Teradata

Teradata Database
Resource Usage Macros and Tables

Release 14.10
B035-1099-112A
August 2014
The product or products described in this book are licensed products of Teradata Corporation or its affiliates.
Teradata, Active Data Warehousing, Active Enterprise Intelligence, Applications-Within, Aprimo Marketing Studio, Aster, BYNET, Claraview,
DecisionCast, Gridscale, MyCommerce, QueryGrid, SQL-MapReduce, Teradata Decision Experts, "Teradata Labs" logo, Teradata ServiceConnect,
Teradata Source Experts, WebAnalyst, and Xkoto are trademarks or registered trademarks of Teradata Corporation or its affiliates in the United
States and other countries.
Adaptec and SCSISelect are trademarks or registered trademarks of Adaptec, Inc.
AMD Opteron and Opteron are trademarks of Advanced Micro Devices, Inc.
Apache, Apache Hadoop, Hadoop, and the yellow elephant logo are either registered trademarks or trademarks of the Apache Software Foundation
in the United States and/or other countries.
Apple, Mac, and OS X all are registered trademarks of Apple Inc.
Axeda is a registered trademark of Axeda Corporation. Axeda Agents, Axeda Applications, Axeda Policy Manager, Axeda Enterprise, Axeda Access,
Axeda Software Management, Axeda Service, Axeda ServiceLink, and Firewall-Friendly are trademarks and Maximum Results and Maximum
Support are servicemarks of Axeda Corporation.
Data Domain, EMC, PowerPath, SRDF, and Symmetrix are registered trademarks of EMC Corporation.
GoldenGate is a trademark of Oracle.
Hewlett-Packard and HP are registered trademarks of Hewlett-Packard Company.
Hortonworks, the Hortonworks logo and other Hortonworks trademarks are trademarks of Hortonworks Inc. in the United States and other
countries.
Intel, Pentium, and XEON are registered trademarks of Intel Corporation.
IBM, CICS, RACF, Tivoli, and z/OS are registered trademarks of International Business Machines Corporation.
Linux is a registered trademark of Linus Torvalds.
LSI is a registered trademark of LSI Corporation.
Microsoft, Active Directory, Windows, Windows NT, and Windows Server are registered trademarks of Microsoft Corporation in the United States
and other countries.
NetVault is a trademark or registered trademark of Dell Inc. in the United States and/or other countries.
Novell and SUSE are registered trademarks of Novell, Inc., in the United States and other countries.
Oracle, Java, and Solaris are registered trademarks of Oracle and/or its affiliates.
QLogic and SANbox are trademarks or registered trademarks of QLogic Corporation.
Quantum and the Quantum logo are trademarks of Quantum Corporation, registered in the U.S.A. and other countries.
Red Hat is a trademark of Red Hat, Inc., registered in the U.S. and other countries. Used under license.
SAP is the trademark or registered trademark of SAP AG in Germany and in several other countries.
SAS and SAS/C are trademarks or registered trademarks of SAS Institute Inc.
SPARC is a registered trademark of SPARC International, Inc.
Symantec, NetBackup, and VERITAS are trademarks or registered trademarks of Symantec Corporation or its affiliates in the United States and
other countries.
Unicode is a registered trademark of Unicode, Inc. in the United States and other countries.
UNIX is a registered trademark of The Open Group in the United States and other countries.
Other product and company names mentioned herein may be the trademarks of their respective owners.
THE INFORMATION CONTAINED IN THIS DOCUMENT IS PROVIDED ON AN "AS-IS" BASIS, WITHOUT WARRANTY OF ANY KIND, EITHER
EXPRESS OR IMPLIED, INCLUDING THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR
NON-INFRINGEMENT. SOME JURISDICTIONS DO NOT ALLOW THE EXCLUSION OF IMPLIED WARRANTIES, SO THE ABOVE EXCLUSION
MAY NOT APPLY TO YOU. IN NO EVENT WILL TERADATA CORPORATION BE LIABLE FOR ANY INDIRECT, DIRECT, SPECIAL, INCIDENTAL,
OR CONSEQUENTIAL DAMAGES, INCLUDING LOST PROFITS OR LOST SAVINGS, EVEN IF EXPRESSLY ADVISED OF THE POSSIBILITY OF
SUCH DAMAGES.
The information contained in this document may contain references or cross-references to features, functions, products, or services that are not
announced or available in your country. Such references do not imply that Teradata Corporation intends to announce such features, functions,
products, or services in your country. Please consult your local Teradata Corporation representative for those features, functions, products, or
services available in your country.
Information contained in this document may contain technical inaccuracies or typographical errors. Information may be changed or updated
without notice. Teradata Corporation may also make improvements or changes in the products or services described in this information at any time
without notice.
To maintain the quality of our products and services, we would like your comments on the accuracy, clarity, organization, and value of this document.
Please email: teradata-books@lists.teradata.com.
Any comments or materials (collectively referred to as "Feedback") sent to Teradata Corporation will be deemed non-confidential. Teradata
Corporation will have no obligation of any kind with respect to Feedback and will be free to use, reproduce, disclose, exhibit, display, transform,
create derivative works of, and distribute the Feedback and derivative works thereof without limitation on a royalty-free basis. Further, Teradata
Corporation will be free to use any ideas, concepts, know-how, or techniques contained in such Feedback for any purpose whatsoever, including
developing, manufacturing, or marketing products or services incorporating Feedback.
Copyright 2000-2014 by Teradata. All Rights Reserved.
Preface
Purpose
This is a reference book for Teradata Database resource usage macros and tables. The macros
can be used to report resource usage data and, when enabled, the tables can be used to:
Observe the balance of disk and CPU usage
Obtain pdisk usage information
Check for bottlenecks
Provide an overall history of the system operation
Monitor the utilization of AMP Worker Tasks (AWTs)
Related Documentation
For information on:
Collecting RSS data, see Application Programming Reference.
Workload management (WM) capacity on demand (COD) and setting limits on resource
usage, refer to the following:
Carrie Ballinger, Workload Management Capacity on Demand: Teradata Database 14.10

and Linux SLES 11, Teradata Database Orange Book 541-0010245
Teradata Viewpoint User Guide
Audience
This book is intended to be used by administrators, programmers, and other Teradata
technical personnel responsible for administering or managing Teradata Database.
Supported Software Releases and Operating

Systems
This book supports Teradata Database 14.10.
Teradata Database 14.10 is supported on:
SUSE Linux Enterprise Server (SLES)10
SLES 11
Teradata Database client applications support other operating systems.
Preface
Changes to This Book
Changes to This Book

Release
Description
Teradata Database 14.10.03
Added a new feature: Workload Management Capacity on Demand.

This feature is available starting with TD 14.10.03.
August 2014
Modified the following field descriptions:

CODFactor
WM CPU COD, WM I/IO COD, and PM I/O COD Spare
columns
LdvReadRespMax
LdvWriteRespMax
NetMrgHeapFails
NetMrgHeapRequests
Added the ProcBlksTime column to the ResUsageSps Table
chapter.
Replaced instances of CASE statement with CASE expression.
Teradata Database 14.10
September 2013

May 2013
Updated the following:

The Data Block Prefetch columns.
The Segment Acquired columns.
The VprType field description in the ResUsageSps table chapter.
Added a note to the AGid, RelWgt, and IOBlks columns about how
they are only valid on SLES 10 or earlier systems.
Added a note to the VPid, WDid, WaitIO, CPURunDelay,
IOSubmitted, IOCompleted, IOCriticalSubmitted, IOSubmittedKB,
IOCompletedKB, IOCriticalSubmittedKB, DecayLevel1IO,
DecayLevel2IO, DecayLevel1CPU, DecayLevel2CPU,
TacticalExceptionIO, and TacticalExceptionCPU columns about
how they are only valid on SLES 11 or later systems.
Replaced references to ResGeneralInfoView with ResSpmaView.
Added the following:
Spare02-11 fields to the ResUsageSpma table
Spare00-04 fields to the ResUsageSldv table
Spare00-01 and Spare03-14 fields to the ResUsageSps table
Spare00-01 fields to the ResUsageSpdsk and the ResUsageSvdsk
tables
Spare00-07 fields to the ResUsageSvpr table
Spare00 field to the ResUsageSawt, ResUsageScpu, ResUsageShst,
ResUsageImpa, ResUsageIvpr tables
SpareInt field to the ResUsageSpma, ResUsageSldv,
ResUsageSpdsk, ResUsageSps, and ResUsageSvdsk tables
An entry and description for VH cache to the Glossary
Preface
Additional Information
Release
Description
Updated the following views:

ResSpmaView
ResSldvView
ResSpdskView
ResSpsView
ResSpdskView
ResSvdskView
ResSvprView
ResSpmaView
ResIvprView
Updated the following column descriptions:
CODFactor
LdvReadRespMax
LdvWriteRespMax
AwtInuseMax
Noted that the message delivery times columns for the
ResUsageIpma and ResUsageIvpr tables are not currently valid.
Replaced instances of LAN-connected with TCP/IP networkconnected.
May 2013 continued
Additional Information
URL
Description
www.info.teradata.com/
Use the Teradata Information Products Publishing Library site

to:
View or download a manual:
1 Under Online Publications, select General Search.
2 Enter your search criteria and click Search.
Download a documentation CD-ROM:

1 Under Online Publications, select General Search.
2 In the Title or Keyword field, enter CD-ROM, and click
Search.
www.teradata.com
The Teradata home page provides links to numerous sources of

information about Teradata. Links include:
Executive reports, white papers, case studies of customer
experiences with Teradata, and thought leadership
Technical information, solutions, and expert advice
Press releases, mentions and media resources
Preface
Teradata Database Optional Features
URL
Description
www.teradata.com/t/TEN/
Teradata Customer Education delivers training that builds skills

and capabilities for our customers, enabling them to maximize
their Teradata investment.
tays.teradata.com/
Use Teradata @ Your Service to access Orange Books, technical

alerts, and knowledge repositories, view and join forums, and
download software patches.
developer.teradata.com/
Teradata Developer Exchange provides articles on using

Teradata products, technical discussion forums, and code
downloads.
To maintain the quality of our products and services, we would like your comments on the
accuracy, clarity, organization, and value of this document. Please email teradatabooks@lists.teradata.com.
Teradata Database Optional Features

This book may include descriptions of the following optional Teradata Database features and
products:
Teradata Columnar
Teradata Row Level Security
Teradata SQL-H
Teradata Temporal
Teradata Virtual Storage (VS)
Unity Source Link
You may not use these features without the appropriate licenses. The fact that these features
may be included in product media or downloads, or described in documentation that you
receive, does not authorize you to use them without the appropriate licenses.
Contact your Teradata sales representative to purchase and enable optional features.
Table of Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Audience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Supported Software Releases and Operating Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Changes to This Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
Additional Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
Teradata Database Optional Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
Chapter 1: Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Benefits of Resource Usage Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Overview of Resource Usage Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Data Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Resource Usage Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Application Programming Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Chapter 2: Planning Your Resource Usage Data . . . . . . . . . . . . . . . 19

Resource Usage Table Settings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Tables Based on Needed Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Types of Resource Usage Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Logging Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Logging Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Active Row Filter Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Resource Usage Logging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Cost of Logging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Logging Cost Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Operational Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
23
23
24
24
Table of Contents
Chapter 3: Resource Usage and Procedures . . . . . . . . . . . . . . . . . . . .27

Enabling RSS Logging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
ctl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
DBW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
General Macro Input Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
Parameter Use for One-Node, Multiple-Node, All-Node, and Group Macros . . . . . . . . .30
Input Format for One-Node Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Input Format for ByGroup Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Retaining Data From Prior Releases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
Macro Execution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
Using ENABLE and DISABLE LOGON Commands. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
Purging Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35
Chapter 4: Resource Usage Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37

Physical Table Naming Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37
Relational Primary Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37
Resource Usage Table Rows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38
Occasional Event Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38
Types of Resource Usage Table Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38
Invalid Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41
About the Mode Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42
Summary Mode in Resource Usage Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42
Chapter 5: ResUsageScpu Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45

Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45
Relational Primary Index Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45
Miscellaneous Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45
Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46
Process Scheduling CPU Utilization Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46
Reserved Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48
Table of Contents
Chapter 6: ResUsageSpma Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Relational Primary Index Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Miscellaneous Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
File System Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Concurrency Control Database Locks Columns . . . . . . . . . . . . . . . . . . . . . . . . .
Host Controller Channel and TCP/IP Network Traffic Columns . . . . . . . . . . . . . . . . . .
Memory Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Net Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Process Scheduling Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
TASM Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Teradata VS Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
User Command Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reserved Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
52
52
55
55
56
57
60
64
65
65
66
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Chapter 7: ResUsageSawt Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
TASM Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Reserved Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Chapter 8: ResUsageShst Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Host Controller Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Reserved Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Table of Contents
Chapter 9: ResUsageSldv Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85

Logical Device Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87
Reserved Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
Chapter 10: ResUsageSpdsk Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93

Teradata VS Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95
Reserved Column. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
Chapter 11: ResUsageSps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101

Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101
Relational Primary Index Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101
Miscellaneous Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102
Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .104
TASM Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .104
Process Scheduling Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109
File System Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112
Memory Allocation Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116
Net Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116
Reserved Column. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117
10
Table of Contents
Chapter 12: ResUsageSvdsk Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Relational Primary Index Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Miscellaneous Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Teradata VS Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Reserved Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Chapter 13: ResUsageSvpr Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
File System Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Concurrency Control Database Locks Columns . . . . . . . . . . . . . . . . . . . . . . . .
Memory Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Net Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Process Scheduling Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Teradata VS Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reserved Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
133
133
146
146
150
151
155
158
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Chapter 14: Resource Usage Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

ResCPUUsageByAMPView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
ResCPUUsageByPEView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
ResSawtView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
ResScpuView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
ResShstView. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
ResSldvView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
ResSpdskView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
ResSpmaView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
ResSpsView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
11
Table of Contents
ResSvdskView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178
ResSvprView . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .179
Chapter 15: Resource Usage Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . .181

Macro Output Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .181
ResAWT Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .183
ResCPUByAMP Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .188
ResCPUByPE Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192
ResCPUByNode Macros. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .195
ResHostByLink Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .198
ResLdvByNode Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .200
ResMemMgmtByNode Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .204
ResNetByNode Macros. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .208
ResNode Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211
ResPdskByNode Macros. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .217
ResPs Macros. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220
ResVdskByNode Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .223
Appendix A: How to Read Syntax Diagrams . . . . . . . . . . . . . . . . . . .227

Syntax Diagram Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .227
Appendix B: ResUsageIpma Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .233

Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .233
Relational Primary Index Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .233
Miscellaneous Housekeeping Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .233
Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .235
Process Scheduling CPU Switching Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .235
Net Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .236
Reserved Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .239
Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .240
12
Table of Contents
Appendix C: ResUsageIvpr Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241

Statistics Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
File System Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Concurrency Control Monitor Management Columns. . . . . . . . . . . . . . . . . . .
Net Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Transient Journal Purge Overhead Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
243
243
248
249
250
Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

Spare Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Appendix D: ResIpmaView and ResIvprView Views . . . . . . . . 253

ResIpmaView. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
ResIvprView. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Appendix E: Partition Assignments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

Table Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Partition Assignment Listing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
13
Table of Contents
14
CHAPTER 1
Introduction
This manual documents the resource usage data and settings for a variety of installation
configurations and environments in Chapter 2: Planning Your Resource Usage Data. To
implement the settings you decide on, see Chapter 3: Resource Usage and Procedures.
The only maintenance required is to purge old data regularly. See Purging Data.
For additional information on performance analysis and system tuning, see the following:
Application Programming Reference
Database Administration
Benefits of Resource Usage Data

Resource usage data is useful for the following purposes:
Measuring system performance
Measuring component performance
Assisting with on-site job scheduling
Identifying potential performance impacts
Planning installation, upgrade, and migration
Analyzing performance degradation and improvement
Identifying problems such as bottlenecks, parallel inefficiencies, down components, and

congestion
Overview of Resource Usage Data

Resource usage data is stored in system tables and views in the DBC database. Macros installed
with Teradata Database generate reports that display the data.
To load the resource usage views and macros, you can run the Database Initialization Program
(dip) script, DIPRUM. For more information on DIPRUM, see Utilities.
Note: You can run DIP scripts one at a time or all at once by running DIPALL. DIP scripts,
such as DIPRUM and DIPSYSFNC, can be executed in any order.
As with other database data, you can access resource usage data using SQL if you have the
proper privileges. You can also write your own queries or macros on resource usage data.
The following table lists the topics covered by resource usage data.
15
Chapter 1: Introduction
Data Reporting
Resource usage data covers
Which includes
BYNET traffic on a node
point-to-point messaging, broadcast messaging, and merge

activities.
client-to-server traffic
data for each communication link.
CPU utilization
overhead, user service, and time of session execution.
storage device traffic
the number of reads/writes and amount of data transferred

as seen from the storage driver.
pdisk device traffic
pdisk I/O, cylinder allocation, and migration statistics.
vdisk device traffic
all the cylinders allocated by an AMP (which can come from

any pdisks in the clique).
Priority Scheduler information
On SLES 10 or earlier systems, data by Performance

