SUN22HP2220

%

80

60

40

20

100

CPU avg

Figure 33. Example of UNIX Perf CPU Utilization by System, Daily Overview


5.R

epo

rts

UNIX Perf Statistics by System, Hourly TrendFor a specific node in the network, this report (see Figure 34) provides hourly, trendinformation about:

¶ Processor utilization (average and maximum)

¶ The number of available pages of memory (average and minimum)

¶ The paging rate (average and maximum)

¶ The number of processes (average and maximum)

¶ The number of users (average and maximum)

A total row is given, which are the averages and maximums calculated for all hours.


Report ID XPERF03


Source XPERF_CPU_H, (described on page 59)

Attributes UNIX, Performance, System, Usage, Hourly, Trend




CPU usage avg (%) The average processor utilization, in percent.

CPU usage max (%) The maximum processor utilization, in percent.

Mem free pages avg The average number of free pages of memory.

Mem free pages min The minimum number of free pages of memory.

Paging rate avg The average paging rate, per second.

Paging rate max The maximum paging rate, per second.

Processes avg The average number of processes.

UNIX Perf Statistics by System, Hourly TrendDate: 1999-12-27

Node name: 'hp111 '

| CPU CPU Mem free Mem free Paging Paging| usage usage pages pages rate rate Processes Processes Users Users

Hour| avg (%) max (%) avg min avg max avg max avg max-----| -------- -------- -------- -------- -------- -------- --------- --------- -------- --------

.| . . . . . . . . . .14| 6.7 43 1099 166 0.03 0.17 90.3 92 7.0 715| 4.8 50 1257 687 0.05 0.12 91.3 96 7.3 1216| 5.6 21 822 125 0.02 0.09 92.3 99 7.0 7.| . . . . . . . . . ..| . . . . . . . . . .| ======== ======== ======== ======== ======== ======== ========= ========= ======== ========

Total| 4.7 50 2447 125 0.06 0.17 85.8 99 6.2 12

Performance Reporter Report: XPERF03

Figure 34. Example of UNIX Perf Statistics by System, Hourly Trend

98 Release 5

Processes max The maximum number of processes.

Users avg Tghe average number of users for this node.

Users max The maximum number of users for this node.


5.R

epo

rts

UNIX Perf Statistics all Systems, Daily OverviewFor all nodes in the network, this report (see Figure 35) provides daily overview informationabout:

¶ Processor utilization (average and maximum)

¶ The number of available pages of memory (average and minimum)

¶ The paging rate (average and maximum)

¶ The number of processes (average and maximum)

¶ The number of users (average and maximum)

The report is produced by period name (for example, PRIME or NIGHT). An hourlybreakdown of the information provided in this report, and for a specific node, is given in“UNIX Perf Statistics by System, Hourly Trend” on page 98.


Report ID XPERF04


Source XPERF_CPU_D, (described on page 59)

Attributes UNIX, Performance, System, Usage, Daily, Overview

Variables Date, Period name



CPU usage avg (%) The average processor utilization, in percent.

CPU usage max (%) The maximum processor utilization, in percent.

Mem free pages avg The average number of free pages of memory.

Mem free pages min The minimum number of free pages of memory.

Paging rate avg The average paging rate, per second.

Paging rate max The maximum paging rate, per second.

Processes avg The average number of processes.

Processes max The maximum number of processes.

Users avg The average number of users for this node.

UNIX Perf Statistics all Systems, Daily OverviewDate: 1999-12-27 Period: 'PRIME '

| CPU CPU Mem free Mem free Paging Paging| usage usage pages pages rate rate Processes Processes Users Users

Node name | avg (%) max /%) avg min avg max avg max avg max---------------| -------- -------- -------- -------- -------- -------- --------- --------- -------- --------hp111 | 5.6 50 1068 125 0.01 2 91.4 96 7.1 8sun333 | 18.2 74 621 62 0.15 18 163.9 202 13.5 17


Figure 35. Example of UNIX Perf Statistics all Systems, Daily Overview

100 Release 5

Users max The maximum number of users for this node.


5.R

epo

rts

UNIX Perf Vol Group and File Syst, Daily OverviewThis report only covers nodes that use AIX. For a specific node in the network, this report(see Figure 36) provides daily overview information about space utilization: averageallocated and free space (in megabytes), and the average percentage used space (by spacename and device name). The report is produced by period name (for example, PRIME orNIGHT).


Report ID XPERF05


Source XPERF_DISK_D, (described on page 60)

Attributes AIX, Performance, Volume, File, Utilization, Usage, Daily, Overview



Volume The volume used by the node.

File system The file system corresponding to the given volume.

Space type The space type, which can be one of the following:FS = File systemVG = Volume group

Space size avg (MB) The average size of the space, in megabytes. This iscalculated as SPACE_SIZE_MB/RECORDS_COLLECTED.

Space free avg (MB) The average size of free space, in megabytes. This iscalculated as SPACE_FREE_MB/RECORDS_COLLECTED.

Space used avg (%) The average size of used space, in percent. This is calculatedas SPACE_USED_PCT/RECORDS_COLLECTED.

UNIX Perf Vol Group and File syst, Daily OverviewDate: 1999-12-27 Period: 'PRIME '

Node name: 'aix555'

File Space | Space size Space free Space usedVolume system type | avg (MB) avg (MB) avg (%)---------------- --------------- -----| ---------- ---------- ----------/ /dev/hd4 FS | 8.0 0.0 92.0/tmp /dev/hd3 FS | 12.0 11.0 6.0/usr /dev/hd2 FS | 556.0 67.0 87.0/usr/lpp/ileaf6. /dev/lv01 FS | 196.0 38.0 80.0/var /dev/hd9var FS | 16.0 5.3 65.0rootvg VG | 2212.0 1160.0 47.6


Figure 36. Example of UNIX Perf Vol Group and File Syst, Daily Overview

102 Release 5

UNIX Perf Disk I/O for a Disk, Hourly TrendFor a specific node in the network and a specific disk, this report (see Figure 37) provideshourly trend information about the disk’s utilization: the average and maximum percentagebusy times, the average and maximum read amounts (in kilobytes), and the average andmaximum write amounts (in kilobytes).


Report ID XPERF06


Source XPERF_DISKIO_H, (described on page 61)

Attributes UNIX, Performance, Disk, I/O, System, Hourly, Trend

Variables Date, Disk, Node name



Busy avg (%) The average time in which the processor was busy, inpercent.

Busy max (%) The maximum time in which the processor was busy, inpercent.

Read avg (KB/sec) The average amount of disk space read per second, inkilobytes. This is calculated asREAD_KB_TOTAL/MEASURED_SEC.

Read max (KB/sec) The maximum amount of disk space read per second, inkilobytes.

Write avg (KB/sec) The average amount of disk space written per second, in

UNIX Perf Disk I/O for a Disk, Hourly TrendDate: 1999-12-09

Node name: 'aix555 ' Disk: 'hdisk8'

Busy avg Busy max Read avg Read max Write avg Write max RW avg RW maxHour (%) (%) (KB/sec) (KB/sec) (KB/sec) (KB/sec) (KB/sec) (KB/sec)

------ ---------- ---------- ---------- ---------- ---------- ---------- --------- ----------12 12.9 110.6 1.7 52.0 5.4 18.0 7.1 70.013 8.9 40.9 0.4 14.0 4.3 8.0 4.7 22.014 8.6 24.2 0.1 6.0 4.4 8.0 4.5 14.015 53.1 361.1 20.0 219.0 11.3 67.0 31.3 286.016 16.3 87.2 3.7 35.0 5.1 13.0 8.8 48.017 17.3 115.3 3.2 55.0 6.2 14.0 9.4 69.018 8.1 19.9 0.2 5.0 4.1 6.0 4.3 11.019 7.6 10.0 1.0 10.2 3.6 5.0 4.6 15.220 7.7 9.7 0.5 4.1 4.9 5.0 5.4 9.121 7.7 10.6 1.5 17.3 4.0 4.0 5.5 21.322 7.7 9.8 0.1 1.9 2.2 5.0 2.3 6.923 7.6 9.8 0.3 2.2 4.5 5.0 4.8 7.2

========== ========== ========== ========== ========== ========== ========= ==========Avg/max 13.6 361.1 2.7 219.0 5.0 67.0 7.7 286.0


Figure 37. Example of UNIX Perf Disk I/O for a Disk, Hourly Trend


5.R

epo

rts

kilobytes. This is calculated asWRITE_KB_TOTAL/MEASURED_SEC.

Write max (KB/sec) The maximum amount of disk space written per second, inkilobytes.

RW avg (KB/sec) The average amount of disk space read and written persecond, in kilobytes. This is calculated asRW_KB_TOTAL/MEASURED_SEC.

RW max (KB/sec) The maximum amount of disk space read and written persecond, in kilobytes.

104 Release 5

UNIX Perf Disk I/O for System, Daily OverviewFor a specific node in the network, this report (see Figure 38) provides daily overviewinformation about the utilization of disks at the node: the average and maximum percentagebusy times, the average and maximum read amounts (in kilobytes), and the average andmaximum write amounts (in kilobytes). The report is produced by period name (for example,PRIME or NIGHT). If you require more detailed hourly trend information about a disk’sutilization, you can proceed to the report “UNIX Perf Disk I/O for a Disk, Hourly Trend” onpage 103.


Report ID XPERF07


Source XPERF_DISKIO_D, (described on page 61)

Attributes UNIX, Performance, Disk, I/O, System, Daily, Overview



Disk name The name of the physical disk.

Busy avg (%) The average time in which the disk was busy, in percent.

Busy max (%) The maximum time in which the disk was busy, in percent.

Read avg (KB/sec) The average amount of disk that was read per second, inkilobytes. This is calculated asREAD_KB_TOTAL/MEASURED_SEC.

Read max (KB/sec) The maximum amount of disk that was read per second, inkilobytes.

Write avg (KB/sec) The average amount of disk that was written per second, inkilobytes. This is calculated asWRITE_KB_TOTAL/MEASURED_SEC.

Write max (KB/sec) The maximum amount of disk that was written per second,in kilobytes.

UNIX Perf Disk I/O for System, Daily OverviewDate: 1999-12-27 Period: 'PRIME '

Node name: 'hp111'

Busy avg Busy max Read avg Read max Write avg Write maxDisk name (%) (%) (KB/sec) (KB/sec) (KB/sec) (KB/sec)---------- ---------- ---------- ---------- ---------- ---------- ----------hdisk0 33.5 17.0 1.0 1.4 11.1 40.0hdisk1 11.4 14.5 1.1 1.4 10.4 12.2hdisk2 10.0 10.0 1.0 1.0 10.0 1.2hdisk3 14.3 19.1 1.5 13.0 11.1 18.8

========== ========== ========== ========== ========== ==========Avg/max 17.3 19.1 1.1 13.0 10.6 40.0


Figure 38. Example of UNIX Perf Disk I/O for System, Daily Overview


5.R

epo

rts

UNIX Perf Page Space Utilization, Hourly TrendFor a specific node in the network, this report (see Figure 39) provides daily overviewinformation about the utilization of page space at the node. For each page space, informationabout the disk name for the page space, page space size, and the amount of page space used(in kilobytes and also as a percentage), is given.


Report ID XPERF08


Source XPERF_PAGING_H, (described on page 62)

Attributes UNIX, Performance, Page, Space, Utilization, Hourly, Trend




Page space name The name of the page space used by the node.

Disk name The physical disk name.

Page space size (MB) The size of the page space, in megabytes.

Page space used (MB) The amount of page space used, in megabytes.

UNIX Perf Page Space Utilization, Hourly TrendDate: 1999-01-14

Node name: 'hp111'

Page space Page space Page space Page spaceHour name Disk name size (MB) used (MB) used (%)----- ----------- ---------- ---------- ---------- ----------

. . . . . .8 hd6 hdisk0 96 63 66.1

9 hd6 hdisk0 96 65 67.7

10 hd6 hdisk0 96 66 68.9

11 hd6 hdisk0 96 67 69.8

12 hd6 hdisk0 96 67 69.8

13 hd6 hdisk0 96 67 69.8

14 hd6 hdisk0 96 68 71.2

15 hd6 hdisk0 96 68 70.8

16 hd6 hdisk0 96 67 70.0

17 hd6 hdisk0 96 68 70.8

18 hd6 hdisk0 96 67 70.3

19 hd6 hdisk0 96 68 70.8

20 hd6 hdisk0 96 68 70.8


Figure 39. Example of UNIX Perf Page Space Utilization, Hourly Trend

106 Release 5

Page space used (%) The amount of page space used, in percent. This iscalculated as 100*PS_USED_AVG_KB/PS_SIZE_AVG_KB.


5.R

epo

rts

108 Release 5

Log Record Definitions

This chapter describes (in alphabetical sequence) the following record definitions:

¶ “ACCT_COMMAND” on page 110

¶ “ACCT_CONNECT” on page 111

¶ “ACCT_DISK” on page 112

¶ “ACCT_PRINT” on page 113

¶ “CONFIG_HW” on page 114

¶ “CONFIG_SW” on page 115

¶ “ERROR” on page 117

¶ “PERF_CPU” on page 118

¶ “PERF_DISK_FS” on page 119

¶ “PERF_DISK_VG” on page 120

¶ “PERF_DISK_IO” on page 121

¶ “PERF_PAGING” on page 122

6


6.L

og

Reco

rdD

efinitio

ns

ACCT_COMMANDThis process accounting record definition is used for selecting records with record typePACCT.

Main section

Field name Offset Length Format Description

RECORD_LENGTH 0 2 BINARY Record length

NODE_NAME 4 16 CHAR Node name

RECORD_TYPE 21 5 CHAR Record type

DTE 27 6 CHAR Date

TME 33 6 TIME(HHMMSS) Time

AC_FLAG 40 5 EXTERNAL INTEGER Acct. flag

AC_STAT 46 5 EXTERNAL INTEGER Exit status

AC_UID 52 8 CHAR User name

AC_GID 61 8 CHAR Group name

AC_TTY 70 10 CHAR Control typewriter

AC_COMM 81 8 CHAR Command name

AC_UTIME 90 10 FLOAT User time

AC_STIME 101 10 FLOAT System time

AC_ETIME 112 10 FLOAT Elapsed time

AC_MEM 123 10 FLOAT Memory usage

AC_IO 134 10 FLOAT Characters Transferred

AC_RW 145 10 FLOAT Blocks Read/Written

110 Release 5

ACCT_CONNECTThis connect time accounting record definition is used for selecting records with record typeWTMP.

Main section





DTE 27 6 CHAR Date


UT_USER 40 8 CHAR User login name

UT_ID 49 14 CHAR /etc/inittab ID

UT_LINE 64 12 CHAR Device name

UT_TYPE 77 2 CHAR Type of entry

UT_PID 80 6 EXTERNAL INTEGER Connect ID

E_TERMINATION 87 2 CHAR Connect term status

E_EXIT 90 2 CHAR Connect exit status

UT_TIME 93 10 EXTERNAL FLOAT Time entry was made (seeNote)

UT_HOST 104 * CHAR Node name

Note::Time is given in seconds, measured from 01.01.1970


6.L

og

Reco

rdD

efinitio

ns

ACCT_DISKThis disk accounting record definition is used for selecting records with record type DTMP.