Group (PG) from the Priority Scheduler and the ability
to report resource usage data by Teradata Active System
Management (TASM) workload definitions (WDs).
On SLES 11 or later systems, data by Priority Scheduler
WD.
Note: Priority Scheduler is managed by TASM on SLES
10 systems, and is configured using the Teradata
Viewpoint workload management portlets. For more
information on those portlets, see Teradata Viewpoint
User Guide.
AMP Worker Task information
AMP Worker Task statistics.
memory management activity
memory allocation.
Data Reporting
Data is reported at the logging period. When a new logging period starts, the data is gathered
in the Gather Buffer, then updated to the Log Buffer and logged to the database resource usage
tables.
16
Resource Usage Macros
Data Collection Macros

and
Routines
Gather Buffer
Log Buffer
ResUsage Write Queue
ResUsage
Tables
ResUsage Reports
1099B001

Resource usage macros produce reports from data logged to the resource usage tables. They
can generate reports for a selected period of time and nodes.
You can use the reports to analyze key operational statistics and evaluate the performance of
your system.
Like other macros, resource usage macros consist of one or more Teradata SQL statements
stored in Teradata Database and executed by a single EXECUTE statement.
Refer to Chapter 3: Resource Usage and Procedures for more information on the resource
usage macros, and SQL Quick Reference for details about how to use the EXECUTE statement.
17
Application Programming Interfaces
Application Programming Interfaces

You can use the Teradata application programming interfaces (APIs) to:
Set the resource monitoring and logging rates
Collect RSS data and return node, vproc, and WD oriented data
Examples of these APIs, include the System PMPC SET RESOURCE RATE, MONITOR
PHYSICAL RESOURCE, MONITOR VIRTUAL RESOURCE, and MONITOR WD requests.
For more information on these APIs, see Application Programming Reference.
18
CHAPTER 2
Planning Your Resource Usage

Data
Resource Usage Table Settings

The default resource usage settings provide a good starting point for system monitoring.The
default results in the ResUsageSpma (SPMA) table being logged every 10 minutes (600
seconds).
The ResUsageSpma table provides a high level summary of how the system is operating and
contains summarized or key elements from most of the other tables. If you want to record
detailed statistics covered by any of the resource usage tables, you should enable them for
logging, along with specifying the largest logging period that will meet your needs. You should
not log data that you do not have a planned need for since this does incur additional database
system overhead and uses up additional database space.
The more tables you enable for logging and the shorter the logging period used, the more
overhead the system will use.
Tables Based on Needed Reports

If you plan on using the report macros provided in Chapter 15: Resource Usage Macros, you
need to enable the associated table.
Related Topics
For...
See...
instructions on setting resource usage

tables
Enabling RSS Logging on page 27.
instructions on using macros
General Macro Input Format on page 29 and

Macro Execution on page 31.
descriptions and examples of the macros
Chapter 15: Resource Usage Macros.
19
Chapter 2: Planning Your Resource Usage Data

Types of Resource Usage Tables
Types of Resource Usage Tables

The following table describes the tables and provides guidance about which ones to enable.
Note: The shaded rows in the table below are the resource usage tables that are generally not
used at customer sites.
Table Name
Covers
When You Should Enable
ResUsageIpma
System-wide node information, intended

primarily for Teradata engineers.
---
ResUsageIvpr
System-wide virtual processor information,

intended primarily for Teradata engineers.
---
ResUsageSawt
Data specific to the AMP Worker Tasks.
When you want to monitor the utilization of the

AMP Worker Task and determine if work is backing
up because the AMP Worker Tasks are all being used.
ResUsageScpu
Statistics on the CPUs within the nodes.
When the performance analysis suggests that the

overall performance is limited or to check if a
program is spinning in an infinite loop on an
individual processor.
For example, saturation of a particular CPU on each
node or on a particular node while others are idle
could indicate a task always uses that CPU.
Also, you should enable when the system is first
brought online to verify the following:
That all CPUs are functioning on all nodes
There is a good load balance among the CPUs
ResUsageShst
Statistics on the host channels and LANs

that communicate with Teradata Database.
To determine details about the traffic over the IBM

Host channels and if there is a bottleneck.
ResUsageSldv
System-wide, logical device statistics

collected from the storage driver.
To observe the balance of disk usage. The storage

device statistics are often difficult to interpret with
disk arrays attached due to multi-path access to disks.
Note: Use the ResUsageSvdsk table first to observe
general system disk utilization unless specifically
debugging at a low level.
ResUsageSpdsk
20
Statistics collected from the pdisk device.
To obtain detailed usage information about pdisks.

Logging Rate
Table Name
Covers
When You Should Enable
ResUsageSpma
System-wide node information provides a

summary of overall system utilization
incorporating the essential information
from most of the other tables.
To provide an overall history of the system operation.
Use the columns in ResUsageSpma to view

BYNET utilization.
Note: The BYNET can transmit and receive
at the same time, resulting in 100%
transmitting and 100% receiving values
simultaneously.
Another method of determining BYNET
utilization and traffic is to use the blmstat
tool.
ResUsageSps
Statistics based on the WD the work is being

performed for.
When you need to track utilization by the WD level.
ResUsageSvpr
Data specific to each virtual processor and

its file system.
To view details about the resources being used by each

vproc on the system. This table is useful for looking
for hot AMPS or PEs that may be CPU bound or
throttled on other resources.
Logging Rate
The logging rate controls the frequency (number of seconds) at which resource usage data is
logged to the resource usage tables.
When you have decided what rate to set, see Chapter 3: Resource Usage and Procedures for
details on how to set the logging rate.
Logging Period
Resource usage logging means the writing of resource data as rows to one or more of the
resource usage database tables. The tables are named DBC.ResUsagexxxx, where xxxx is the
name of the resource usage table (for example, SPMA, IPMA, and so on) as listed in Types of
Resource Usage Tables on page 20.
The shorter the logging period, the more frequently data is logged, and the more disk space is
used.
When the system is so busy that the resource usage table logging gets backed up, RSS will
automatically double the logging period which effectively summarizes the data by providing
values for a time period twice that provided by the previous logging period.
If you see the resource usage logging rates change without user intervention, this means that
the database is busy. When no longer busy, the system resumes logging as before.
21

Summary Mode
Note: Events in the event logs related to this doubling of the logging period do not represent
fatal errors but are informational to indicate that the automatic operations of the RSS are
attempting to maintain data logging.
Rule
The following rule on the logging rate is imposed by the system.
Intervals must evenly divide into 3600 (the number of seconds in an hour). The following
table shows the valid logging rate.
The white area of the table shows rates recommended only for short-term use for
debugging a specific issue.
The highlighted area of the table shows rates recommended for production processing.
1
10
12
15
16
18
20
24
25
30
36
40
45
48
50
60
72
75
80
90
100
120
144
150
180
200
225
240
300
360
400
450
600
720
900
1200
1800
3600
A practical minimum log interval during production processing is 60 seconds. Intermediate

log intervals, such as 120 seconds or 300 seconds can also be used. The default rate is 600
seconds.
Rates and enabled tables may be changed at any time and the changes take effect immediately.
Summary Mode
You can use Summary Mode to reduce the system overhead from logging tables that produce
multiple rows per logging period. Summary Mode helps reduce overhead by combining data
from multiple rows into one or more summary rows based on specific criteria for each table.
For example, if you want to log information provided in the ResUsageSvpr table but do not
need data for each individual vproc, use Summary Mode to produce one row per vproc type
instead of one row per vproc.
The ResUsageSpma table, in comparison, provides node level summary of key fields from
most of the other resource usage tables. When more details are required than the
ResUsageSpma table provides then the next level of information is provided by using
Summary Mode logging for the table of interest. This helps minimize the cost of the data
logging.
22