Main section





DTE 27 6 CHAR Date


USERID 40 11 EXTERNAL INTEGER User ID number

USER_NAME 52 8 CHAR User name

DISK_BLOCKS 61 8 EXTERNAL INTEGER Disk blocks

112 Release 5

ACCT_PRINTThis print accounting record definition is used for selecting records with record typeQACCT.

Main section





DTE 27 6 CHAR Date


PRINT_QUEUE 40 8 CHAR Printer queue remote name

PAGES 49 8 EXTERNAL INTEGER Printed pages

COPIES 58 8 EXTERNAL INTEGER The number of printouts

PR_USER 67 * CHAR(*) User and requesting node


6.L

og

Reco

rdD

efinitio

ns

CONFIG_HWThis hardware configuration record definition is used for selecting records with record typeCONF and record subtypes A, D, or S.

Main section





DTE 27 6 CHAR Date


RECORD_SUBTYPE 40 2 CHAR Record subtype

DEVICE_NAME 43 16 CHAR Device name

LOCATION 60 16 CHAR Location

DEVICE_CLASS 77 16 CHAR Device class

DEVICE_TYPE 94 16 CHAR Device type

DEVICE_SUBCLASS 111 16 CHAR Device subclass

DEVICE_DESC 128 * CHAR(*) Device descr.

114 Release 5

CONFIG_SWThis software configuration record definition is used for selecting records with record typeCONF and record subtypes H, I, or L.

Main section





DTE 27 6 CHAR Date



Record section SUBTYPE_HPresent if RECORD_SUBTYPE has the value H.Offset 42Length 36


RELEASE 1 15 CHAR Release number

STATUS 17 10 CHAR Installation status

INST_DATE 28 8 CHAR Installation date

Record section SUBTYPE_IPresent if RECORD_SUBTYPE has the value I.Offset 42Length 37


PRODUCT_ID 1 10 CHAR Product ID

FEATURE_ID 12 4 CHAR Feature ID

PRODUCT_NAME 17 20 CHAR Product name

Record section SUBTYPE_LPresent if RECORD_SUBTYPE has the value L.Offset 42Length 13


STATE 1 12 CHAR Product state

Record section FILE_NAMEOffset 79Length Length of FILE_NAME_L


FILE_NAME_L 1 4 EXTERNAL INTEGER Length of sectionFILE_NAME


6.L

og

Reco

rdD

efinitio

ns


FILE_NAME 6 * CHAR(*) File name

Record section PATH_NAMEOffset Dependent upon length of previous sectionLength Length of PATH_NAME_L


PATH_NAME_L 1 4 EXTERNAL INTEGER Length of sectionPATH_NAME

PATH_NAME 6 * CHAR(*) Path name

Record section PRODUCT_DESCPresent if RECORD_SUBTYPE has the value L.Offset Dependent upon length of previous sectionLength Length of PRODUCT_DESC_L


PRODUCT_DESC_L 1 4 EXTERNAL INTEGER Length of sectionPRODUCT_DESC

PRODUCT_DESC 6 * CHAR(*) Product_ description

116 Release 5

ERRORThis error record definition is used for selecting records with record type ERRPT.

Main section





DTE 27 6 CHAR Date


ERROR_ID 40 8 CHAR Error ID

ERROR_TYPE 49 1 CHAR Error type

ERROR_CLASS 51 1 CHAR Error class

RESOURCE_NAME 53 14 CHAR Resource name

DESCRIPTION 68 * CHAR(*) Error descr.


6.L

og

Reco

rdD

efinitio

ns

PERF_CPUThis CPU performance record definition is used for selecting records with record type PERFand record subtype VM.

Main section





DTE 27 6 CHAR Date



INTERVAL 43 4 EXTERNAL INTEGER Measured time

MEMORY_MIN 48 7 EXTERNAL INTEGER Free pages, min

MEMORY_AVG 56 9 EXTERNAL FLOAT Free pages, avg

MEMORY_MAX 66 7 EXTERNAL INTEGER Free pages, max

PAGING_MIN 74 7 EXTERNAL INTEGER Paging rate, min

PAGING_AVG 82 9 EXTERNAL FLOAT Paging rate, avg

PAGING_MAX 92 7 EXTERNAL INTEGER Paging rate, max

CPU_MIN 100 7 EXTERNAL INTEGER CPU usage, min

CPU_AVG 108 9 EXTERNAL FLOAT CPU usage, avg

CPU_MAX 118 7 EXTERNAL INTEGER CPU usage, max

USERS 126 5 EXTERNAL INTEGER Number of users

PROCESSES 132 5 EXTERNAL INTEGER Number of processes

118 Release 5

PERF_DISK_FSThis disk space performance record definition is used for selecting records with record typeDISK and record subtype FS.

Main section





DTE 27 6 CHAR Date



Record section FILE_SYSTEMOffset 42Length 24


TOTAL_SPACE_KB 1 9 EXTERNAL INTEGER Total space, in kilobytes

FREE_SPACE_KB 11 9 EXTERNAL INTEGER Free space, in kilobytes

USED_SPACE_PCT 21 3 EXTERNAL INTEGER Used space, in percent

Record section DEVICE_NAMEOffset 66Length Length of DEVICE_NAME_L


DEVICE_NAME_L 1 4 EXTERNAL INTEGER Length of sectionDEVICE_NAME

DEVICE_NAME 6 * CHAR(*) Device name

Record section SPACE_NAMEOffset 66Length Length of SPACE_NAME_L


SPACE_NAME_L 1 4 EXTERNAL INTEGER Length of sectionSPACE_NAME

SPACE_NAME 6 * CHAR(*) Space name


6.L

og

Reco

rdD

efinitio

ns

PERF_DISK_VGThis disk space performance record definition is used for selecting records with record typeDISK and record subtype VG.

Main section





DTE 27 6 CHAR Date



Record section VOLUME_GROUPOffset 42Length 30


TOTAL_SPACE_MB 1 9 EXTERNAL INTEGER Total space, in megabytes

FREE_SPACE_MB 11 9 EXTERNAL INTEGER Free space, in megabytes

USED_SPACE_MB 21 9 EXTERNAL INTEGER Used space, in megabytes

Record section SPACE_NAMEOffset Dependent upon length of previous sectionLength Length of SPACE_NAME_L


SPACE_NAME_L 1 4 EXTERNAL INTEGER Length of sectionSPACE_NAME

SPACE_NAME 6 * CHAR(*) Space name

120 Release 5

PERF_DISK_IOThis disk I/O performance record definition is used for selecting records with record typePERF and record subtype IO.

Main section





DTE 27 6 CHAR Date



INTERVAL 43 4 EXTERNAL INTEGER Measurement interval

MINREAD 48 7 EXTERNAL INTEGER KB read—minimum

AVGREAD 56 9 EXTERNAL FLOAT KB read—average

MAXREAD 66 7 EXTERNAL INTEGER KB read—maximum

TOTREAD 74 8 EXTERNAL INTEGER KB read—total

MINWRITE 83 7 EXTERNAL INTEGER KB write—minimum

AVGWRITE 91 9 EXTERNAL FLOAT KB write—average

MAXWRITE 101 7 EXTERNAL INTEGER KB write—maximum

TOTWRITE 109 8 EXTERNAL INTEGER KB write—total

MINRW 118 7 EXTERNAL INTEGER KB read andwritten—minimum

AVGRW 126 9 EXTERNAL FLOAT KB read andwritten—average

MAXRW 136 7 EXTERNAL INTEGER KB read andwritten—maximum

TOTRW 144 8 EXTERNAL INTEGER KB read and written—total

BUSY 153 5 EXTERNAL FLOAT Busy %

DISK 159 * CHAR(*) Physical disk name


6.L

og

Reco

rdD

efinitio

ns

PERF_PAGINGThis paging space performance record definition is used for selecting records with recordtype PERF and record subtype PS.

Main section





DTE 27 6 CHAR Date



PAGE_SPACE 43 20 CHAR Page space name

DISK 64 20 CHAR Physical disk name

PSSIZE 85 5 EXTERNAL INTEGER Page space size in KB

PSUSE 91 5 EXTERNAL INTEGER Page space used in KB

122 Release 5

Migrating from SP6000 feature

The UNIX Performance component replaces the earlier RISC System/6000 SystemPerformance feature (SP6000 feature) which was delivered at Version 1 Release 1.1 leveland which only supported the AIX operating system.

The UNIX Performance component provides support for additional operating systems to AIX(HP-UX and Sun Solaris).

If you already have SP6000 feature installed, you must:

¶ Migrate the host part of SP6000 feature to the UNIX Performance component level.

¶ Replace the agent programs on the nodes where SP6000 feature is currently installed,with the agent programs provided with the UNIX Performance component.

These activities are now described.

Migrating the host part of SP6000 featureWhen you install UNIX Performance component, any existing data tables containing datathat was created when SP6000 feature was installed will not be invalidated. All tabledefinitions for SP6000 feature are compatible with those used with UNIX Performancecomponent. Two columns have been added to the XPERF_DISKIO_H, _D, and _M tables,and log record definitions and update definitions have been changed to enable other UNIXsystems to be supported.

To bring your Performance Reporter database to the UNIX Performance component supportlevel, you must:

1. Check that you have recreated the Performance Reporter system tables. If the systemtables have not been recreated, refer to the instructions for migrating from PerformanceReporter 1.1.1 to 1.2.0 provided in the Administration Guide.

2. Edit the job provided in DRL120.SDRLCNTL(DRLJWUNX), and then submit this job.This job removes the component statistic entries for AIX and replaces them with theequivalent UNIX Performance component entries.

3. Select 2. Components from the Administration Dialog.

4. Note: The names of the components are different from the previous RISC System/6000System Performance feature. The status field displays INSTALLED for your previouslyinstalled components which now have the name UNIX. This is a normal part of themigration process and you should select these components under their new names.

Select the components you wish to reinstall, and press F6 (Install).

5. Remove obsolete objects to complete the installation:

7


7.M

igratin

gfro

mS

P6000

feature

¶ The log definition AIX.

a. Select 3. Logs from the Administration Dialog.

b. Select AIX and then from the Log pull-down select 6. Delete.

c. From the Confirmation window, press Enter to confirm that the log should bedeleted.

¶ The update definition XPERF_DISK_D

a. Select 4. Tables from the Administration Dialog.

b. Select XPERF_DISK and then press F5 (Updates).

c. Select XPERF_DISK_D and then press F11 (Delete).

d. From the Confirmation window, press Enter to confirm that the update should bedeleted.

The components that were previously installed with SP6000 feature are now migrated toUNIX Performance component.

Replacing agent programs on the nodesTo replace the SP6000 feature agent programs with UNIX Performance component agentprograms, you must:

1. Remove all programs contained in the directory /opt/epdmaix, by entering thesecommands:cd /opt/epdmaixrm bin/* etc/* lib/* epdmaix.tarrmdir bin etc libcd /optrmdir epdmaix

2. Remove the links on /usr/bin, by entering these commands:cd /usr/binrm drlparmget zccmsgwrite drlpsperfd drlactperfd drldeactperfd drlsend

3. Remove log files on AIX operating sytems, by entering these commands:umount /var/epdmrmfs -r /var/epdm

Remove log files on HP and Sun operating sytems, by entering these commands:rm /var/epdmaix/*rmdir /var/epdmaix

124 Release 5

Part III — ARM GuideChapter 8. Introducing the ARM component . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

Chapter 9. Installing and Configuring the ARM Component. . . . . . . . . . . . 129Installing the ARM component on OS/390 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129Updating the lookup tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

Updating X_NODE_NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130Testing the installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130Putting the ARM component into production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130


126 Release 5

Introducing the ARM component

The ARM (Application Response Measurement) component enables you to display andmanage data that is retrieved from the ARM Agents for Tivoli Distributed Monitoring. Thedata provides the response time for all of the application programs that have ARM API callsin their code.

The agent gathers the data, writes records as often as you specify, and stores the data in alocal log file. You can close the log file whenever you want and then transfer it forcollection, via FTP, to the OS/390 host on which Tivoli Performance Reporter is installed.Performance Reporter for OS/390 will then process the log file.

For more information about the ARM agents, see the Tivoli Distributed Monitoring ARMAgents User’s Guide.

8


8.In

trod

ucin

gth

eA

RM

com

po

nen

t

128 Release 5

Installing and Configuring the ARMComponent

This chapter supplements the procedure in the Administration Guide for installing acomponent, with information specific to the ARM component.


¶ Install the ARM component on OS/390

¶ Update the Performance Reporter lookup tables

¶ Test the ARM component installation

¶ Put the ARM component into production

Installing the ARM component on OS/390For information on how to install the Tivoli ARM component on OS/390, refer to the section"Installing a component" in the chapter "Working with components" in the AdministrationGuide

Updating the lookup tablesThe ARM component includes the X_NODE_NAME lookup table that you can customize tospecify the groupings you want reflected in your reports.

If you specify online installation, Performance Reporter displays the Lookup Tables window.To edit a lookup table using ISPF edit, select the table and press Enter.

If you specify batch mode installation, you can edit the lookup table using the ISPF editor,after the component is installed. To enter an ISPF edit from Performance Reporter, performthe following steps:

1. Select ’Performance Reporter Administration’ from the Performance Reporter PrimaryMenu.

2. Select 4. Tables.

3. Select the lookup table that you want to edit, select the Edit pull-down, and press Enter.

4. Select 3. ISPF Editor from the Edit pull-down.

To customize the ARM component, you must update the X_NODE_NAME lookup tabledescribed in the following section.

9


9.In

stalling

the

AR

MC

om

po

nen

t

Updating X_NODE_NAMEThis lookup table contains the UNIX node names, accounting groups, internet addresses, andperiod plans. It is used during the collect, to translate the node name to its accounting groupand period plan.

If you are installing online, you will be given the opportunity to edit the contents using theISPF editor.

A sample of the lookup table contents is given in “Lookup Tables” on page 139. Add theinformation needed for all the UNIX nodes for which you are going to do reporting.

Testing the installationBefore you start to use the ARM component on a regular basis, run a few tests to checkthat:

1. The installation was successful.¶ Performance Reporter is collecting the correct data.¶ The data is being stored correctly.¶ The correct data is being used for the creation of reports.


Refer to the Administration Guide for details about the steps involved in testing componentinstallation.

Putting the ARM component into productionAfter you run the tests and verify that the installation was successful, you can put the ARMcomponent into production.


To run reports in batch, you must first use the Administration Dialog to set the batchparameters for each report.


Check that thefile transferprogram hastransferred data.