Active Row Filter Mode
You can select Summary Mode for each table individually. For details on how Summary Mode
affects that particular table, see the description for each table.
For example, for the ResUsageSvpr table in Summary Mode, all the individual vproc rows of
the same vproc type are combined into a single row. Since the data values are added together,
you need to divide the summary row data value by the number of rows that made up the
Summary Mode row to get the average per vproc. For example, divide the AMP summary row
data value by the number of AMPs on that node to determine the average value per AMP. A
similar computation needs to be done to derive the average value per PE from the summary
row data value.
Note: To determine the number of AMP, PEs, and all other vproc types on your system, you
can use the ResUsageSpma table or use the Vproc Manager utility.
Resource usage columns that represent a maximum statistic are not summed together. Instead
the maximum value from the rows is used. For example, the ResUsageSvpr table
MsgWorkQLenMax column in the Summary Mode row for the AMPs will contain the
maximum value from all the AMP rows that would have been logged in non-Summary Mode.
The columns that represent a minimum statistic are summarized by storing the minimum
value from all the constituent rows.
Summary Mode has either no effect on the values of the Housekeeping Columns or it is
specifically detailed in the description of each affected field.
To enable Summary Mode, see Enabling RSS Logging on page 27.
For more information on Summary Mode, see Summary Mode in Resource Usage Tables on
page 42.
Active Row Filter Mode

Active Row Filter Mode reduces the overhead of logging for some of the resource usage table
by limiting the data rows that are logged to the database.
When you enable active row filtering, it may appear that rows are missing when looking at the
query results. This is because the index values of the inactive rows varies over time so that a
row with one index may be logged one period but not in another. To determine if rows are not
being logged to the database, you should look in the event logs for messages indicating that
rows have been lost.
Active Row Filtering should not be disabled for the ResUsageSps table.
Resource Usage Logging

The Cost of Logging
Logging resource usage data to database tables incurs costs:
23

Writing to the database adds to the system I/O load. On a heavily loaded system, this could
affect the production workload throughput.
The rows written to the database take up space. If this space is never reclaimed, it will
eventually grow to consume all available space in user DBC.
In an extremely loaded system, it is possible that the RSS can fall behind in writing data to
the database. Although it caches such data and eventually catches up if given a chance, the
RSS is forced to start discarding rows if the system load persists and its cache capacity is
exceeded.
Logging Cost Contributors

Logging costs are difficult to quantify. They depend on a number of interrelated factors:
How busy is the system
Which resource usage tables are enabled
What resource usage logging rates are in effect
The system configuration (vproc, CPU, host driver, logical devices or device controllers)
Operational Methods
To optimize performance and reduce the cost of resource usage logging on your system,
Teradata recommends that you:
1
Use Summary Mode to reduce the number of rows inserted into the resource usage tables
if Summary Mode data provides sufficient information for your needs.
Do not disable Active Row Filter Mode for the tables that it is by default enabled for (for
example, the ResUsageSps table). Active Row Filter Mode limits the number of rows
written to the database each logging period and minimizes the amount of system resources
used.
Avoid unnecessarily using or exhausting available disk space by doing the following:
Never enable logging on tables that you do not intend to use.

For example, logging only to the ResUsageSpma table provides a lot of useful
information with a minimal operational load on the system.
Use the largest rates that provide enough detail information for your purposes.
Generally, you should use a logging rate no smaller than 60. The default rate is 600.
These values can be adjusted any time, regardless of whether the database system is
busy. New values take effect as soon as the adjustment command is issued. (For
example, with ctl, when you issue the WRITE command.)
24
Purge old data from the resource usage tables periodically.

Related Topics
For instructions on...
See...
enabling resource usage tables, setting the

logging rates, and summarizing or filtering
rows
Enabling RSS Logging on page 27.
purging old data from resource usage tables
Purging Data on page 35.
25

26
Resource Usage and Procedures
CHAPTER 3
Enabling RSS Logging

You can enable the resource usage tables using the Control GDO Editor (ctl) utility or the
database commands in Database Window (DBW).
Before you enable the resource usage tables, determine which tables apply to the resource
usage macros you want to run. For more information, see:
Resource Usage Table Settings on page 19.
Logging Rate on page 21.
ctl
You can set various Teradata Database configuration settings using ctl. The RSS screen in ctl
allows you to specify the rates of resource usage data logging. For detailed information on
starting ctl and modifying the settings, see ctl in Utilities.
DBW
Note: For instructions on starting the Database Window, see Database Window (xdbw) in
Utilities.
To enable RSS logging from DBW
1
Open the Supvr window.
Set the Node Logging Rate using the database command below.
SET RESOURCE
LOGGING
number
LOG
1099D002
where number is the number of seconds.

Note: A rate of zero disables the logging function.
27
Chapter 3: Resource Usage and Procedures

Enabling RSS Logging
3
Specify the table you want to enable logging to using the database command below.
SET LOGTABLE
tablename
ON
ALL
OFF
FE0CA030
where tablename is the suffix part of ResUsageXxxx. For example, for the
DBC.ResUsageSpma table, the tablename would be Spma.
After the table is enabled for logging, you can log rows in Summary Mode. For more
information, see Summary Mode on page 22.
Note: To log rows in Summary Mode, you must enable the table specified in both the RSS
Table Logging Enable group and in the RSS Summary Mode Enable group.
4
(Optional) Enable Summary Mode on the table specified using the command below.
SET SUMLOGTABLE
tablename
ON
OFF
1095A010
Example
The following example shows you how to enable table logging and set the Logging rate using
the database commands in DBW. Suppose you want to enable the ResUsageShst table and set
the logging rate for 10 minutes (600 seconds). You would enter the following:
set logtable shst on
set resource log 600
28
General Macro Input Format

As shown in the table below, there are four kinds of macros:
Multiple-node
One-node
All-node
ByGroup
For any given line in the following table, the macros on that line report the same statistics for either multiple nodes, one node,
all nodes, or group nodes as indicated.
Description
Multinode Macro
AWTs in use by node
ResAWTByAMP
ResAWTByNode
CPU usage by AMP Vprocs
ResCPUByAMP
ResCPUByAMPOneNode
ResAmpCpuByGroup
CPU usage by PE Vprocs
ResCPUByPE
ResCPUByPEOneNode
ResPeCpuByGroup
CPU usage by nodes
ResCPUByNode
ResCPUOneNode
ResCpuByGroup
Host statistics
One-Node Macro
All-Node Macro
ByGroup Macro
ResAWT
ResHostOneNode
ResHostByLink
ResHostByGroup
Ldv disk statistics
ResLdvByNode
ResLdvOneNode
ResLdvByGroup
Memory management
ResMemMgmtByNode
ResMemMgmtOneNode
ResMemByGroup
General network statistics
ResNetByNode
ResNetOneNode
ResNetByGroup
General node-level statistics
ResNodeByNode
ResOneNode
pdisk level I/O statistics
ResPdskByNode
ResPdskOneNode
Priority Scheduler and TASM

Workload statistics
ResPsByNode
AMP level I/O statistics
ResVdskByNode
ResNode
ResNodeByGroup
ResPdskByGroup
ResPsByGroup
ResVdskOneNode
ResVdskByGroup
29

General Macro Input Format
Parameter Use for One-Node, Multiple-Node, All-Node, and Group Macros

The following table explains parameter use for one-node, multiple-node, all-node, and group
macros.
Macro Type
Number of Parameters
Node Parameters Used
Multiple node
Six
FromNode, ToNode
One node
Five
Node
All node
Four
None; this macro reports system-wide

statistics.
Group
Four
None; this macro reports statistics for

all nodes in the group.
For instructions on using these macros, see Macro Execution on page 31.
Input Format for One-Node Macros

One-node macro versions are primarily used on single-node systems. Alternatively, you can
use the corresponding multiple-node macro to report on just one node by supplying equal
FromNode and ToNode parameters. One-node versions are recommended, however, because
they eliminate redundant report columns on a single-node system. Examples of redundant
columns are the NodeID column and columns that focus on cross-node load balancing.
OneNode macros have the same general input format as the other macros. The only
differences are that the single-node version of each macro has both of the following:
OneNode qualifier in the macro name.
A single node specification, instead of the FromNode and ToNode parameters to specify a
range of nodes. The default is '001-01'.
Input Format for ByGroup Macros

ByGroup macro versions are used on systems with co-existing nodes. In Teradata Database,
co-existing nodes are nodes of different model types in the same configurations. Because of
the differences, the nodes may become bottlenecks in the throughput of the system as a whole.
Therefore, ByGroup macros were developed to provide the system user with a summary of the
performance data based on node groupings.
Note: The Database Administrator must identify the groupings of nodes when the system is
first configured.
ByGroup macros are similar to the other macros. The only difference is that they use the
GroupId column of the views to report system usage for a specific set of nodes grouped by a
GroupId. The input format of the ByGroup macros is the same as the other macros except
ByGroup appears as the qualifier in the macro name.
For an example of a ByGroup macro, see ResAmpCpuByGroup Sample Output on
page 190.
30

Macro Execution
Retaining Data From Prior Releases

If you expect an ongoing need to retain and analyze data from prior releases, ask your System
Administrator to retain two sets of view and macro Data Definition Language (DDL) files in
separate places. Rename the views and macros so that you can use either.
You could, for example, use ResNodeR12 as the name of the resource usage macro from an
older release, and use it when you want to analyze the data from that release.
Macro Execution
Function
Macro execution is illustrated in the following diagram. For details about each macro and its
resulting report, see Chapter 15: Resource Usage Macros.
EXECUTE MACRO Syntax

The execution of each resource usage macro has the following form. For information on
interpreting the syntax diagrams, see Appendix A: How to Read Syntax Diagrams.
EXECUTE
MacroNameMultiNode
FromDate
EXEC
MacroNameAllNode
(
FromDate
MacroNameOneNode
,
ToTime
ToDate
,
FromTime
ToDate
FromDate
A
,
ToDate
FromDate
MacroNameByGroup
ToDate
FromNode
);
ToNode
B
FromTime
ToTime
,
C
FromTime
,
ToTime
Node
D
FromTime
ToTime
1099A004
where:
31

Macro Execution
Syntax element
Description
MacroNameMultiNode
Name of a multinode resource usage macro:
MacroNameAllNode
ResAwtByNode
ResCPUByAMP
ResCPUByPE
ResCPUByNode
ResLdvByNode
ResMemMgmtByNode
ResNetByNode
ResNodeByNode
ResPdskByNode
ResPsByNode
ResVdskByNode
Name of an all-node resource usage macro:

ResNode
ResHostByLink
The ResHostByLink and ResNode macros do not use the FromNode and
ToNode parameters.
MacroNameByGroup
Name of a ByGroup resource usage macro:
FromDate
ResAmpCpuByGroup
ResCPUByGroup
ResHostByGroup
ResLdvByGroup
ResMemByGroup
ResNetByGroup
ResNodeByGroup
ResPeCpuByGroup
ResPdskByGroup
ResPsByGroup
ResVdskByGroup
Start date to report resource usage data.

The date may be entered either as a character string (for example,
character format for May 31, 2007 would appear as '2007-05-31') or as a
numeric value (for the same date in numeric format, 1070531). The
character string is the recommended format. The default is the current
system date.
For more information on using numeric dates with macros, see SQL
Data Types and Literals.
Note: The character string date format is changed from yymmdd to
'yyyy-mm-dd' to accommodate dates in the 21st century.
ToDate
End date to report resource usage data.

See the FromDate syntax element column for a further explanation of
date formats.
The character string is the recommended format.
32
FromTime
Start time to report resource usage data. The format is hhmmss. The
default is 000000.
ToTime
End time to report resource usage data. The format is hhmmss. The
default is 999999.

Macro Execution
Syntax element
Description
FromNode
Starting range of nodes to report resource usage data. The format is

'nnn-nn'. A hyphen must be included in the fourth character position.
The default is '000-00'.
Note: To identify the node ID numbers for your system, type
get config in the DBW Supervisor Window (Supvr).
ToNode
Ending range of nodes to report resource usage data. The format is 'nnnnn'. A hyphen must be included in the fourth character position. The
default is '999-99'.
Note: To identify the node ID numbers for your system, type
get config in the DBW Supervisor Window (Supvr).
Node
Single-node ID to report resource usage data. The format is 'nnn-nn',

and hyphen must be included in the forth character position. For
example,1-01 should be typed out as '001-01'. The default is '001-01'.
Example 1: The ResCPUByAMP Macro

The following statement executes the ResCPUByAMP macro, producing a report for the
period beginning 8:00 a.m. on December 25, 2006 and ending 12:00 midnight, on December
31, 2006. It includes data for nodes 123-02 through 125-04.
EXECUTE ResCPUByAmp('2006-12-25','2006-12-31', 080000, 240000,
'123-02','125-04');
where:
Statement Element
Description
ResCPUByAMP
Name of the resource usage macro
'2006-12-25'
Starting date of December 25, 2006
'2006-12-31'
Ending date of December 31, 2006
080000
Starting time of 8:00 a.m.
240000
Ending time of 12:00 midnight
'123-02'
Starting node of a range of nodes
'125-04'
Ending node of a range of nodes
For information on using numeric values for dates, see SQL Data Types and Literals.
Example 2: The ResCPUByAMPOneNode Macro

The following statement executes the OneNode version of the ResCPUByAMP macro shown
in Example 1. It uses the same starting and ending dates and times (using character string
format), except the report is for a single node, node 123-02.
33

Using ENABLE and DISABLE LOGON Commands
EXECUTE ResCpuByAmpOneNode ('2006-12-25','2006-12-31',080000,
240000,'123-02');
where:
Statement Element
Description
ResCPUByAMPOneNode
'2006-12-25'
'2006-12-31'
080000
240000
'123-02'
Node
Example 3: The ResAMPCpuByGroup Macro

The following statement executes the ByGroup version of the ResCPUByAmp macro shown in
Example 1. It uses the same starting and ending dates and times (using character string
format), except the report is for a node grouping.
EXECUTE ResAMPCpuByGroup ('2006-12-25','2006-12-31',080000,
240000);
where:
Statement Element
Description
ResCPUByAMPByGroup
'2006-12-25'
'2006-12-31'
080000
240000
Using ENABLE and DISABLE LOGON Commands

The DISABLE LOGONS command prevents new sessions from logging on. When logons are
disabled, resource usage data stops logging to the tables even if there are still active sessions
34

Purging Data
logged on. (DISABLE ALL LOGONS prevents all users, including user DBC, from logging on
and also stops logging to the tables.)
To enable logons from:
Database Window, run ENABLE LOGONS or ENABLE ALL LOGONS.
Teradata Command Prompt, use the Start With Logons field of the Screen Debug menu of
ctl. For more information, see Control GDO Editor (ctl) in Utilities.
For more information on enabling and disabling logons, see Changing Logon States and
Restarting the System in Database Administration.
Purging Data
The RSS does not automatically delete data from the resource usage tables. You need to purge
data you no longer need on a regular basis.
You can directly remove old resource usage data by submitting SQL statements. For example,
use the following SQL statement to remove data more than seven days old from the
ResUsageSpma table:
DELETE FROM ResUsageSpma WHERE TheDate < CURRENT_DATE - 7;
For more information about the DELETE syntax, see Statement Syntax in SQL Data
Manipulation Language.
35

Purging Data
36
CHAPTER 4
Resource Usage Tables
Physical Table Naming Conventions

Each physical table name follows this general naming convention:
ResUsage Information_type Table_name
where:
Element
Is one of the following...
Information_type
S = System-wide information
I = Internal Teradata Database information
Table_name
pma = Node information

vpr = Vproc information
cpu = CPU-specific information
ldv = Logical device statistics
pdsk = Pdisk device statistics
vdsk = Vdisk device statistics
awt = AMP Worker Task statistics
sps = WD resolution statistics
hst = Mainframe and workstation host information
Relational Primary Index

All resource usage tables have the same nonunique primary index:
The nonunique primary index consists of TheDate, TheTime, and NodeID columns.
The primary index is nonunique because:
Of duplicate rows that will appear with the same timestamp during daylight savings
time.
Some tables have multiple rows per logging period.
Rows that have duplicate timestamps can be distinguished by the GmtTime column.
Because the primary index is nonunique, all resource usage tables are created as
MULTISET tables. This prevents the system from checking for duplicate rows.
37
Chapter 4: Resource Usage Tables

Resource Usage Table Rows
For more information on MULTISET tables, see CREATE TABLE (Table Kind Clause) in
SQL Data Definition Language or Duplicate Rows in Tables in SQL Fundamentals.
Resource Usage Table Rows

For information on how rows will be inserted into these tables based on the current resource
usage control settings, see Chapter 2: Planning Your Resource Usage Data. For information
on the number of rows inserted in a resource usage table for each applicable log period, refer
to Summary Mode on page 22.
Occasional Event Data