RunCOLLECTjob

Performdatabase

maintenance

Createreports


130 Release 5

Part IV — ARM ReferenceChapter 10. Data Flow and Performance Reporter Objects. . . . . . . . . . . . . 133ARM Agents general data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

Chapter 11. Data Tables, Lookup Tables, and Control tables. . . . . . . . . . . 137Naming standard for tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137Table descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

ARM_TABLE_H,_D,_M . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138Lookup Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139X_NODE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Example of table contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Chapter 12. Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141ARM report definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141ARM reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

UNIX Application Monitoring Hourly Response Times trend report . . . . . . . . . . . . . . . . . . . . . . . . . . 142UNIX Application Monitoring Hourly Service Level Summary report. . . . . . . . . . . . . . . . . . . . . . . . . 143

Chapter 13. Log Record Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145ARM log collector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145ARM log definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146ARM record definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147


132 Release 5

Data Flow and Performance ReporterObjects


¶ The general data flow, starting with the gathering of data into log files at the DistributedMonitoring nodes where the ARM agents are running, and ending with the production ofPerformance Reporter reports


¶ The data flow for the ARM, including the names of the ARM log files, and thePerformance Reporter records, tables, and reports

10


10.D

ataF

low

and

Rep

orter

Ob

jects

ARM Agents general data flow


�1� and �2�Process ARM data and send the data to OS/390.




1. Process the data and send it to OS/390, Steps �1�and �2�

To accomplish this, you must send the ARM log file to OS/390.


In processing each record contained in the log data set, the collect procedure uses a logcollector to update the Performance Reporter table with the record. To do this, the logcollector uses:

¶ The log definition and record definitions to update the Performance Reporter tablewith the reformatted record.

¶ An update definition to decide which reformatted record fields are to be included inwhich Performance Reporter table, including further summarizing into other tables

DM Nodesrunning the ARM agents

Log files

Log data sets


Tables

Reports

Control tablesLookup tables


P.R.“Collect”

Lookup tables

1

2

3

4

Figure 41. General ARM data flow

134 Release 5

(for example, updating the daily table ARM_TABLE_D from the information usedfor updating the hourly table ARM_TABLE_H).

¶ Lookup tables (which contain user-defined information that defines an organization’soperating environment) to add user-defined data to the Performance Reporter tablerecord.

A description of the collect procedure is provided in the Language Guide and Reference.

A description of the use of control tables is provided in the Administration Guide


A description of how to create new reports is provided in the Guide to the Reporting Dialog

The reports that can be created when using the ARM are described in “Reports” onpage 141.


10.D

ataF

low

and

Rep

orter

Ob

jects

136 Release 5

Data Tables, Lookup Tables, and Controltables



¶ The format that is used for defining ARM table names and views.

¶ The layout of the tables that are used in this chapter to describe each table and lookuptable.

Naming standard for tablesThe names of ARM tables use the following format:

ARM_TABLE_suffix

where suffix indicates the summarization level of the data in the table (for example,ARM_TABLE_D for command statistics summarized by day).

A table name can have these summarization-level suffixes:_H The table holds data summarized by hour (hourly data)._D The table holds data summarized by day (daily data)._M The table holds data summarized by month (monthly data).

Lookup tables do not have a suffix.


¶ Key columns are marked with a: k. They are sorted in the sequence they appear in thetable.

¶ Data columns follow the last key column.

The descriptions of most key columns and data columns contain references to the fieldsfrom which they are derived in the record. For an explanation of these fields, refer to theapplicable product documentation.

11


11.D

ataTab

les,L

oo

kup

Tables,

and

Co

ntro

ltab

les

ARM_TABLE_H,_D,_MThese tables provide hourly, daily, and monthly statistics on ARM data. The source file is inthe member DRLTXARM, contained in the dataset xxxxxxxx.SDRLDEFS, where xxxxxxxx isyour high level identifier.

The default retention period are:

¶ 7 days for ARM_TABLE_H

¶ 30 days for ARM_TABLE_D

¶ 765 days for ARM_TABLE_M

Primary keys: DATE, TIME, NODE_NAME, PERIOD_NAME, APPL_SYMBOL_NAME,APPL_USERID, and TRANS_SYMB_NAME.

Table 9. ARM table definitionsColumn name Data type Description

DATE k DATE Date when the record was written. For ARM_TABLE_M this is the dateof the first day of the month.

TIME k TIME Time when the record was written. The format is HH.MM.SS. FromTME. This field is present only in ARM_TABLE_H.

NODE_NAME k CHAR(16). Node name. From NODE_NAME.

PERIOD_NAME k CHAR(8). Name of the period. The name is derived by using the fieldsNODE_NAME, DTE, and TME from the record as parameters in thePERIOD function. If no match is found, this column is set to "?".

APPL_SYMB_NAME k CHAR(16). Application symbolic name.

APPL_USERID k CHAR(16). Application user ID.

TRANS_SYMB_NAME k CHAR(16). Transaction symbolic name.

NODE_NET_ADDR CHAR(16). Node network address.

NODE_SYMB_NAME CHAR(16). Node symbolic name.

RECORDS_COLLECTED INTEGER Number of records collected. This is the count of records.

MAX_RESP_TIME INTEGER Maximum response time.

MIN_RESP_TIME INTEGER Minimum response time.

AVG_RESP_TIME FLOAT Average response time.

NR_TRANS INTEGER Number of transactions.

NR_SUM_SQUARES INTEGER Number of sum of squares.

BUCKET_1_B INTEGER Bucket bound 1.






BUCKET_1 INTEGER Bucket 1.







NR_FAIL_TRANS INTEGER Number of failed transactions.

138 Release 5

Table 9. ARM table definitions (continued)Column name Data type Description

NR_ABORT_TRANS INTEGER Number of aborted transactions.

Lookup Tables

This section describes the X_NODE_NAME lookup table.

For descriptions of common lookup tables used by other Performance Reporter features,refer to the Administration Guide.








PERIOD_PLAN CHAR(8) Period plan. This value can be used as a parameter when using the PERIODfunction looking up PERIOD_NAME.

Figure 42. X_NODE_NAME lookup table


11.D

ataTab

les,L

oo

kup

Tables,

and

Co

ntro

ltab

les

140 Release 5

Reports

This chapter describes the reports provided with the ARM component.

ARM report definitionsA new report group has been defined. It is called UNIX Application Monitoring and it isnamed ARMREPG.

It contains the following reports:

1. UNIX Application Monitoring Hourly Response Times Trend (ARM01)

2. UNIX Application Monitoring Hourly Service Level Summary (ARM02)

The source file for the group definitions is in the member DRLOXARM, which is containedin the dataset xxxxxxxx.SDRLRENU, where xxxxxxxx is a high-level identifier. The sourcefiles for the related queries and forms for the report definitions are in the membersDRLQARyy and DRLFARyy respectively, where yy is 01 or 02. They are located in thedataset xxxxxxxx.SDRLRENU, where xxxxxxxx is a high-level identifier.

12


12.R

epo

rts

ARM reports

UNIX Application Monitoring Hourly Response Times trend reportThe following information identifies the report:

Report ID ARM01

Report group ARM reports

Source ARM_TABLE_H

Attributes UNIX, ARM, OVERVIEW, RESPONSE, PERFORMANCE

Variables DATE, NODE_NAME, APPL_NAME, TRANS_NAME

Figure 43 shows an example of this report:

The report contains the following columns:

TIME

NODE_NAME

TRANS_SYMB_NAME

APPL_SYMB_NAME

MAX_RESP_TIME

AVG_RESP_TIME

MIN_RESP_TIME

UNIX Application Monitoring Hourly Response Times trend

MAX AVG MINNODE TRANS APPL RESP RESP RESP

HOUR NAME NAME NAME TIME TIME TIME-------- -------------------- -------------------- ---------------- --------- --------- ------14-00-00 cspm1 ple_transaction_ sample program 100 6.980E+01 60

Performance Reporter: ARM01

Figure 43. UNIX Application Monitoring Hourly Response Times trend report

142 Release 5

UNIX Application Monitoring Hourly Service Level Summary reportThe following information identifies the report:

Report ID ARM02

Report group ARM reports

Source ARM_TABLE_H

Attributes UNIX, ARM, OVERVIEW, SERVICE, SUMMARY

Variables DATE, SNMP_SUBAGENT, APPL_NAME, TRANS_NAME

Figure 44 shows a partial view of this report:


TIME

NODE_NAME

TRANS_SYMB_NAME

APPL_SYMB_NAME

NR_TRANS

NR_FAIL_TRANS

NR_ABORT_TRANS

MAX_RESP_TIME

AVG_RESP_TIME

MIN_RESP_TIME

BUCKET_1

BUCKET_2

BUCKET_3

BUCKET_4

BUCKET_5

BUCKET_6

BUCKET_7

UNIX Application Monitoring Hourly Service Level Summary

NUM NUM MAX AVG MINNODE TRANS APPL NUM FAIL ABORT RESP RESP RESP

HOUR NAME NAME NAME TRANS TRANS TRANS TIME TIME TIME BUCKET1 BUCKET2-------- --------- ------------------ --------------- ----- ----- ----- ------ --------- ------ ------- -------14-00-00 cspm1 ple_transaction_ sample program 64 0 0 100 6.980E+01 60 0 0

0 0

Performance Reporter: ARM02

Figure 44. UNIX Application Monitoring Hourly Service Level Summary report, partial view


12.R

epo

rts

144 Release 5


ARM log collectorThe log collector component processes the log file. During the processing, the ARM data isstored in the database with the needed summarizations. When the processing is complete youcan use both graphical and tabular reports to display the data.

The definitions for the log collector are contained in the log definition, the record definition,and the table definition.

¶ The log definition describes the format of the log file to the log collector.

¶ The record definition describes the record format of the ARM application to the logcollector.

¶ The table definition consists of two parts:

v The table definition that tells the log collector how the table for ARM data is built.This table will contain the data.

v The update table definition that tells the log collector how to process the data thatcomes from the log file before the log collector stores the data in the data table.

The report definition describes the report layouts, data, and defaults.

13


13.L

og

Reco

rdD

efinitio

ns

ARM log definitionsThe ARM log file is a flat file that contains several records. Each record is written by theARM agent. The format of each record is defined in the member DRLLXARM, which iscontained in the dataset xxxxxxxx.SDRLDEFS, where xxxxxxxx is a high-level identifier. Itsformat is as follows:

Table 10. ARM log file formatField Offset Length Type Description


RECORD_VERSION 6 2 CHAR Record version

INTERVAL 8 8 EXTERNAL HEX Interval

DTE 66 6 DATE(YYMMDD) Date


146 Release 5

ARM record definitionsThe ARM record definition describes to the log collector the data that is written to the logfile. The record format for the RT record is defined in the member DRLRXARM, which iscontained in the dataset xxxxxxxx.SDRLDEFS, where xxxxxxxx is a high-level identifier.

It is identified by:

RECORD_TYPE = RT

and its fields are:

Table 11. ARM record definition fieldsField Offset Length Type Description


RECORD_VERSION 6 2 CHAR Record version

INTERVAL 8 8 EXTERNAL HEX Interval

BUCKET_1_B 16 8 EXTERNAL HEX Bucket bound 1






CENTURY 64 2 EXTERNAL INTEGER Century

DTE 66 6 CHAR Date


SECOND_100 78 2 EXTERNAL INTEGER Seconds/100

MAX_RESP_TIME 80 8 EXTERNAL HEX Max response time

MIN_RESP_TIME 88 8 EXTERNAL HEX Min response time

AVG_RESP_TIME 96 8 EXTERNAL HEX Average response time

NR_TRANS 104 8 EXTERNAL HEX Number of transactions

NR_SUM_SQUARES 112 8 EXTERNAL HEX Number of squares

BUCKET_1 120 8 EXTERNAL HEX Bucket 1







NR_FAIL_TRANS 176 8 EXTERNAL HEX Number of failed transactions

NR_ABORT_TRANS 184 8 EXTERNAL HEX Number of aborted transactions

NODE_NAME_L 192 4 EXTERNAL INTEGER Length of next field

NODE_NAME * CHAR Node name

NODE_NET_ADDR_L * 4 EXTERNAL INTEGER Length of next field

NODE_NET_ADDR * CHAR Node network address

NODE_SYMB_NAME_L * 4 EXTERNAL INTEGER Length of next field

NODE_SYMB_NAME * CHAR Symbolic node name

APPL_SYMB_NAME_L * 4 EXTERNAL INTEGER Length of next field

APPL_SYMB_NAME * CHAR Symbolic application name

APPL_USERID_L * 4 EXTERNAL INTEGER Length of next field

APPL_USERID * CHAR Application user ID

TRANS_SYMB_NAME_L * 4 EXTERNAL INTEGER Length of next field

TRANS_SYMB_NAME * CHAR Symbolic transaction name


13.L

og

Reco

rdD

efinitio

ns

* The offset of those fields is variable and is obtained by adding up the current offset withthe length of the previous fields.

148 Release 5

Part V — Distributed Monitoring GuideChapter 14. Introducing Distributed Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . 151

Chapter 15. Configuring Distributed Monitoring . . . . . . . . . . . . . . . . . . . . . . . . 153Processing Distributed Monitoring data and sending it to OS/390 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

Create the task from a UNIX system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155Create the task from an OS/390 UNIX system service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156Create the task from a Windows NT system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157Execute the task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

Updating the lookup tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159Updating X_NODE_NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159Updating SENTRY_MONITOR_NT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159Updating SENTRY_MONITOR_UN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Testing the installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160Putting the Distributed Monitoring component into production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160


||

150 Release 5

Introducing Distributed Monitoring

With the Performance Reporter Distributed Monitoring Component (Sentry) you can display14 predefined reports on data that is retrieved from the Tivoli Distributed Monitoringapplication. The data is gathered by Tivoli Distributed Monitoring 3.01 (or later), stored in alog file, and sent via FTP to the Performance Reporter product, which in turn processes thedata. At the end of processing, Tivoli Distributed Monitoring data is available in thedatabase for reporting. The reports can display data that is retrieved from all supportedUNIX platforms and from Windows NT 3.51 and 4.0.

Note: Distributed Monitoring Component (Sentry) requires OS/390 Version 2.4 or later.

Distributed Monitoring reports the status of a variety of networked resources, such as hosts,applications, and processes. You can use Distributed Monitoring to gather information aboutlocal and remote resources, as well as system resources. In addition, Distributed Monitoringcan generate events and alarms on a netwok-wide basis.

The data produced by Distributed Monitoring can be stored on either agent or server. Theserver runs on the following platforms:¶ UNIX¶ Windows NT¶ OS/390 (since Tivoli Distributed Monitoring 3.6).

The following list shows the predefined reports provided with the Distributed MonitoringComponent (Sentry):

Table 12. Distributed Monitoring predefined reportsType of Monitor Description

All Distributed Monitoring NT Monitors A complete view of 67 NT monitors gathered by Tivoli Distributed Monitoring,detailed by hourly, daily, and weekly trends.