Occasional event data is considered outside the scope of resource usage and is, therefore,
logged in the ERRORLOG and the DBCINFO tables rather than in the resource usage tables.
Types of Resource Usage Table Columns

This manual describes what each of the resource usage table columns report (that is, what
each DBC.ResUsageXxxx.ColumnName reports) in a table format.
Note: The actual table definitions are obtainable by executing the SHOW TABLE statement.
For more information about SHOW TABLE, see SQL Data Definition Language.
All columns described in the following chapters and appendixes are type FLOAT unless
otherwise specified in the description of that column. All nonexistent values are stored as
NULL.
For each resource usage table column, this manual describes the:
Column Name
Mode
For descriptions of the different modes of data reporting, see About the Mode Column
on page 42.
Description
Data Type
For descriptions of the data types described in this manual, see SQL Data Types and
Literals.
The columns are grouped into either housekeeping columns or statistics columns. Statistic
columns are further grouped by category and subcategory as shown in the following table.
38

Column Name
Mode
Description
Data Type
HOUSEKEEPING OR STATISTICS COLUMNS

CATEGORY
Subcategory
Each table has:
Housekeeping columns which contain statistics on timestamp and current logging

characteristics.
Statistics columns which can be further categorized into subcategories. Categories and
subcategories may vary from table to table.
The following table shows the types of statistics subdivided into their respective subcategories.
Category
Subcategories
Description
File System
Some of the columns can be viewed as a subset of

memory columns by expanding on the operations
performed on disk memory segments. Operations
counted are logical memory and physical disk reads
and writes (including aging) and locking control
activities. Other columns identify the purpose of
operations being performed on disk segments such
as cylinder migration or data updates, the requests
being made by database software on the file system,
the number of cylinders that were selected or
scanned by AutoCylPack, or the number of data
blocks that have been compressed or uncompressed.
The WAL columns identify the log-based file system
recovery scheme in which modifications to
permanent data are written to a log file, the WAL
log.
AutoCylPack
Block-level compression (BLC)
Cylinder Defragmentation Overhead
Cylinder Management Overhead
Events
Cylinder MiniCylPack Overhead
Cylinder Split and Migrate Overhead
File Segment (FSG) Cache Wait
FSG I/O
Data Block Creations
Data Block Merge
Data Block Prefetches
Data Block Update Operations
Data Segment Lock Requests
Depot
Master Index (MI)
Multi-Row Requests
Segments Acquired
Segments Released
Single-Row Requests
Synchronized Full Table Scans
Transient Journal Requests
Transient Journal Overhead
Write Ahead Logging (WAL)
39

Category
Subcategories
Description
General Concurrency
Control
Database Locks
Monitor Management
These columns identify concurrency control

activities. These activities provided are subdivided
into control done for user level processing, system
overhead processing, and database locks. It does not
include control specific to disk, memory or net
concurrency control, which are included in the disk,
memory or net columns.
Host Controller (SHST)
Channel Traffic
Channel Management
Controller Overhead
These columns identify traffic on the host-to-node

channels and TCP/IP networks. Some also give
overhead and management information on the host
channel and TCP/IP network.
Memory
Memory Allocations
Memory Availability Management
Memory Pages
Memory Resident
Paging
Task Context Segment Usage
These columns collect memory allocation and

deallocation, logical memory and physical disk
reads and writes (including paging), access, deaccess
and memory control.
Input and Output Traffic

Outstanding Requests
Response Time
Seek Statistics
These columns identify individual logical device

activities for external storage components
connected through the buses.
Broadcast Net Traffic

Group Coordination
Message Type
Message Delivery Times
Merge Services
Net Controller Status and
Miscellaneous Management
Net Circuit Management
Net Miscellaneous Contention
Management
Net Queues
Network Transport Data
Point-to-point Net Traffic
Work Mailbox Queue
These columns identify traffic over the BYNET

through the number and direction of messages,
subdivided into the type of transmission, as well as
physical utilization of the BYNET. Logical messages
and direction are identified through subdivisions of
the message class. Controller overhead, channel
utilization, and Teradata net contention are
identified as well.
Logical Device
Net
40
Memory management columns are also provided to

identify events leading up to paging and aging
activities. Finally, a detailed snapshot of the memory
is provided by tracking the current states per
memory types.
The storage device statistics are calculated only on

what can be derived from statistics collected by the
operating system, since the disk array controllers do
not provide us with any useful data for resource
usage.

Category
Subcategories
Description
Process Scheduling
These columns provide a CPU-level snapshot of

work started with current characteristics and states.
Expanded detail is provided for work started but
waiting on resources. This identifies the ability or
inability of the system to effectively utilize resources.
Time allotments are tracked by monitoring the time
spent waiting for resources or processing code.
Some of these columns also track the number of
times processing was switched to another process
for multitasking purposes or to perform interrupt
services; and others can provide information about
wether the UDFs were doing work for the AMP and
PE vprocs.
ChnSignal Status Tracking

CPU Utilization
Cylinder Read
Interrupted CPU Switching
PE and AMP User-Defined Function
(UDF) CPU
Process Allocation
Process Pending Snapshot
Process Block Counts
Process Pending Wait Time
Scheduled CPU Switching
Reserved
None
These columns are not used.
Spare
None
These columns are reserved for a future release or

are for internal manipulation by Teradata
developers.
TASM
These columns report statistics on the AMP Worker

Tasks (AWTs) and Priority Scheduler. These
columns also provide RSS data to the System PMPC
APIs (for example, the MONITOR WD request and
MonitorWD function). For information, see
Application Programming Reference.
Transient Journal
Management
Purge Overhead
These columns identify background overhead

associated with the occasional transient journal
purge operation.
User Commands
User command
User command Arrival and
Departure
These columns describe the types of commands

given to Teradata Database by the user and the
progress of those commands.
Teradata Virtual Storage

(VS)
These columns identify individual pdisk and vdisk

device activities.
AMP Worker Task

In use and Max Array Data
Priority Scheduler
Worktype Descriptions
Monitor WD
Allocation
I/O
Migration
Node Agent
Invalid Columns
Some of the resource usage table columns described in this manual are not currently valid.
These columns are shaded in gray. For example:
Column Name
Mode
Description
Data Type
SampleColumn
track
An example of a column that is

not currently valid.
FLOAT
41

About the Mode Column
About the Mode Column

The Mode column describes the kind of data reported by the resource usage column.
IF the Mode column is ....
THEN these resource usage columns ...
count
tally the number of times an event occurred, such as the number of

disk reads or writes during the logging period.
max
report the maximum value recorded during the logging period. Most
of the max columns have a max suffix in the column name (for
example, IoRespMax).
min
report the minimum value recorded during the logging period. The
min columns have a min suffix in the column name (for example,
AvailableMin).
track
show the value of a countable item achieved at the end of the current
logging period. An example of a countable item is a queue length.
Summary Mode in Resource Usage Tables

Summary Mode combines data from the multiple data rows normally generated into one or
more rows. When multiple rows are condensed into a single row, the data is combined using
the rules in the following table.
For the following
Mode
Summary fields are combined by
Count
summing the values from all the contributing rows.

For example, the ResUsageSawt.FlowCtlCnt field provides the total number of times that the
system entered flow control state from a non-flow control state.
In normal mode, the values are reported per AMP.
In Summary Mode, the value needs to be divided by the number of AMPs if you wish to
determine the average number per AMP rather than the total.
Max
taking the maximum value from all the contributing rows.

In Summary Mode, the reported value for a Max field such as
ResUsageSpdsk.ConcurrentWriteMax is the maximum value from all the rows that are combined
into a single summary row.
Min
taking the minimum value from all the contributing rows.

In Summary Mode, the reported value for a Min field, such as AvailableMin is the minimum value
from all the rows that are combined into a single summary row.
42

For the following

Mode
Summary fields are combined by
Track
summing the values from all the contributing rows.

In Summary Mode, the Track values from each row to be combined are summed together. For
example, ResUsageSawt.FlowControlled is a track field so that in Summary Mode, all the AMP
vproc rows are combined into a single row and the FlowControlled field will report the summed
value from each of the AMP vproc rows.
Summary Mode is applicable to all tables except ResUsageSpma, ResUsageIpma, and

ResUsageSps. If the information for a row of a table is in Summary Mode, the SummaryFlag
value is set to 'S'. If the row is being logged normally, the SummaryFlag value is set to 'N'.
43

44
CHAPTER 5
ResUsageScpu Table
This resource usage table contains resource usage information specific to the CPUs within the
nodes. The ResUsageScpu table includes resource usage data for available system-wide,
CPU information.
Note: This table is created as a MULTISET table. For more information see Relational
Primary Index on page 37.
Housekeeping Columns
Relational Primary Index Columns
These columns, taken together form the nonunique primary index.
Column Name
Mode
Description
Data Type
TheDate
n/a
Date of the log entry.
DATE
TheTime
n/a
Nominal time of the log entry.
FLOAT
Note: Under conditions of heavy system load, entries may be logged

late (typically, by no more than one or two seconds), but this
column will still contain the time value when the entry should have
been logged. See the Secs and NominalSecs columns.
NodeID
n/a
Node ID. The Node ID is formatted as CCC-MM, where CCC

denotes the three-digit cabinet number and MM denotes the twodigit chassis number of the node. For example, a node in chassis 9 of
cabinet 3 has a node ID of '003-09'.
INTEGER
Note: Symmetric Multi-Processing (SMP) nodes have a chassis and

cabinet number of 1. For example, the node ID of an SMP node is
'001-01'.
Miscellaneous Housekeeping Columns

These columns provide statistics on current logging characteristics.
Column Name
Mode
Description
Data Type
GmtTime
n/a
Greenwich Mean Time is not affected by the Daylight Savings Time

adjustments that occur twice a year.
FLOAT
NodeType
n/a
Type of node, representing the per node system family type. For
example, 5600C or 5555H.
CHAR(8)
45
Chapter 5: ResUsageScpu Table

Statistics Columns
Column Name
Mode
Description
Data Type
CPUId
n/a
Identifies the CPU within this node. The values are 0 through
NCPUs-1.
SMALLINT
In Summary Mode, the value is -1.

Secs
n/a
Actual number of seconds in the log period represented by this row.

Normally the same as NominalSecs, but can be different in three
cases:
SMALLINT
The first interval after a log rate change

A sample logged late because of load on the system
System clock adjustments affect reported Secs
Useful for normalizing the count statistics contained in this row, for
example, to a per-second measurement.
CentiSecs
n/a
Number of centiseconds in the logging period. This column is

useful when performing data calculations with small elapsed times
where the difference between centisecond-based data and whole
seconds results in a percentage error.
INTEGER
NominalSecs
n/a
Specified or nominal number of seconds in the logging period.
SMALLINT
SummaryFlag
n/a
Summarization status of this row. Possible values are 'N' if the row
is a non-summary row, and 'S if the row is a summary row.
CHAR (1)
Active
max
Controls whether or not the rows will be logged to the resource

usage tables if Active Row Filter Mode is enabled.
FLOAT
If Active is set to a non-zero value, the row contains data columns.

If Active is set to a zero value, none of the data columns in the row
have been updated during the logging period.
For example, if you enable Active Row Filter Mode, the rows that
have a zero Active column value will not be logged to the resource
usage tables.
TheTimestamp
n/a
Number of seconds since midnight, January 1, 1970.
BIGINT
This column is useful for aligning data with the DBQL log.
CODFactor
n/a
Platform Metering (PM) CPU COD value in one tenths of percent.

For example, a value of 500 represents a PM CPU COD value of
50%.
SMALLINT
The value is set to 1000 if the PM CPU COD is disabled.
Statistics Columns
Process Scheduling CPU Utilization Columns
These columns count all CPU activities, including activities performed for virtual processors.
The CPU utilization columns are aggregates representing all CPUs on the node. CPU
utilization by user code is further subdivided by the vproc tables.
46

Summary Mode
CPU idle time = CPUIdle + CPUIoWait
CPU busy time = CPUUServ + CPUUExec
Theoretically, the values of these four columns, for any given interval, account for total CPU
time on the node. That is, these columns should total to 100 * Secs * number of CPUs on the
node, since each CPU is always in exactly one of these four states. In practice, there is
occasionally a very small plus or minus difference from this theoretical total.
Column Name
Mode
Description
Data Type
CPUIdle
count
Time in centiseconds the CPU is idle and not waiting for I/O.
FLOAT
CPUIoWait
count
Time in centiseconds CPU is waiting for I/O completion.
FLOAT
Note: This represents another variety of Idle, since the CPU is only
recorded as being in this state if there are no processes eligible for
execution. This is because if there were any such process, the CPU
would be immediately dispatched for that process.
CPUUServ
count
Time in centiseconds CPU is busy executing kernel system calls or

servicing I/O and timer hardware interrupts.
FLOAT
CPUUExec
count
Time in centiseconds CPU is busy executing user execution code,

that is, time spent in a user state on behalf of a process.
FLOAT
Reserved Columns
Column Name
Mode
Description
Data Type
Reserved
n/a
Note: This column is not used.
CHAR (1)
Reserved2
n/a
CHAR (2)
Summary Mode
When Summary Mode is active for tables in this group, one row is written to the database for
each node, summarizing all CPUs per node, for each log interval.
You can determine if a row is in Summary Mode by checking the SummaryFlag column for
that row.
IF the SummaryFlag column value is
THEN the data for that row is being logged
'S'
in Summary Mode.
'N'
normally.
47

Spare Columns
Spare Columns
The ResUsageScpu table spare fields are named Spare00 through Spare04, and SpareInt.
The SpareInt field has a 32-bit internal resolution while all other spare fields have a 64-bit
internal resolution. All spare fields default to count data types but can be converted to min,
max, or track type data fields if needed when they are used.
The following table describes the Spare field currently being used.
Column Name
Description
Spare00
Workload management (WM) CPU COD value in one tenths of a

percent. For example, a value of 500 represents a WM CPU COD
value of 50.0%.
The value is set to 1000 if the WM CPU COD is disabled.
Note: This field will be converted to the WM_CPU_COD column in
Teradata Database 15.0. You can access resource usage data for this
field using the WM_CPU_COD column name in the ResScpuView
view. For details, see ResScpuView on page 169
Note: WM CPU COD is not supported on SLES 10. Its value is set to
1000 on SLES 10.
Related Topics
For details on the different type of data fields, see About the Mode Column on page 42.
48
CHAPTER 6
ResUsageSpma Table
The ResUsageSpma table includes resource usage data for available system-wide information.
You can use the ResUsageSpma table to identify node-level skew by comparing maximum
CPU usage to the average CPU consumed across all nodes.
The ResUsageSpma table is similar to the ResUsageIpma table. For information on this table,
see Appendix B: ResUsageIpma Table.
Note: Summary Mode is not applicable to the ResUsageSpma table.
This table is created as a MULTISET table. For more information, see Relational Primary
Index on page 37.
Note: Always use the views provided in Chapter 14: Resource Usage Views to access the data
rather than accessing the resource usage table directly.
These columns taken together form the nonunique primary index.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT
Note: Under conditions of heavy system load, entries may be

logged late (typically, by no more than one or two seconds), but this
column will still contain the time when the entry should have been
logged. For more information, see the Secs and NominalSecs
columns.
NodeID
n/a

denotes the three-digit cabinet number and MM denotes the twodigit chassis number of the node. For example, a node in chassis 9
of cabinet 3 has a node ID of '003-09'.
INTEGER
Note: SMP nodes have a chassis and cabinet number of 1. For

example, the node ID of an SMP node is '001-01'.

These columns provide a generalized picture of the vprocs running on this node, shown as
Type n virtual processors where n = 1 to 7. Under the current implementation, only Type 1
49
Chapter 6: ResUsageSpma Table

(AMP), Type 2 (PE), Type 3 (GTW), Type 4 (RSG), and Type 5 (TVS) vprocs exist; vproc
types 6 through 7 are not currently used.
Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
NodeType
n/a
example, 5100C, 5600C or 5555H.
CHAR(8)
TheTimestamp
n/a
BIGINT
CentiSecs
n/a
Actual number of centiseconds in the logging period.
INTEGER
This column is useful when performing data calculations with small

elapsed times where the difference between centisecond-based data
and whole seconds results in a percentage error.
Secs
n/a

cases:
SMALLINT

NominalSecs
n/a
SMALLINT
CODFactor
n/a
PM CPU COD value in one tenths of percent. For example, a value

of 500 represents a PM CPU COD value of 50%.
SMALLINT

NCPUs
n/a
Number of CPUs on this node.
SMALLINT
This column is useful for normalizing the CPU utilization column

values for the number of CPUs on the node. This is especially
important in coexistence systems where the number of CPUs can
vary across system nodes.
Vproc1
n/a
Current count of type 1 (AMP) virtual processors running on the

node.
SMALLINT
Vproc2
n/a
Current count of type 2 (PE) virtual processors running under the

node.
SMALLINT
Vproc3
n/a
Current count of type 3 (GTW) virtual processors running under

the node.
SMALLINT
Vproc4
n/a
Current count of type 4 (RSG) virtual processors running under

the node.
SMALLINT
Vproc5
n/a
Current count of type 5 (TVS) virtual processors running under the

node.
SMALLINT
50

Column Name
Mode
Description
Data Type
Vproc6
n/a
Current count of type 6 virtual processors running under the node.
SMALLINT
This column reports zeros and (blanks).