All Distributed Monitoring UNIX Monitors A complete view of 27 UNIX monitors gathered by Tivoli Distributed Monitoring,detailed by hourly, daily, and weekly trends.

Distributed Monitoring UNIX Disk Resources Monitors A detailed view of the UNIX disk resources monitors.

Distributed Monitoring UNIX System Monitors A detailed view of the UNIX system monitors.

Distributed Monitoring UNIX Network Monitors A detailed view of the network monitors.

Distributed Monitoring NT Processor Monitors A detailed view of the NT processor monitors.

Distributed Monitoring NT TCP Monitors A detailed view of the NT TCP monitors.

Distributed Monitoring NT Memory Monitors A detailed view of the NT memory monitors.

Distributed Monitoring NT Server Monitors A detailed view of the NT server monitors.

Distributed Monitoring NT System Monitors A detailed view of the NT system monitors.

14


|||||

14.In

trod

ucin

gD

istribu

tedM

on

itorin

g

152 Release 5

Configuring Distributed Monitoring

This chapter describes how to configure Tivoli Distributed Monitoring. The chapter alsoprovides information that supplements the procedure in the Administration Guide forinstalling a component that is specific to the Distributed Monitoring Component (Sentry).


¶ Process Distributed Monitoring data and transmit this data to Performance Reporter onOS/390 by creating and executing a task.

¶ Update the Performance Reporter lookup tables.

¶ Test the Distributed Monitoring Component (Sentry) installation.

¶ Put the Distributed Monitoring Component (Sentry) into production.

Note: For information on how to install the Tivoli Distributed Monitoring Component(Sentry) on OS/390 refer to the Administration Guide.

Processing Distributed Monitoring data and sending it to OS/390To ensure that Distributed Monitoring data can be displayed with Performance Reporter, thedata must be processed and sent to the OS/390 host that is running Performance Reporter.

The log file that Distributed Monitoring creates must conform to the following criteria to beprocessed by Performance Reporter:

¶ The message style of the log file must be Local format 1.

¶ The log file path must be as follows:v For UNIX and OS/390 USS: /tmp/filenamesentry.logv For Windows NT: d:\temp\filenamesentry.log

where:

d: is the same drive where the managed node is installed (NT).

filename is the name you assign to the log file.

¶ The name of the Distributed Monitoring profile that generates the log file must notcontain a vertical bar (|).

For more information about Distributed Monitoring Component (Sentry) profiles, see theTivoli Distributed Monitoring User’s Guide.

To process and send Distributed Monitoring data from a UNIX, OS/390 USS, or WindowsNT system to Performance Reporter on OS/390, you must create a task by running a script.

15


|||

|

||

|

|

|

||

|

||

|

||

||

||

15.C

on

figu

ring

Distrib

uted

Mo

nito

ring

The scripts are located on the OS/390 host in the following Performance Reporter members:

Table 13. Performance Reporter members by operating systemOperating system Performance Reporter member

UNIX DRLSEUNX

OS/390 USS DRLSE390

Windows NT DRLSENT

The members are under the partitioned dataset xxxxxxxx.SDRLWS, where xxxxxxxx is yourhigh-level identifier.

154 Release 5

|

|

Create the task from a UNIX system1. Download the Performance Reporter member DRLSEUNX from OS/390 to a directory

on the UNIX machine on which Tivoli Distributed Monitoring is installed. Use anydirectory, for example, /tmp.

Perform the following steps to download the member as a file with the name prmvs.tar:

a. Enter ftp hostname, where hostname is the OS/390 hostname.

b. Enter the user ID and password when prompted.

c. Enter bin

d. Enter lcd /tmp

This command changes the current directory on the UNIX system.

e. Enter get 'xxxxxxxx.SDRLWS(DRLSEUNX)' prmvs.tar

This command gets the script from the partitioned dataset on OS/390, gives it thename prmvs.tar, and puts it in the /tmp directory on the UNIX machine.

f. Enter quit.

2. Enter tar -xvf prmvs.tar

The following files are created:¶ prrepinst¶ prmvsinstall¶ prrepmvs.tll¶ SentryFormats.cat

3. Enter /etc/Tivoli/setup_env.sh

This command configures the Tivoli Distributed Monitoring environment variables.

4. Enter prrepinst

This script copies the SentryFormats.cat file to the Tivoli message catalogs directory andcreates the following:¶ A policy region named Reporter Region¶ A task library named Reporter Library¶ A task named SentryLogToReporter.


|

15.C

on

figu

ring

Distrib

uted

Mo

nito

ring

Create the task from an OS/390 UNIX system service1. Download the Performance Reporter member DRLSE390 from OS/390 to a directory on

the OS/390 USS machine on which Tivoli Distributed Monitoring is installed. Use anydirectory, for example, /tmp.

Perform the following steps to download the member as a file with the name prmvs.tar:



c. Enter bin

d. Enter lcd /tmp

This command changes the current directory on the OS/390 USS system.

e. Enter get 'xxxxxxxx.SDRLWS(DRLSE390)' prmvs.tar

This command gets the script from the partitioned dataset on OS/390, gives it thename prmvs.tar, and puts it in the /tmp directory on the OS/390 USS machine.

f. Enter quit.

2. Enter tar -xvf prmvs.tar

The following files are created:¶ prrepinst¶ prmvsinstall¶ prrepmvs.tll¶ SentryFormats.cat

3. Enter /etc/Tivoli/setup_env.sh


4. Enter prrepinst

This script copies the SentryFormats.cat file to the Tivoli message catalogs directory andcreates the following in the /tmp/prmvs directory:¶ A policy region named Reporter Region¶ A task library named Reporter Library¶ A task named SentryLogToReporter.

156 Release 5

|

|||

|

|

|

|

|

|

|

||

|

|

|||||

|

|

|

|||||

|

Create the task from a Windows NT system1. Download the Performance Reporter member DRLSENT from OS/390 to a directory on

the Windows NT machine on which Tivoli Distributed Monitoring is installed. Use anydirectory, for example, \temp.

Perform the following steps to download the member as a file with the nameprwinmvs.tar:



c. Enter bin

d. Enter lcd \temp

This command changes the current directory on the Windows NT system.

e. Enter get 'xxxxxxxx.SDRLWS(DRLSENT)' prwinmvs.tar

This command gets the script from the partitioned dataset on OS/390, gives it thename prwinmvs.tar, and puts it in the \temp directory on the Windows NT machine.

f. Enter quit.

2. Enter tar -xvf prwinmvs.tar

The following files are created:¶ prwinrepinst¶ prmvsinstall¶ prrepmvs.tll¶ SentryFormats.cat

3. c:\winnt40\system32\drivers\etc\Tivoli\setup_env.cmd


4. Enter sh prwinrepinst

This script copies the SentryFormats.cat file to the Tivoli message catalogs directory andcreates the following:¶ A policy region named Reporter Region¶ A task library named Reporter Library¶ A task named SentryLogToReporter.


|

|

15.C

on

figu

ring

Distrib

uted

Mo

nito

ring

Execute the taskYou can execute the task manually or by scheduling a job. To execute a task by scheduling ajob, see the Tivoli Framework User’s Guide

To execute the task manually, perform the following steps:

1. From Tivoli Distributed Monitoring, double-click on the Reporter Region icon.

2. On the Policy Region: Reporter Region window, double-click on the Reporter Library:icon.

3. On the Task Library: Reporter Library window, click with the right mouse button on theSendSentryLogToReporter icon, and select Execute Task.

4. On the Execute Task window, from the Available Task Endpoints list, select theworkstations on which to execute the task, and click on the left arrow button.

The selected workstations are moved to the Selected Task Endpoints list.

5. Ensure that the specified Timeout value shown in the Execution Parameters list is atleast as large as the time required to process the Distributed Monitoring log file.

Note: The user who executes the task must have the same authorization as the user whocreated the log. If the Distributed Monitoring log file was created with rootauthority (in UNIX or OS/390 USS) or administrator authority (in Windows NT),you will receive the following warning message after the file is transferred toOS/390: Unable to remove file. Not owner.

To delete the file when you execute the task (and avoid the above-mentionedproblem), you must modify the execution privileges of theSendSentryLogToReporter task in the following way:

a. Click with the right button on the Task icon, and select Edit task.

b. In the Execution Privileges panel of the Edit Task window, enter root (inUNIX or OS/390 USS) or administrator (in Windows NT), in the UserName field.

c. To save, select the Change and Close button.

The log file filenamesentry.log will be deleted when you execute the task inthe following step. (As an alternative, you can delete the file manually afteryou execute the task.)

6. Click on the Execute push button.

7. On the SendSentryLogToReporter window, specify the address, login, and password ofthe OS/390 host on which Performance Reporter resides, in the following fields:¶ Reporter Host Address¶ Reporter Host Login¶ Reporter Host Password

8. In the Destination DS High Level Qualifier field, specify the dataset name (sequential)in which to store the log file at the destination.

9. In the Destination Volume field, specify the volume in which to store the log file at thedestination.

158 Release 5

|

|||||

|||

|

|||

|

|||

|

Each log file is sent to the destination after being renamed node.sentry.dyymmdd, wherenode is the name of the node at which the data was gathered and yymmdd is the date of thelast day on which the data was gathered.

Updating the lookup tablesThe Distributed Monitoring Component (Sentry) includes the following lookup tables thatyou can customize to specify the groupings you want reflected in your reports:¶ X_NODE_NAME¶ SENTRY_MONITOR_NT¶ SENTRY_MONITOR_UN

If you specify online installation, Performance Reporter displays the Lookup Tables window.To edit a lookup table using ISPF edit, select a table and press Enter.

If you specify batch mode installation, you can edit the lookup tables using the ISPF editor,after the component is installed. To enter an ISPF edit from Performance Reporter, performthe following steps:1. Select ’Performance Reporter Administration’ from the Performance Reporter Primary

Menu2. Select ’4. Tables’3. Select the lookup table that you want to edit, select the ’Edit’ pull-down, and press Enter4. Select ’3. ISPF Editor’ from the Edit pull-down.

With the Distributed Monitoring component, to customize you must update theX_NODE_NAME, SENTRY_MONITOR_NT, and SENTRY_MONITOR_UN lookup tablesdescribed in the following sections.

Updating X_NODE_NAMEThis lookup table contains the UNIX, OS/390 USS, and Windows NT node names,accounting groups, internet addresses, and period plans. It is used during the collect, totranslate the node name to its accounting group and period plan.


A sample of the lookup table contents is given in Figure 42 on page 139. Add theinformation needed for all the UNIX, OS/390 USS, and Windows NT nodes for which youare going to do reporting.

Updating SENTRY_MONITOR_NTThis lookup table contains the Windows NT information that is used during the collect.


A sample of the lookup table contents is given in Figure 48 on page 171. Add theinformation needed for all the Windows NT nodes for which you are going to do reporting.

Updating SENTRY_MONITOR_UNThis lookup table contains the UNIX and OS/390 USS information that is used during thecollect.


||

15.C

on

figu

ring

Distrib

uted

Mo

nito

ring


A sample of the lookup table contents is given in Figure 49 on page 171. Add theinformation needed for all the UNIX and OS/390 USS nodes for which you are going to doreporting.

Testing the installationBefore you start to use the Distributed Monitoring Component (Sentry) on a daily basis, runa few tests to check that:

1. The installation was successful:¶ Performance Reporter is collecting the correct data.¶ The data is being stored correctly.¶ The correct data is being used for the creation of reports.


Refer to the Administration Guide for details about the steps involved in testing componentinstallation.

Putting the Distributed Monitoring component into productionAfter you run the tests and verify that the installation was successful, you can put theDistributed Monitoring Component (Sentry) into production.


To run reports in batch, you must first use the Administration Dialog to set the batchparameters for each report.


Check that thefile transferprogram hastransferred data

RunCOLLECTjob

Performdatabasemaintenance

Createreports


160 Release 5

|||

Part VI Distributed Monitoring ReferenceChapter 16. Data Flow and Performance Reporter Objects. . . . . . . . . . . . . 163Tivoli Distributed Monitoring general data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

Chapter 17. Data Tables, Lookup Tables, and Control Tables . . . . . . . . . . 167Naming standard for tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167Table descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167SENTRY_N_NT_H,_D,_W . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168SENTRY_N_UN_H,_D,_W . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169Lookup Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

X_NODE_NAME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170Example of table contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

SENTRY_MONITOR_NT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171Example of table contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

SENTRY_MONITOR_UN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171Example of table contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Chapter 18. Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173Distributed Monitoring report definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

DM Service Level Details Hourly (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174DM Service Level Details Daily (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176DM Service Level Details Weekly (UNIX). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178DM System Resources Monitors Daily (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180DM Network Monitors Daily (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181DM Disk Resource Monitors Daily (UNIX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182DM Service Level Details Hourly (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183DM Service Level Details Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186DM Service Level Details Weekly (NT). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189DM TCP Monitors Daily (NT). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192DM System Monitors Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193DM Server Monitors Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194DM Processor Monitors Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195DM Memory Details Daily (NT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

Chapter 19. Log Record Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197Distributed Monitoring log collector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197Distributed Monitoring log definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197Distributed Monitoring record definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198


162 Release 5

Data Flow and Performance ReporterObjects


¶ The general data flow, starting with the gathering of data at the Distributed Monitoringnodes into log files, and ending with the production of Performance Reporter reports


¶ The data flow for Distributed Monitoring, including the names of the DistributedMonitoring log files, and the Performance Reporter records, tables, and reports.

16


16.D

ataF

low

and

Rep

orter

Ob

jects

Tivoli Distributed Monitoring general data flow


�1� and �2�Process Distributed Monitoring data and send the data to OS/390.




1. Process the data and send it to OS/390, Steps �1�and �2�

To accomplish this, you must create and execute a task. See “Processing DistributedMonitoring data and sending it to OS/390” on page 153 for a detailed description of howto do this.


In processing each record contained in the log data set, the collect procedure uses a logcollector to update the Performance Reporter table with the record. To do this, the logcollector uses:

¶ The log definition and record definitions to update the Performance Reporter tablewith the reformatted record.

¶ An update definition to decide which reformatted record fields are to be included inwhich Performance Reporter table, including further summarizing into other tables

DM Nodes

Log files

Log data sets


P.R."Collect"

Tables Control tablesLookup tables

Reports Lookup tables


1

2

3

4

Figure 46. General Distributed Monitoring data flow

164 Release 5

(for example, updating the weekly table SENTRY_N_NT_W from the informationused for updating the daily table SENTRY_N_NT_D).

¶ Lookup tables (which contain user-defined information that defines an organization’soperating environment) to add user-defined data to the Performance Reporter tablerecord.

A description of the collect procedure is provided in the Language Guide and Reference.

A description of the use of control tables is provided in the Administration Guide.


A description of how to create new reports is provided in the Guide to the ReportingDialog.

The reports that can be created when using Distributed Monitoring are described in“Reports” on page 173.


16.D

ataF

low

and

Rep

orter

Ob

jects

166 Release 5

Data Tables, Lookup Tables, and ControlTables



¶ The format that is used for defining Distributed Monitoring table names and views.

¶ The layout of the tables that are used in this chapter to describe each DistributedMonitoring Component (Sentry) table and lookup table.

Naming standard for tablesThe names of Distributed Monitoring Component (Sentry) tables use the following format:

SENTRY_N_content_suffix

where:

¶ content is a description (for example, NT for data coming from Distributed Monitoringrunning on Windows NT, and UN for data coming from Distributed Monitoring runningon UNIX.

¶ suffix indicates the summarization level of the data in the table (for example,SENTRY_N_NT._D for command statistics summarized by day).

A table name can have these summarization-level suffixes:_H The table holds data summarized by hour(hourly data)._D The table holds data summarized by day(daily data)._M The table holds data summarized by month(monthly data).

Lookup tables do not have a suffix.


¶ Key columns are marked with a k. They are sorted in the sequence they appear in thetable.

¶ Data columns follow the last key column.

17


17.D

ata,L

oo

kup

,an

dC

on

trol

Tables

The descriptions of most key columns and data columns contain references to the fieldsfrom which they are derived in the record. For an explanation of these fields, refer to theapplicable product documentation.

SENTRY_N_NT_H,_D,_WThese tables provide hourly, daily, and weekly statistics on Distributed Monitoring data forWindows NT. The source file is in the member DRLTXSEN, which is contained in thedataset xxxxxxxx.SDRLDEFS, where xxxxxxxx is a high-level identifier.

The default retention periods are:¶ 7 days for SENTRY_N_NT_H¶ 31 days for SENTRY_N_NT_D¶ 400 days for SENTRY_N_NT_W

Primary keys: DTE, TME, NODE_NAME, PERIOD_NAME, and STATUS

Table 14. Distributed Monitoring table definition for NT dataColumn name Data type Description

DTE k DATE Date when the record was written. For SENTRY_N_NT_W this is the date ofthe first day of the week. From DTE.

TME k CHAR(8). Time when the record was written. The format is HH.MM.SS. From TME. Thisfield is present only in SENTRY_N_NT_H.


PERIOD_NAME k CHAR(8). Name of the period. The name is derived by using the fields NODE_NAME,DTE, and TME from the record as parameters in the PERIOD function. If nomatch is found, this column is set to "?".

STATUS k CHAR(8). The status of the monitor as set by Distributed Monitoring. From STATUS.

MONITOR1.....60 FLOAT Monitor value.

MONITORC1.....60 INTEGER Number of monitor occurrences in the log file.

168 Release 5

SENTRY_N_UN_H,_D,_WThese tables provide hourly, daily, and weekly statistics on Distributed Monitoring data forUNIX. The source file is in the member DRLTXSEN, which is contained in the datasetxxxxxxxx.SDRLDEFS, where xxxxxxxx is a high-level identifier.

The default retention periods are:¶ 7 days for SENTRY_N_UN_H¶ 31 days for SENTRY_N_UN_D¶ 400 days for SENTRY_N_UN_W

Primary keys: DTE, TME, NODE_NAME, PERIOD_NAME, and STATUS.

Table 15. Distributed Monitoring table definition for UNIX dataColumn name Data type Description

DTE k DATE Date when the record was written. For SENTRY_N_UN_W this is the date ofthe first day of the week. From DTE.

TME k CHAR(8). Time when the record was written. The format is HH.MM.SS. From TME.This field is present only in SENTRY_N_UN_H.


PERIOD_NAME k CHAR(8). Name of the period. The name is derived by using fields NODE_NAME,DTE, and TME from the record as parameters in the PERIOD function. If nomatch is found, this column is set to "?".

STATUS k CHAR(8). The status of the monitor as set by Distributed Monitoring. From STATUS.

MONITOR61.....99 FLOAT Monitor value.

MONITORC61.....99 INTEGER Number of monitor occurrences in the log file.

The various monitors are defined and can be customized in the lookup tablesSENTRY_MONITOR_NT (monitors 1 through 60) and SENTRY_MONITOR_UN (monitors61 through 99). These lookup tables are defined in the member DRLTXSEL, which iscontained in the dataset xxxxxxxx.SDRLDEFS, where xxxxxxxx is a high-level identifier.


17.D

ata,L

oo

kup

,an

dC

on

trol

Tables

Lookup TablesThis section describes the X_NODE_NAME, SENTRY_MONITOR_NT, andSENTRY_MONITOR_UN lookup tables.

For descriptions of common lookup tables used by other Performance Reporter features,refer to the Administration Guide.








PERIOD_PLAN CHAR(8) Period plan. This value can be used as a parameter when using the PERIODfunction looking up PERIOD_NAME.

Figure 47. X_NODE_NAME lookup table

170 Release 5

SENTRY_MONITOR_NTThis lookup table for Windows NT contains node names and their associated types ofmonitors and unique identifiers.


NODE MONITOR MONITORNAME DESC MONITOR N---------------- --------------------------- -------- -------% File Read Bytes/sec% FRDBYTES 1% File Read Operations/sec% FRDOPER 2% File Write Bytes/sec% FWRBYTES 3% File Write Operations/sec% FWROPER 4

SENTRY_MONITOR_UNThis lookup table for UNIX contains node names and their associated types of monitors andunique identifiers.


NODE MONITOR MONITORNAME DESC MONITOR N---------------- --------------------------- -------- -------% Inodes free(/) INODEFR 61% Inodes used(/) INODEUS 62% Percent inodes used(/) PINODEUS 63



MONITOR_DESC k VARCHAR(80) Type of monitor.

MONITOR CHAR(8) Unique identifier for the monitor.

MONITOR_N SMALLINT Unique internal monitor number.

Figure 48. SENTRY_MONITOR_NT lookup table


NODE_NAME k CHAR(16) Node name. This is an alias name for the node. This may contain global searchcharacters.

MONITOR_DESC k VARCHAR(80) Type of monitor.

MONITOR CHAR(8) Unique identifier for the monitor.

MONITOR_N SMALLINT Unique internal monitor number.

Figure 49. SENTRY_MONITOR_UN lookup table


17.D

ata,L

oo

kup

,an

dC

on

trol

Tables

172 Release 5

Reports

This chapter describes the reports provided with the Distributed Monitoring Component(Sentry).

Distributed Monitoring report definitionsTwo new report groups have been defined. They contain a total of 14 new reports:

The TMEUNIX Tivoli Distributed Monitoring UNIX report group, with the following 6reports:

1. Distributed Monitoring Service Level Details Hourly (SEQ01)

2. Distributed Monitoring Service Level Details Daily (SEQ02)

3. Distributed Monitoring Service Level Details Weekly (SEQ03)

4. Distributed Monitoring System Resources Monitors Daily (SEQ04)

5. Distributed Monitoring Network Monitors Daily (SEQ05)

6. Distributed Monitoring Disk Resource Monitors Daily (SEQ06)

The TMENTTivoli Distributed Monitoring NT report group, with the following 8 reports:

1. Distributed Monitoring Service Level Details Hourly (SEQ07)

2. Distributed Monitoring Service Level Details Daily (SEQ08)

3. Distributed Monitoring Service Level Details Weekly (SEQ09)

4. Distributed Monitoring TCP Monitors Daily (SEQ10)

5. Distributed Monitoring System Monitors Daily (SEQ11)

6. Distributed Monitoring Server Monitors Daily (SEQ12)

7. Distributed Monitoring Processor Monitors Daily (SEQ13)

8. Distributed Monitoring Memory Details Daily (SEQ14)

The source file for the group definitions is in the member DRLOXSEN, which is containedin the dataset xxxxxxxx.SDRLRENU, where xxxxxxxx is a high-level identifier. The sourcefiles for the related queries and forms for the report definitions are in the membersDRLQSEyy and DRLFSEyy respectively, where yy can be a number from 01 to 14. Theyare located in the dataset xxxxxxxx.SDRLRENU, where xxxxxxxx is a high-level identifier.

See Table 12 on page 151 for a list of the predefined reports that are provided with theDistributed Monitoring component of Performance Reporter.

18


18.R

epo

rts

Reports

DM Service Level Details Hourly (UNIX)The following information identifies the report:

Report ID SEQ01

Report group Distributed Monitoring (Sentry) UNIX Reports

Source SENTRY_N_UN_H

Attributes AIX, SERVICE, OVERVIEW, UNIX

Variables DATE, NODE_NAME



HOURINODES_FREE(/)INODES_USED(/)PERCENT_INODES_USED(/)PERCENT_SPACE_USED(/)SPACE_FREE(/)SPACE_USED(/)INODES_FREE(/TMP)INODES_USED(/TMP)

DM Service Level Details Hourly report (UNIX)

PERCENT PERCENT PERCENTINODES INODES INODES SPACE SPACE SPACE INODES INODES INODES

HOUR FREE(/) USED(/) USED(/) USED(/) FREE(/) USED(/) FREE(/TMP) USED(/TMP) USED(/TMP)-------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------12 - - - - - - - - -

TIVOLIPERCENT PERCENT PERCENT DBSPACE SPACE SPACE INODES INODES INODES SPACE SPACE SPACE FREE

USED(/TMP) FREE(/TMP) USED(/TMP) FREE(/USR) USED(/USR) USED(/USR) USED(/USR) FREE(/USR) USED(/USR) SPACE----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------

- - - - - - - - - -

USERLOGINS USER CLIENT INPUT

FILE FILE BY LOGGED RPC NETWORK NETWORK NFS BAD PACKET INPUTCHECKSUM SIZE USER IN TIMEOUT COLLISIONS COLLISIONS/ CALLS ERRORS PACKETS

----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

JOBSOUTPUT RPC AVAILABLE LINGERING MAIL IN TOTALPACKET OUTPUT BAD SWAP TERMINATED LOAD QUEUE PAGE-OUTS PRINT SIZEERRORS PACKETS CALLS SPACE(AVG) PROC. AVERAGE LENGHT(AVG) (AVG) QUEUE(AVG) QUEUE(AVG)

----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

Performance Reporter: SEQ01

Figure 50. DM Service Level Details Hourly Report (UNIX)

174 Release 5

PERCENT_INODES_USED(/TMP)PERCENT_SPACE_USED(/TMP)SPACE_FREE(/TMP)SPACE_USED(/TMP)INODES_FREE(/USR)INODES_USED(/USR)PERCENT_INODES_USED(/USR)PERCENT_SPACE_USED(/USR)SPACE_FREE(/USR)SPACE_USED(/USR)TIVOLI_DB_FREE_SPACEFILE_CHECKSUMFILE_SIZEUSER_LOGINS_BY_USERUSER_LOGGED_INCLIENT_RPC_TIMEOUTNETWORK_COLLISIONSNETWORK_COLLISIONS/PCKTNFS BAD_CALLSINPUT_PACKET_ERRORSINPUT_PACKETSOUTPUT_PACKET_ERRORSOUTPUT_PACKETSRPC_BAD_CALLSAVAILABLE_SWAP_SPACE(AVG)LINGERING_TERMINATED_PROC.LOAD_AVERAGEMAIL_QUEUE_LENGTH(AVG)PAGE-OUTS_(AVG)JOBS_IN_PRINT_QUEUE(AVG)TOTAL_SIZE_QUEUE(AVG)


18.R

epo

rts

DM Service Level Details Daily (UNIX)The following information identifies the report:

Report ID SEQ02

Report group Distributed Monitoring UNIX Reports

Source SENTRY_N_UN_D





DATEINODES_FREE(/)INODES_USED(/)PERCENT_INODES_USED(/)PERCENT_SPACE_USED(/)SPACE_FREE(/)SPACE_USED(/)INODES_FREE(/TMP)INODES_USED(/TMP)PERCENT_INODES_USED(/TMP)PERCENT_SPACE_USED(/TMP)

DM Service Level Details Daily report (UNIX)


DATE FREE(/) USED(/) USED(/) USED(/) FREE(/) USED(/) FREE(/TMP) USED(/TMP) USED(/TMP)---------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------1998-04-09 - - - - - - - - -



- - - - - - - - - -



----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -


----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -


Figure 51. DM Service Level Details Daily Report (UNIX)

176 Release 5

SPACE_FREE(/TMP)SPACE_USED(/TMP)INODES_FREE(/USR)INODES_USED(/USR)PERCENT_INODES_USED(/USR)PERCENT_SPACE_USED(/USR)SPACE_FREE(/USR)SPACE_USED(/USR)TIVOLI_DB_FREE_SPACEFILE_CHECKSUMFILE_SIZEUSER_LOGINS_BY_USERUSER_LOGGED_INCLIENT_RPC_TIMEOUTNETWORK_COLLISIONSNETWORK_COLLISIONS/PCKTNFS BAD_CALLSINPUT_PACKET_ERRORSINPUT_PACKETSOUTPUT_PACKET_ERRORSOUTPUT_PACKETSRPC_BAD_CALLSAVAILABLE_SWAP_SPACE(AVG)LINGERING_TERMINATED_PROC.LOAD_AVERAGEMAIL_QUEUE_LENGTH(AVG)PAGE-OUTS_(AVG)JOBS_IN_PRINT_QUEUE(AVG)TOTAL_SIZE_QUEUE(AVG)


18.R

epo

rts

DM Service Level Details Weekly (UNIX)The following information identifies the report:

Report ID SEQ03


Source SENTRY_N_UN_W


Variables FROM_DATE, TO_DATE, NODE_NAME



DATEINODES_FREE(/)INODES_USED(/)PERCENT_INODES_USED(/)PERCENT_SPACE_USED(/)SPACE_FREE(/)SPACE_USED(/)INODES_FREE(/TMP)INODES_USED(/TMP)PERCENT_INODES_USED(/TMP)PERCENT_SPACE_USED(/TMP)

DM Service Level Details Weekly report (UNIX)


DATE FREE(/) USED(/) USED(/) USED(/) FREE(/) USED(/) FREE(/TMP) USED(/TMP) USED(/TMP)---------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------1998-04-08 - - - - - - - - -



- - - - - - - - - -



----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -


----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -


Figure 52. DM Service Level Details Weekly Report (UNIX)

178 Release 5

SPACE_FREE(/TMP)SPACE_USED(/TMP)INODES_FREE(/USR)INODES_USED(/USR)PERCENT_INODES_USED(/USR)PERCENT_SPACE_USED(/USR)SPACE_FREE(/USR)SPACE_USED(/USR)TIVOLI_DB_FREE_SPACEFILE_CHECKSUMFILE_SIZEUSER_LOGINS_BY_USERUSER_LOGGED_INCLIENT_RPC_TIMEOUTNETWORK_COLLISIONSNETWORK_COLLISIONS/PCKTNFS BAD_CALLSINPUT_PACKET_ERRORSINPUT_PACKETSOUTPUT_PACKET_ERRORSOUTPUT_PACKETSRPC_BAD_CALLSAVAILABLE_SWAP_SPACE(AVG)LINGERING_TERMINATED_PROC.LOAD_AVERAGEMAIL_QUEUE_LENGTH(AVG)PAGE-OUTS_(AVG)JOBS_IN_PRINT_QUEUE(AVG)TOTAL_SIZE_QUEUE(AVG)


18.R

epo

rts

DM System Resources Monitors Daily (UNIX)The following information identifies the report:

Report ID SEQ04



Attributes AIX, SERVICE, OVERVIEW, UNIX, SYSTEM

Variables FROM_DATE, TO_DATE, NODE_NAME, PERIOD_NAME



DATEAVAILABLE_SWAP_SPACE_(AVG)LOAD_AVERAGEMAIL_QUEUE_LENGTH_(AVG)PAGE-OUTS_(AVG)

DM System Resources Monitors Daily report (UNIX)

AVAILABLE MAILSWAP QUEUESPACE LOAD LENGHT PAGE-OUTS

DATE (AVG) AVERAGE (AVG) (AVG)---------- ----------- ----------- ----------- -----------1998-04-09 - - - -


Figure 53. DM System Resources Monitors Daily Report (UNIX)

180 Release 5

DM Network Monitors Daily (UNIX)The following information identifies the report:

Report ID SEQ05



Attributes AIX, SERVICE, OVERVIEW, UNIX, NETWORK




DATENETWORK_COLLISIONSNETWORK_COLLISIONS_/PCKTINPUT_PACKET_ERRORSINPUT_PACKETSOUTPUT_PACKET_ERRORSOUTPUT_PACKETS

DM Network Monitors Daily report (UNIX)

NETWORK INPUT OUTPUTNETWORK COLLISIONS PACKET INPUT PACKET OUTPUT

DATE COLLISIONS /PCKT ERRORS PACKETS ERRORS PACKETS---------- ----------- ----------- ----------- ----------- ----------- -----------1998-04-09 - - - - - -


Figure 54. DM Network Monitors Daily Report (UNIX)


18.R

epo

rts

DM Disk Resource Monitors Daily (UNIX)The following information identifies the report:

Report ID SEQ06



Attributes AIX, SERVICE, OVERVIEW, UNIX, DISK




DATEINODES_FREE(/)INODES_USED(/)SPACE_FREE(/)SPACE_USED(/)INODES_FREE(/TMP)INODES_USED(/TMP)SPACE_FREE(/TMP)SPACE_USED(/TMP)INODES_FREE(/USR)INODES_USED(/USR)SPACE_FREE(/USR)SPACE_USED(/USR)TIVOLI_DB_FREE_SPACE

DM Disk Resource Monitors Daily report (UNIX)

INODES INODES SPACE SPACE INODES INODES SPACE SPACE INODESDATE FREE(/) USED(/) FREE(/) USED(/) FREE(/TMP) USED(/TMP) FREE(/TMP) USED(/TMP) FREE(/USR)---------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------1998-04-09 - - - - - - - - -

TIVOLIDB

INODES SPACE SPACE FREEUSED(/USR) FREE(/USR) USED(/USR) SPACE

----------- ----------- ----------- ------------ - - -


Figure 55. DM Disk Resource Monitors Daily Report (UNIX)

182 Release 5

DM Service Level Details Hourly (NT)The following information identifies the report:

Report ID SEQ07

Report group Distributed Monitoring NT Reports

Source SENTRY_N_NT_H

Attributes NT, SERVICE, OVERVIEW


Figure 56 shows an example of this report.

DM Service Level Details Hourly report (NT)

FILE FILE PERCENT PERCENT PERCENTFILE READ FILE WRITE TOT. TOT. TOT. PAGEREAD OPERATIONS WRITE OPERATIONS INT. PRIV. USER TOTAL FAULTS/SEC

HOUR BYTES(AVG) (AVG) BYTES(AVG) (AVG) TIME(%) TIME(%) TIME(%) INTERRUPTS/ (AVG)-------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------17 0 29 0 210 44 - - - -

BYTES CONTEXTPAGE PAGE TRANSITION WRITE TOTAL/SEC BLOCK ERRORS ERRORS

READS/SEC INPUT/SEC PAGES/SEC FAULTS/SEC COPIES/SEC (NT SER.) QUEUED/SEC ACCESS GRANTED ERRORS(AVG) (AVG) (AVG)' (AVG) (AVG) (AVG) (AVG) PERMISSIONS ACCESS LOGON

----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - 1 - - - - - - -

BYTES FILE READ WRITEFILES TOTAL/SEC DATA PACKETS/SEC BYTES BYTESOPENED LOGON/SEC (REDIRE.) OPERATIONS/ NETWORK (REDIRECTOR NETWORK/SEC NETWORK/SEC SEGMENT SEGMENTTOTAL (AVG) (AVG) (AVG) ERRORS (AVG) (AVG) (AVG) RECV/SEC RETRANS/SEC

----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

BYTES FRAMES BYTES FRAMESTOTAL/SEC REJ./SEC FRAMES/SEC PACKETS/SEC TOTAL/SEC REJ./SEC FRAMES/SEC PACKETS/SEC

SEGMENT (NW IPX) (NW IPX) (NW IPX) (NW IPX) (NW BIOS) (NW BIOS) (NW BIOS) (NW BIOS) INTERRUPTSSENT/SEC (AVG) (AVG) (AVG) (AVG) (AVG) (AVG) (AVG) (AVG) /SEC

----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

AVGREAD WRITE DISK

PERCENT PERCENT PERCENT BYTES OPERATIONS WRITE OPERATIONS BYTES/READINTERRUPT PRIVILEGED USER RECV/SEC BYTES SENT/ READ BYTES/ /SEC BYTES/SEC /SEC (LOG)TIME(%) TIME(%) TIME(%) (AVG) (AVG) (AVG) (AVG) (AVG) (AVG) (AVG)

----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

Figure 56. DM Service Level Details Hourly Report (UNIX) (Part 1 of 2)


18.R

epo

rts


HOURFILE_READ_BYTES(AVG)FILE_READ_OPERATIONS_(AVG)FILE_WRITE_BYTES(AVG)FILE_WRITE_OPERATIONS_(AVG)PERCENT_TOT._INT._TIME(%)PERCENT_TOT._PRIV._TIME(%)PERCENT_TOT._USER_TIME(%)TOTAL_INTERRUPTS/SECPAGE_FAULTS/SEC_(AVG)PAGE_READS/SEC_(AVG)PAGE_INPUT/SEC_(AVG)PAGES/SEC_(AVG)TRANSITION_FAULTS/SEC_(AVG)WRITE_ COPIES/SEC_(AVG)BYTES_TOTAL/SEC_(NT SER.)_(AVG)CONTEXT_BLOCK_QUEUED/SEC_(AVG)ERRORS_ACCESS_PERMISSIONSERRORS_GRANTED_ACCESSERRORS_LOGONFILES_OPENED_TOTALLOGON/SEC_(AVG)BYTES_TOTAL/SEC_(REDIRE.)_(AVG)FILE_DATA_OPERATIONS/SEC_(AVG)NETWORK_ERRORSPACKETS/SEC_(REDIRECTOR)_(AVG)READ_BYTES_NETWORK/SEC_(AVG)WRITE_BYTES_NETWORK/SEC_(AVG)SEGMENT_RECV/SECSEGMENT_RETRANS/SECSEGMENT_SENT/SECBYTES_TOTAL/SEC_(NW IPX)_(AVG)FRAMES_REJ./SEC_(NW IPX)_(AVG)FRAMES/SEC_(NW IPX)_(AVG)

PERCENTAVG PERCENT PERCENT AVG DISK

DISK DISK DISK READ DISK WRITE DISK AVG READBYTES/WRITE READS/SEC DISK TIME TIME BYTES/READ DISK DISK DISK TIME

(LOG) (LOG) WRITES/SEC( (LOG) (LOG) (PHY) BYTES/WRITE READS/SEC(P WRITES/SEC( (PHY)(AVG) (AVG) (AVG) (%) (%) (AVG) (AVG) (AVG) (AVG) (%)

----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

PERCENTDISKWRITETIME(PHY)(%)

------------


Figure 56. DM Service Level Details Hourly Report (UNIX) (Part 2 of 2)

184 Release 5

PACKETS/SEC_(NW IPX)_(AVG)BYTES_TOTAL/SEC_(NW BIOS)_(AVG)FRAMES_REJ./SEC_(NW BIOS)_(AVG)FRAMES/SEC_(NW BIOS)_(AVG)PACKETS/SEC_(NW BIOS)_(AVG)INTERRUPTS_/SECPERCENT_INTERRUPT_TIME(%)PERCENT_PRIVILEGED_TIME(%)PERCENT_USER_TIME(%)BYTES_RECV/SEC_(AVG)BYTES SENT/SEC_(AVG)READ BYTES/SEC_(AVG)READ_OPERATIONS_/SEC_(AVG)WRITE_BYTES/SEC_(AVG)WRITE_OPERATIONS_/SEC_(AVG)AVG_DISK_BYTES/READ_(LOG)_(AVG)AVG_DISK_BYTES/WRITE_(LOG)_(AVG)DISK_READS/SEC_(LOG)_(AVG)DISK_WRITES/SEC(LOG)_(AVG)PERCENT_DISK READ_TIME_(LOG)_(%)PERCENT_DISK WRITE_TIME_(LOG)_(%)AVG_DISK_BYTES/READ_(PHY)_(AVG)AVG_DISK_BYTES/WRITE(PHY)_(AVG)DISK_READS/SEC(PHY)_(AVG)DISK_WRITES/SEC(PHY)_(AVG)PERCENT_DISK_READ_TIME_(PHY)_(%)PERCENT_DISK_WRITE_TIME_(PHY)_(%)


18.R

epo

rts

DM Service Level Details Daily (NT)The following information identifies the report:

Report ID SEQ08


Source SENTRY_N_NT_D


Variables FROM_DATE, TO_DATE, PERIOD_NAME, NODE_NAME


DM Service Level Details Daily report (NT)


DATE BYTES(AVG) (AVG) BYTES(AVG) (AVG) TIME(%) TIME(%) TIME(%) INTERRUPTS/ (AVG)---------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------1998-02-18 - - - - - - - - -



----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------208 - 2 - - - - - - -


----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -



----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

AVGREAD WRITE DISK


----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

Figure 57. DM Service Level Details Daily Report (NT) (Part 1 of 2)

186 Release 5


DATAFILE_READ_BYTES(AVG)FILE_READ_OPERATIONS_(AVG)FILE_WRITE_BYTES(AVG)FILE_WRITE_OPERATIONS_(AVG)PERCENT_TOT._INT._TIME(%)PERCENT_TOT._PRIV._TIME(%)PERCENT_TOT._USER_TIME(%)TOTAL_INTERRUPTS/SECPAGE_FAULTS/SEC_(AVG)PAGE_READS/SEC_(AVG)PAGE_INPUT/SEC_(AVG)PAGES/SEC_(AVG)TRANSITION_FAULTS/SEC_(AVG)WRITE_ COPIES/SEC_(AVG)BYTES_TOTAL/SEC_(NT SER.)_(AVG)CONTEXT_BLOCK_QUEUED/SEC_(AVG)ERRORS_ACCESS_PERMISSIONSERRORS_GRANTED_ACCESSERRORS_LOGONFILES_OPENED_TOTALLOGON/SEC_(AVG)BYTES_TOTAL/SEC_(REDIRE.)_(AVG)FILE_DATA_OPERATIONS/SEC_(AVG)NETWORK_ERRORSPACKETS/SEC_(REDIRECTOR)_(AVG)READ_BYTES_NETWORK/SEC_(AVG)WRITE_BYTES_NETWORK/SEC_(AVG)SEGMENT_RECV/SECSEGMENT_RETRANS/SECSEGMENT_SENT/SECBYTES_TOTAL/SEC_(NW IPX)_(AVG)FRAMES_REJ./SEC_(NW IPX)_(AVG)FRAMES/SEC_(NW IPX)_(AVG)




----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -


------------


Figure 57. DM Service Level Details Daily Report (NT) (Part 2 of 2)


18.R

epo

rts


188 Release 5

DM Service Level Details Weekly (NT)The following information identifies the report:

Report ID SEQ09


Source SENTRY_N_NT_W




DM Service Level Details Weekly report (NT)


DATE BYTES(AVG) (AVG) BYTES(AVG) (AVG) TIME(%) TIME(%) TIME(%) INTERRUPTS/ (AVG)---------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------1998-02-17 - - - - - - - - -



----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------208 - 2 - - - - - - -


----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -



----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

AVGREAD WRITE DISK


----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -

Figure 58. DM Service Level Details Weekly Report (NT) (Part 1 of 2)


18.R

epo

rts


DATEFILE_READ_BYTES(AVG)FILE_READ_OPERATIONS_(AVG)FILE_WRITE_BYTES(AVG)FILE_WRITE_OPERATIONS_(AVG)PERCENT_TOT._INT._TIME(%)PERCENT_TOT._PRIV._TIME(%)PERCENT_TOT._USER_TIME(%)TOTAL_INTERRUPTS/SECPAGE_FAULTS/SEC_(AVG)PAGE_READS/SEC_(AVG)PAGE_INPUT/SEC_(AVG)PAGES/SEC_(AVG)TRANSITION_FAULTS/SEC_(AVG)WRITE_ COPIES/SEC_(AVG)BYTES_TOTAL/SEC_(NT SER.)_(AVG)CONTEXT_BLOCK_QUEUED/SEC_(AVG)ERRORS_ACCESS_PERMISSIONSERRORS_GRANTED_ACCESSERRORS_LOGONFILES_OPENED_TOTALLOGON/SEC_(AVG)BYTES_TOTAL/SEC_(REDIRE.)_(AVG)FILE_DATA_OPERATIONS/SEC_(AVG)NETWORK_ERRORSPACKETS/SEC_(REDIRECTOR)_(AVG)READ_BYTES_NETWORK/SEC_(AVG)WRITE_BYTES_NETWORK/SEC_(AVG)SEGMENT_RECV/SECSEGMENT_RETRANS/SECSEGMENT_SENT/SECBYTES_TOTAL/SEC_(NW IPX)_(AVG)FRAMES_REJ./SEC_(NW IPX)_(AVG)FRAMES/SEC_(NW IPX)_(AVG)




----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- ------------ - - - - - - - - -


------------


Figure 58. DM Service Level Details Weekly Report (NT) (Part 2 of 2)

190 Release 5



18.R

epo

rts

DM TCP Monitors Daily (NT)The following information identifies the report:

Report ID SEQ10



Attributes NT, SERVICE, OVERVIEW, TCP




DATESEGMENT_RECEIVED/SECSEGMENT_RETRANS/SECSEGMENT_SENT/SEC

DM TCP Monitors Daily report (NT)

SEGMENT SEGMENT SEGMENTDATE RECEIVED/SE RETRANS/SEC SENT/SEC---------- ----------- ----------- -----------1998-02-18 - - -


Figure 59. DM TCP Monitors Daily Report (NT)

192 Release 5

DM System Monitors Daily (NT)The following information identifies the report:

Report ID SEQ11



Attributes NT, SERVICE, OVERVIEW, SYSTEM




DATEFILE READ_BYTES_(AVG)FILE READ_OPERATIONS_(AVG)FILE WRITE_BYTES_(AVG)FILE WRITE_OPERATIONS_(AVG)PERCENT_TOT. INT._TIME_(%)PERCENT_TOT. PRIV._TIME_(%)PERCENT_TOT. USER_TIME_(%)TOTAL_INTERRUPTS_/SEC

DM System Monitors Daily report (NT)

PERCENT PERCENT PERCENTFILE READ FILE READ FILE WRITE FILE WRITE TOT. INT. TOT. PRIV. TOT. USER TOTALBYTES OPERATIONS BYTES OPERATIONS TIME TIME TIME INTERRUPTS

DATE (AVG) (AVG) (AVG) (AVG) (%) (%) (%) /SEC---------- ----------- ----------- ----------- ----------- ----------- ----------- ----------- -----------1998-02-18 - - - - - - - -


Figure 60. DM System Monitors Daily Report (NT)


18.R

epo

rts

DM Server Monitors Daily (NT)The following information identifies the report:

Report ID SEQ12



Attributes NT, SERVICE, OVERVIEW, SERVER




DATEBYTES_TOTAL/SEC_(NT SER.)_(AVG)CONTEXT_BLOCK_QUEUED/SEC_(AVG)ERRORS_ACCESS_PERMISSIONSERRORS_GRANTED_ACCESSERRORS_LOGONFILES_OPENED_TOTALLOGON/SEC_(AVG)

DM Server Monitors Daily report (NT)

BYTES CONTEXTTOTAL/SEC BLOCK ERRORS ERRORS FILES(NT SER.) QUEUED/SEC ACCESS GRANTED ERRORS OPENED LOGON/SEC

DATE (AVG) (AVG) PERMISSIONS ACCESS LOGON TOTAL (AVG)---------- ----------- ----------- ----------- ----------- ----------- ----------- -----------1998-02-18 - - - - - - -


Figure 61. DM Server Monitor Daily (NT)

194 Release 5

DM Processor Monitors Daily (NT)The following information identifies the report:

Report ID SEQ13



Attributes NT, SERVICE, OVERVIEW, PROCESSOR




DATEINTERRUPTS_/SECPERCENT_INTERRUPT_TIME_(%)PERCENT_PRIVILEGED_TIME_(%)PERCENT_USER_TIME_(%)

DM Processor Monitors Daily report (NT)

PERCENT PERCENT PERCENTINTERRUPT PRIVILEGED USER

INTERRUPTS TIME TIME TIMEDATE /SEC (%) (%) (%)---------- ----------- ----------- ----------- -----------1998-02-18 - - - -


Figure 62. DM Processor Monitors Daily Report (NT)


18.R

epo

rts

DM Memory Details Daily (NT)The following information identifies the report:

Report ID SEQ14



Attributes NT, SERVICE, OVERVIEW, MEMORY




DATEPAGE_FAULTS/SEC_(AVG)PAGE_READS/SEC_(AVG)PAGE_INPUT/SEC_(AVG)PAGES/SEC_(AVG)TRANSITION_FAULTS/SEC_(AVG)WRITE_COPIES/SEC_(AVG)

DM Memory Details Daily report (NT)

PAGE PAGE PAGE TRANSITION WRITEFAULTS/SEC READS/SEC INPUT/SEC PAGES/SEC FAULTS/SEC COPIES/SEC

DATE (AVG) (AVG) (AVG) (AVG) (AVG) (AVG)---------- ----------- ----------- ----------- ----------- ----------- -----------1998-02-18 - - - - - -


Figure 63. DM Memory details Daily Report (NT)

196 Release 5


Distributed Monitoring log collectorThe log collector component processes the log file. During the processing, the DistributedMonitoring data is stored in the database with the needed summarizations. When theprocessing is complete you can use both graphical and tabular reports to display the data.

Fourteen reports are provided with the Distributed Monitoring Component (Sentry) featureand users can also use queries to create customized reports.

The definitions for the log collector are contained in the log definition, the record definition,and the table definition.

¶ The log definition describes the format of the log file to the log collector.

¶ The record definition describes the record format of the Distributed MonitoringComponent (Sentry) application to the log collector.

¶ The table definition consists of two parts:

v The table definition that tells the log collector how the table for DistributedMonitoring Component (Sentry) data is built. This table will contain the data.

v The update table definition that tells the log collector how to process the data thatcomes from the log file before the log collector stores the data in the data table.

The report definition describes the report layouts, data, and defaults.

Distributed Monitoring log definitionsThe Distributed Monitoring Component (Sentry) log file is a flat file that contains severalrecords. Each record is written by the Distributed Monitoring Component (Sentry)application. The format of each record is defined in the member DRLLXSEN, which iscontained in the dataset xxxxxxxx.SDRLDEFS, where xxxxxxxx is a high-level identifier. Itsformat is as follows:

Table 16. Distributed Monitoring log file formatField Offset Length Type Description





19


19.L

og

Reco

rdD

efinitio

ns

Distributed Monitoring record definitionsThe Distributed Monitoring record definition describes to the log collector the data that iswritten to the log file. The record format for the PRSENTRYN record is defined in themember DRLRXSEN, which is contained in the dataset xxxxxxxx.SDRLDEFS, wherexxxxxxxx is a high-level identifier.

It is identified by:

RECORD_TYPE = PRSENTRYN

and its fields are:

Table 17. Distributed Monitoring record definition fieldsField Offset Length Type Description





MONITOR_PROFILE 41 20 CHAR Profile name

MONITOR_STATUS 61 8 CHAR Status

MONITOR_CVALUE 69 16 EXTERNAL FLOAT Current value

MONITOR_PVALUE 85 16 EXTERNAL FLOAT Previous value

MONITOR_DESC 101 * CHAR(*) Description

198 Release 5

Part VII — Appendixes


200 Release 5

List of abbreviations

These abbreviations appear in this book:

ARM Application Response Measurement

CD-ROM compact disk—read-only memory

CMF CICS monitoring facility

CPU central processing unit

DB2 DATABASE 2

DL/I Data Language/I

DM Distributed Monitoring

IMS Information Management System

I/O input/output

IS information system

ISPF Interactive System Productivity Facility

NPM NetView Performance Monitor

P.R. Acronym for Performance Reporter for OS/390 (only for specific cases).

SMF system management facilities

SQL Structured Query Language

SRB service control block

TCB task control block

TCP/IP Transmission Control Protocol/Internet Protocol

TSO time-sharing option


List

of

abb

reviation

s

202 Release 5

Glossary

A

administrationA Performance Reporter task that includes maintaining the database, updating environment information, andensuring the accuracy of data collected.

administration dialogThe set of host windows used to administer Performance Reporter.

agentIn UNIX Performance feature, a set of programs running on a network node that gather data about the node andwrite the data to log files. Agents can gather data by issuing UNIX commands, and by reading files such asUNIX accounting files.

C

COLLECTA process used by Performance Reporter to read data from input log data sets, interpret records in the data set,and store the data in DB2 tables in the Performance Reporter database.

componentAn optionally installable part of an Performance Reporter feature.

control tableA predefined Performance Reporter table that controls results returned by some log collector functions.

D

Data Language/I (DL/I)An IBM database-management facility.

data tableA Performance Reporter table that contains performance data used to create reports.

distributed transaction processingThe distribution of processing among transactions that communicate synchronously with each other overintersystem or interregion links.

E

environment informationAll of the information that is added to the log data to create reports. This information can include data such asperformance groups, shift periods, installation definitions, and so on.

G

gatherA process performed by agents in which performance data is entered into log files. The log files are located onthe same nodes as the agents.

H

help topicsAn online table of contents for the Performance Reporter online help information.


Glo

ssary

K

key columnsThe columns of a DB2 table that together constitute the key.

key valueValue used to sort records into groups.

L

logAny sequential data set used as input to Performance Reporter.

log collectorA Performance Reporter program that processes log data sets and provides other services.

log collector languagePerformance Reporter statements used to supply definitions to and invoke services of the log collector.

log data setAny sequential data set used as input to Performance Reporter.

log definitionThe description of a log data set processed by the log collector.

log procedureA program module that is used to process all record types in certain log data sets.

logical unit (LU)A port through which a user gains access to the services of a network.

lookup expressionAn expression that specifies how a value is obtained from a lookup table.

lookup tableA Performance Reporter DB2 table that contains grouping, translation, or substitution information.

O

objectAn integral part of a feature component needed for data collection (for example, record definitions, recordprocedures, and update definitions).

P

Performance Reporter databaseA set of DB2 tables that includes data tables, lookup tables, system tables, and control tables.

purge conditionInstruction for purging unneeded data from the Performance Reporter database.

R

record definitionThe description of a record type contained in the log data sets used by Performance Reporter, including detailedrecord layout and data formats.

record procedureA program module that is called to process some types of log records.

record typeThe classification of records in a log data set.

204 Release 5

regionA section of the dynamic area that is allocated to a job step or system task.

report definition languagePerformance Reporter statements used to define reports and report groups.

report groupA collection of Performance Reporter reports that can be referred to by a single name.

reporting dialogA set of host or workstation windows used to request reports.

resourceAny facility of the computing system or operating system required by a job or task, including central storage,input/output devices, the processing unit, data sets, and control or processing programs.

resource groupA collection of resources identified as belonging to a particular department or division. Resources are organizedinto groups to reflect the structure of an organization.

resource informationEnvironment information that describes the elements in a system (for example, a network).

S

sectionA structure within a record that contains one or more fields and may contain other sections.

sourceIn an update definition, the record or DB2 table that contains the data used to update an Performance ReporterDB2 table.

subcomponentAn optionally installable part of an Performance Reporter feature component.

system tableA DB2 table that stores information for controlling log collector processing, Performance Reporter dialogs, andreporting.

T

targetIn an update definition, the DB2 table in which Performance Reporter stores data from the source record or table.

thresholdThe maximum or minimum acceptable level of usage. Usage measurements are compared with threshold levels.

Transmission Control Protocol/Internet Protocol (TCP/IP)A non-proprietary communications protocol for linking workstations to host computers and to other hardware.

U

update definitionInstructions for entering data into DB2 tables from records of different types or from other DB2 tables.

updatesInstructions in Performance Reporter on how to process data from log data sets to DB2 tables.


Glo

ssary

V

viewAn alternative representation of data from one or more tables. A view can include all or some of the columnscontained in the table on which it is defined.

206 Release 5

Index

Aabbreviations used in this book, list of 201accounting component

data flow 41reports 69tables 51

accounting component data tables 51XACCT_COMMAND_D 51XACCT_COMMAND_M 51XACCT_CONNECT_D 54XACCT_CONNECT_M 54XACCT_DISK_D 52XACCT_DISK_M 52XACCT_DISK_MV 55XACCT_PRINT_D 53XACCT_PRINT_M 53

accounting component reports 69UNIX Acct Cmd Resource Consumption, Daily

Overview 74UNIX Acct Commands by User, Daily Overview 70UNIX Acct Disk Blocks by User, Monthly Overview 77UNIX Acct Disk Blocks in 1000s, Monthly Trend 78UNIX Acct Printed Pages by System, Monthly

Overview 81UNIX Acct Printed Pages by User, Monthly Overview 80UNIX Acct User Resource Usage, Monthly Overview 76UNIX Acct Users and Connects, Daily Overview 79UNIX Acct Users by Command, Daily Overview 72

accounting function in UNIX, implementing 14agents 3ARM component data tables

ARM_TABLE_H,_D,_M 138ARM component lookup tables

X_NODE_NAME 130, 139ARM component reports

UNIX Application Monitoring Hourly Response TimeTrend 142

UNIX Application Monitoring Hourly Service LevelSummary 143

attributes, report 68audience for this book, intended xix

Bbooks, Performance Reporter xxii

Ccapturing UNIX data 14changes to this book xxiCOLLECT job (JCL) 29collect procedure, description 39, 164collecting performance data 3collection optimization, data 29communication between UNIX and MVS 10, 12, 18, 22component installation 24configuration component


configuration component data tables 56XCONFIG_HW 56XCONFIG_SW 57

configuration component reports 82UNIX Configuration of HW for a System, Overview 83UNIX Configuration of HW for Device Class,

Overview 85UNIX Configuration of SW for a System, Overview 86UNIX Configuration of SW for Object, Overview 88

connects, made by users 79control tables 39, 165CPU utilization 70, 72, 74, 76, 96, 97, 98, 100

Ddaily reports

DM Disk Resource Monitors, Daily (SEQ06) 182DM Memory Details, Daily (SEQ13) 196DM Network Monitors, Daily (SEQ05) 181DM Processor Monitors, Daily (SEQ13) 195DM Server Monitors, Daily (SEQ12) 194DM Service Level Details, Daily (SEQ02) 176DM Service Level Details, Daily (SEQ08) 186DM System Monitors, Daily (SEQ11) 193DM System Resources Monitors, Daily (SEQ04) 180DM TCP Monitors, Daily (SEQ10) 189UNIX Acct Cmd Resource Consumption, Daily Overview

(XACCT03) 74UNIX Acct Commands by User, Daily Overview

(XACCT01) 70UNIX Acct Users and Connects, Daily Overview

(XACCT07) 79UNIX Acct Users by Command, Daily Overview

(XACCT02) 72UNIX Error by Class, Daily Overview (XERROR03) 92UNIX Error by ID, Daily Overview (XERROR01) 90UNIX Error by Resource, Daily Overview

(XERROR04) 93UNIX Error by Type, Daily Overview (XERROR02) 91


Ind

ex

daily reports (continued)UNIX Perf CPU Utilization by System, Daily Overview

(XPERF02) 97UNIX Perf Disk I/O for System, Daily Overview

(XPERF07) 105UNIX Perf Statistics all Systems, Daily Overview

(XPERF04) 100UNIX Perf Vol Group and File Syst, Daily Overview

(XPERF05) 102data

sources of performance 3data collection, optimizing 29data flow from ARM to Performance Reporter 134data flow from distributed monitoring to Performance

Reporter 164data flow from UNIX to Performance Reporter

accounting component 41configuration component 43general flow 38job statistics component 45performance component 47

data tablesin the accounting component 51

XACCT_COMMAND_D 51XACCT_COMMAND_M 51XACCT_CONNECT_D 54XACCT_CONNECT_M 54XACCT_DISK_D 52XACCT_DISK_M 52XACCT_DISK_MV 55XACCT_PRINT_D 53XACCT_PRINT_M 53

in the ARM componentARM_TABLE_H,_D,_M 138

in the configuration component 56XCONFIG_HW 56XCONFIG_SW 57

in the distributed monitoring componentSENTRY_N_NT_H,_D,_W 168SENTRY_N_UN_H,_D,_W 169

in the error component 58XERROR_D 58XERROR_M 58

in the performance component 59XPERF_CPU_D 59XPERF_CPU_H 59XPERF_CPU_M 59XPERF_DISK_D 60XPERF_DISK_M 60XPERF_DISKIO_D 61XPERF_DISKIO_H 61XPERF_DISKIO_M 61XPERF_PAGING_D 62XPERF_PAGING_H 62XPERF_PAGING_M 62

naming standard 49, 137, 167summarization-level suffixes 49, 137, 167

disk blocksallocated to node 78allocated to user 77

disk spacefor installation 9utilization 103, 105

distributed monitoring component data tablesSENTRY_N_NT_H,_D,_W 168SENTRY_N_UN_H,_D,_W 169

distributed monitoring component lookup tablesSENTRY_MONITOR_NT 171SENTRY_MONITOR_UN 171X_NODE_NAME 159, 170

distributed monitoring reportsDM Disk Resource Monitors Daily (UNIX) 182DM Memory Details Daily (NT) 196DM Network Monitors Daily (UNIX) 181DM Processor Monitors Daily (NT) 195DM Server Monitors Daily (NT) 194DM Service Level Details Daily (NT) 186DM Service Level Details Daily (UNIX) 176DM Service Level Details Hourly (NT) 183DM Service Level Details Hourly (UNIX) 174DM Service Level Details Weekly (NT) 189DM Service Level Details Weekly (UNIX) 178DM System Monitors Daily (NT) 193DM System Resources Monitors Daily (UNIX) 180DM TCP Monitors Daily (NT) 192

documentation, Performance Reporter xxii

Eelapsed time 70, 72, 74, 76error component


error component data tables 58XERROR_D 58XERROR_M 58

error component reports 89UNIX Error by Class, Daily Overview 92UNIX Error by ID, Daily Overview 90UNIX Error by Resource, Daily Overview 93UNIX Error by Resource, Monthly Trend 94UNIX Error by Type, Daily Overview 91

errors, by classnumber 92percentage of total 92

errors, by IDnumber 90percentage of total 90

errors, by resourcemonthly trend 94number 93percentage of total 93

errors, by typenumber 91percentage of total 91

evaluation and planning 9

208 Release 5

Ffile transfer 18format, Performance Reporter report 67

Ggather of data 3Generation Data Group (GDG) 27, 28glossary 203groups, report 68

Hhardware configuration

for a device class 85for a system 83

hourly reportsDM Service Level Details Hourly (SEQ01) 174DM Service Level Details Hourly (SEQ07) 183UNIX Perf CPU Utilization by System, Hourly Trend

(XPERF01) 96UNIX Perf Disk I/O for a Disk, Hourly Trend

(XPERF06) 103UNIX Perf Page Space Utilization, Hourly Trend

(XPERF08) 106UNIX Perf Statistics by System, Hourly Trend

(XPERF03) 98how to use this book xix

II/O (kilobytes) 70, 72, 74, 76, 103, 105identifiers, report 67information logs 22installation

component 24distributed monitoring component 153requirements 9

making input data available 28optimizing data collection 29putting feature into production 30updating lookup tables 26, 159

testing the installation 30UNIX perfomance component 7

intended audience for this book xix

JJCL

COLLECT with specified Performance Reporter tables 29General COLLECT 29

Llibrary, Performance Reporter xxiilist of abbreviations used in this book 201list of terms used in this book 203log definitions, Performance Reporter 40logs

UNIX 40UNIX PerformanceUNIX Performance feature

information 22

Mmigrating from SP6000 feature 123monthly reports

UNIX Acct Disk Blocks by User, Monthly Overview(XACCT05) 77

UNIX Acct Disk Blocks in 1000s, Monthly Trend(XACCT06) 78

UNIX Acct Printed Pages by System, Monthly Overview(XACCT09) 81

UNIX Acct Printed Pages by User, Monthly Overview(XACCT08) 80

UNIX Acct User Resource Usage, Monthly Overview(XACCT04) 76

UNIX Error by Resource, Monthly Trend (XERROR05) 94

Nnaming standards for tables 49, 167

Ooperating UNIX PerformanceUNIX Performance feature on

UNIX 22optimizing data collection 29overview reports

UNIX Acct Cmd Resource Consumption, Daily Overview(XACCT03) 74

UNIX Acct Commands by User, Daily Overview(XACCT01) 70

UNIX Acct Disk Blocks by User, Monthly Overview(XACCT05) 77

UNIX Acct Printed Pages by System, Monthly Overview(XACCT09) 81

UNIX Acct Printed Pages by User, Monthly Overview(XACCT08) 80

UNIX Acct User Resource Usage, Monthly Overview(XACCT04) 76

UNIX Acct Users and Connects, Daily Overview(XACCT07) 79

UNIX Acct Users by Command, Daily Overview(XACCT02) 72

UNIX Configuration of HW for a System, Overview(XCONFIG01) 83


Ind

ex

overview reports (continued)UNIX Configuration of HW for Device Class, Overview

(XCONFIG02) 85UNIX Configuration of SW for a System, Overview

(XCONFIG03) 86UNIX Configuration of SW for Object, Overview

(XCONFIG04) 88UNIX Error by Class, Daily Overview (XERROR03) 92UNIX Error by ID, Daily Overview (XERROR01) 90UNIX Error by Resource, Daily Overview

(XERROR04) 93UNIX Error by Type, Daily Overview (XERROR02) 91UNIX Perf CPU Utilization by System, Daily Overview

(XPERF02) 97UNIX Perf Disk I/O for System, Daily Overview

(XPERF07) 105UNIX Perf Statistics all Systems, Daily Overview

(XPERF04) 100UNIX Perf Vol Group and File Syst, Daily Overview

(XPERF05) 102

Ppage space, utilization 98, 100, 106paging rate 98, 100performance component


performance component data tables 59XPERF_CPU_D 59XPERF_CPU_H 59XPERF_CPU_M 59XPERF_DISK_D 60XPERF_DISK_M 60XPERF_DISKIO_D 61XPERF_DISKIO_H 61XPERF_DISKIO_M 61XPERF_PAGING_D 62XPERF_PAGING_H 62XPERF_PAGING_M 62

performance component reports 95UNIX Perf CPU Utilization by System, Daily Overview 97UNIX Perf CPU Utilization by System, Hourly Trend 96UNIX Perf Disk I/O for a Disk, Hourly Trend 103UNIX Perf Disk I/O for System, Daily Overview 105UNIX Perf Page Space Utilization, Hourly Trend 106UNIX Perf Statistics all Systems, Daily Overview 100UNIX Perf Statistics by System, Hourly Trend 98UNIX Perf Vol Group and File Syst, Daily Overview 102

performance managementcollecting data (overview) 3concepts xxiii

Performance Reporter product library xxiiplanning the implementation process 7print resources used

pages, by user 80, 81printouts, by user 81

processes, number of 98, 100

product library, Performance Reporter xxii

Rrecord definition descriptions

ACCT_COMMAND 110ACCT_CONNECT 111ACCT_DISK 112ACCT_PRINT 113CONFIG_HW 114CONFIG_SW 115ERROR 117PERF_CPU 118PERF_DISK_FS 119PERF_DISK_IO 121PERF_DISK_VG 120PERF_PAGING 122

record definitions for Performance Reporter 40report format 67report groups 4report IDs 67

XACCT01 (UNIX Acct Commands by User, DailyOverview) 70

XACCT02 (UNIX Acct Users by Command, DailyOverview) 72

XACCT03 (UNIX Acct Cmd Resource Consumption, DailyOverview) 74

XACCT04 (UNIX Acct User Resource Usage, MonthlyOverview) 76

XACCT05 (UNIX Acct Disk Blocks by User, MonthlyOverview) 77

XACCT06 (UNIX Acct Disk Blocks in 1000s, MonthlyTrend) 78

XACCT07 (UNIX Acct Users and Connects, DailyOverview) 79

XACCT08 (UNIX Acct Printed Pages by User, MonthlyOverview) 80

XACCT09 (UNIX Acct Printed Pages by System, MonthlyOverview) 81

XCONFIG01 (UNIX Configuration of HW for a System,Overview) 83

XCONFIG02 (UNIX Configuration of HW for Device Class,Overview) 85

XCONFIG03 (UNIX Configuration of SW for a System,Overview) 86

XCONFIG04 (UNIX Configuration of SW for Object,Overview) 88

XERROR01 (UNIX Error by ID, Daily Overview) 90XERROR02 (UNIX Error by Type, Daily Overview) 91XERROR03 (UNIX Error by Class, Daily Overview) 92XERROR04 (UNIX Error by Resource, Daily

Overview) 93XERROR05 (UNIX Error by Resource, Monthly Trend) 94XPERF01 (UNIX Perf CPU Utilization by System, Hourly

Trend) 96XPERF02 (UNIX Perf CPU Utilization by System, Daily

Overview) 97XPERF03 (UNIX Perf Statistics by System, Hourly

Trend) 98

210 Release 5

report IDs 67 (continued)XPERF04 (UNIX Perf Statistics all Systems, Daily

Overview) 100XPERF05 (UNIX Perf Vol Group and File Syst, Daily

Overview) 102XPERF06 (UNIX Perf Disk I/O for a Disk, Hourly

Trend) 103XPERF07 (UNIX Perf Disk I/O for System, Daily

Overview) 105XPERF08 (UNIX Perf Page Space Utilization, Hourly

Trend) 106reporting dialogs xxiireports

attributes 68format and general description 67identifiers, unique 67in the accounting component 69

UNIX Acct Cmd Resource Consumption, DailyOverview 74

UNIX Acct Commands by User, Daily Overview 70UNIX Acct Disk Blocks by User, Monthly

Overview 77UNIX Acct Disk Blocks in 1000s, Monthly Trend 78UNIX Acct Printed Pages by System, Monthly

Overview 81UNIX Acct Printed Pages by User, Monthly

Overview 80UNIX Acct User Resource Usage, Monthly

Overview 76UNIX Acct Users and Connects, Daily Overview 79UNIX Acct Users by Command, Daily Overview 72

in the ARM componentUNIX Application Monitoring Hourly Response

Times 142UNIX Application Monitoring Hourly Service Level

Summary 143in the configuration component 82

UNIX Configuration of HW for a System, Overview 83UNIX Configuration of HW for Device Class,

Overview 85UNIX Configuration of SW for a System, Overview 86UNIX Configuration of SW for Object, Overview 88

in the distributed monitoring componentDM Disk Resource Monitors Daily (UNIX) 182DM Memory Details Daily (NT) 196DM Network Monitors Daily (UNIX) 181DM Processor Monitors Daily (NT) 195DM Server Monitors Daily (NT) 194DM Service Level Details Daily (NT) 186DM Service Level Details Daily (UNIX) 176DM Service Level Details Hourly (NT) 183DM Service Level Details Hourly (UNIX) 174DM Service Level Details Weekly (NT) 189DM Service Level Details Weekly (UNIX) 178DM System Monitors Daily (NT) 193DM System Resources Monitors Daily (UNIX) 180DM TCP Monitors Daily (NT) 192

in the error component 89UNIX Error by Class, Daily Overview 92UNIX Error by ID, Daily Overview 90UNIX Error by Resource, Daily Overview 93

reports (continued)in the error component 89 (continued)

UNIX Error by Resource, Monthly Trend 94UNIX Error by Type, Daily Overview 91

in the performance component 68, 95UNIX Perf CPU Utilization by System, Daily

Overview 97UNIX Perf CPU Utilization by System, Hourly

Trend 96UNIX Perf Disk I/O for a Disk, Hourly Trend 103UNIX Perf Disk I/O for System, Daily Overview 105UNIX Perf Page Space Utilization, Hourly Trend 106UNIX Perf Statistics all Systems, Daily Overview 100UNIX Perf Statistics by System, Hourly Trend 98UNIX Perf Vol Group and File Syst, Daily

Overview 102source tables 68variables 68

requirements, installation 9disk space 9software 10UNIX 9

retry interval, file transfer 19retry limit, file transfer 18

Ssoftware configuration

for a system 86for an object 88

source tables, report 68SP6000 feature, migrating from 123space utilization 102suffixes, data-table 49, 167system time 70, 72, 74, 76

Ttables

ARM_TABLE_D 138ARM_TABLE_H 138ARM_TABLE_M 138lookup 63SENTRY_N_NT_D 168SENTRY_N_NT_H 168SENTRY_N_NT_W 168SENTRY_N_UN_D 169SENTRY_N_UN_H 169SENTRY_N_UN_W 169X_NODE_NAME 27, 64, 159, 170XACCT_COMMAND_D 26, 51, 159XACCT_COMMAND_M 51XACCT_CONNECT_D 54XACCT_CONNECT_M 54XACCT_CONNECT_TYPE 27, 63XACCT_DISK_D 52XACCT_DISK_M 52


Ind

ex

tables (continued)XACCT_DISK_MV 55XACCT_PRINT_D 53XACCT_PRINT_M 53XACCT_PRINT_TYPE 27, 65XCONFIG_HW 56XCONFIG_SW 57XERROR_D 58XERROR_M 58XPERF_CPU_D 59XPERF_CPU_H 59XPERF_CPU_M 59XPERF_DISK_D 60XPERF_DISK_M 60XPERF_DISKIO_D 61XPERF_DISKIO_H 61XPERF_DISKIO_M 61XPERF_PAGING_D 62XPERF_PAGING_H 62XPERF_PAGING_M 62

TCP/IP 4terms used in this book 203testing the installation 30transfer of files to MVS 14trend reports

UNIX Acct Disk Blocks in 1000s, Monthly Trend(XACCT06) 78

UNIX Error by Resource, Monthly Trend (XERROR05) 94UNIX Perf CPU Utilization by System, Hourly Trend

(XPERF01) 96UNIX Perf Disk I/O for a Disk, Hourly Trend

(XPERF06) 103UNIX Perf Page Space Utilization, Hourly Trend

(XPERF08) 106UNIX Perf Statistics by System, Hourly Trend

(XPERF03) 98

UUNIX lookup tables 63

X_NODE_NAME 27, 64XACCT_CONNECT_TYPE 27, 63XACCT_PRINT_TYPE 27, 65

UNIX Performanceinstallation of UNIX PerformanceUNIX Performance

feature 9updating lookup tables

for distributed monitoring component 159for UNIX component 26

use of this book, intended xixuser time 70, 72, 74, 76

Vvariables, report 68view table for accounting component 55

XX_NODE_NAME lookup table 27, 139, 159, 170XACCT_CONNECT_TYPE lookup table 27XACCT_PRINT_TYPE lookup table 27

212 Release 5

Program Number: 5695-101

Printed in Denmark by IBM Danmark A/S

SH19-4018-04

Date post:	27-May-2020
Category:	Documents
Upload:	others
View:	13 times
Download:	0 times

Performance Reporter OS/390 Distributed Systems...

Documents