Vproc7
n/a
SMALLINT

VprocType1
n/a
Type of virtual processor for Vproc1. When the vproc is present on

the node, the value is AMP.
CHAR(4)
VprocType2
n/a

the node, the value is PE.
CHAR(4)
VprocType3
n/a

the node, the value is GTW.
CHAR(4)
VprocType4
n/a

the node, the value is RSG.
CHAR(4)
VprocType5
n/a

the node, the value is TVS.
CHAR(4)
VprocType6
n/a
Type of virtual processor for Vproc6.
CHAR(4)
Note: This column is not currently valid. It should not be used.

VprocType7
n/a
Type of virtual processor for Vproc7.
CHAR(4)

MemSize
n/a
Amount of memory on this node in mega bytes. Useful for

performing memory usage calculations.
INTEGER
NodeNormFactor
n/a
A per CPU normalization factor that is used to normalize the

reported CPU values of the ResUsageSpma table to a single 5100
CPU.
INTEGER
This value is scaled by a factor of 100. For example, if the actual

factor is 5.25, the value of the NodeNormFactor will be 525.
Active
max
Gets set to a non-zero value whenever one of the other data

columns in the row is set.
FLOAT
NetSamples
count
Sample count for sampled statistics for a Bynet.
FLOAT
Note: NetSamples is used to normalize all net time monitored

statistics to a percent-of-time basis. For example, dividing
(NetTxIdle/NetSamples) yields the transmitter-idle time ratio for
the net statistics.
51

Statistics Columns
Statistics Columns
File System Columns
Segments Acquired Columns
These columns identify the total disk memory segments acquired by the File System during
the log period.
For more information, see Segment Acquires Columns in the Chapter 13: ResUsageSvpr
Table.
Column Name
Mode
Description
Data Type
FileAcqKB
count
Total KB logically acquired by FileAcqs.
FLOAT
Note: Use the views provided in Chapter 14: Resource Usage

Views instead of accessing the data for this column directly from
this table.
FileAcqOtherKB
count
Total number of scratch disk segments acquired in KB.
FLOAT
FileAcqReadKB
count
Total KB physically read by FileAcqReads.
FLOAT
FileAcqReads
count
Total number of disk segment acquired that caused a physical read.
FLOAT
FileAcqs
count
Total number of disk segments acquires that were logically

acquired.
FLOAT
Note: Only the FileAcqs column is counted as a cache hit.

FileAcqsOther
count
Total number scratch disk segments that were logically acquired.
FLOAT
Segments Released Columns

These columns identify the total disk memory segments released by the File System, as well as
those segments that are dropped from memory during the log period.
For more information, see Segments Released Columns in Chapter 13: ResUsageSvpr
Table.
Column Name
Mode
Description
Data Type
FileRels
count
Total number of disk segments released by tasks.
FLOAT
FileRelsOther
count
Total number of scratch disk segments released as part of them

being deleted.
FLOAT
FileRelKB
count
Total KB logically released by FileRels.
FLOAT

this table.
FileRelOtherKB
52
count
Total number of scratch disk segments released in KB.
FLOAT

Statistics Columns
Column Name
Mode
Description
Data Type
FileWriteKB
count
Total KB physically written by FileWrites.
FLOAT
Note: This column may produce a negative number due to

overflow. To avoid this issue, see ResSpmaView on page 173 to
access the ResUsageSpma table data.
FileWrites
count
Total number of disk segment immediate or delayed physical

writes.
FLOAT
Data Block Prefetches Columns

These columns summarize the effects of prefetching data blocks on the file system. For more
information, see Data Block Prefetches Columns in Chapter 13: ResUsageSvpr Table.
A prefetch is either a cylinder read operation or individual block reads operation. Either of
these operations are generically called a prefetch.
When all cylinder slots are in use, the cylinder reads revert back to the original algorithm of a
block-at-a-time read ahead. So the column FilePreKB is the sum of the size of data blocks
logically read by either cylinder reads or data block pre-reads. This also applies to the physical
pre-reads. FilePreReadKB includes both physical cylinder reads and single block pre-reads.
The number of data blocks that are pre-read at a time is controlled by the DBS Control
performance parameter ReadAhead Count. The default is one block at a time pre-read.
If you enable cylinder reads, there will be extra sectors read in on cylinder reads. An accurate
calculation of the wasted KB read by cylinder read is not possible since there are legitimate
logical pre-reads that do not incur physical pre-reads.
For more information about cylinder reads, see Database Administration.
Column Name
Mode
Description
Data Type
FilePreKB
count
Sum of the sizes of data blocks logically loaded with data prefetches
(for example, either cylinder reads or individual block reads).
FLOAT
For cylinder reads, this column does not include the disk sectors in
between the loaded data blocks.
this table.
FilePreReadKB
count
Size of the data prefetch (cylinder section or individual blocks being

read) that is physically loaded from disk.
FLOAT
For cylinder reads, this column includes the disk sectors in between
the loaded data blocks.
Note: This column may produce a negative number due to
overflow. To avoid this issue, see ResSpmaView on page 173 to
access the ResUsageSpma table data.
53

Statistics Columns
Column Name
Mode
Description
Data Type
FilePreReads
count
Number of times a data prefetch was physically performed either as

a cylinder read or individual blocks read.
FLOAT
FilePres
count
Total number of times a logical data prefetch was performed (either

as a cylinder read or individual block reads).
FLOAT
Data Segment Lock Requests Columns

These columns summarize the number of lock requests, blocks, and deadlocks on a disk
segment.
For more information, see Data Segment Lock Requests Columns in Chapter 13:
ResUsageSvpr Table.
Column Name
Mode
Description
Data Type
FileLockBlocks
count
Number of lock requests that were blocked.
FLOAT
FileLockDeadlocks
count
Number of deadlocks detected on lock requests.
FLOAT
FileLockEnters
count
Number of times a lock was requested.
FLOAT
Depot Columns
These columns summarize the physical writes to the Depot used to protect in-place
modifications.
Column Name
Mode
Description
Data Type
FileLargeDepotBlocks
count
Total number of blocks (either WAL or database) that have been

protected by large Depot writes.
FLOAT
Since a large Depot write protects multiple blocks, the following

calculation results in the average number of blocks protected by
each large Depot write:
FileLargeDepotBlocks / FileLargeDepotWrites
FileLargeDepotWrites
54
count
Number of large writes to the Depot performed to protect in-place

modifications. Each large Depot write protects multiple in-place
writes of either WAL data blocks or database data blocks. The large
Depot is typically used when blocks age out of memory in the
background. Large Depot writes are also counted against
FileWrites. Therefore, FileWrites still indicates the total writes
regardless of whether it was a Depot write or a database write.
FLOAT

Statistics Columns
Column Name
Mode
Description
Data Type
FileSmallDepotWrites
count
Number of small writes to the Depot performed to protect in-place

modifications. Each small Depot write protects a single in-place
write of either a write ahead logging (WAL) data block or a
database data block. The small Depot is typically used when the inplace writes are initiated by a foreground task. Small Depot writes
are also counted against FileWrites. Therefore, FileWrites still
indicates the total writes regardless of whether it was a Depot write
or a database write.
FLOAT
General Concurrency Control Database Locks Columns

These columns identify database locking occurrences.
Column Name
Mode
Description
Data Type
DBLockBlocks
count
Number of times a database lock was blocked.
FLOAT
DBLockDeadlocks
count
Number of times a database lock was deadlocked.
FLOAT
Host Controller Channel and TCP/IP Network Traffic Columns

These columns identify the traffic between the host and the node channel and TCP/IP
network connections in three levels of granularity:
Blocks
Messages
KB
Blocks are made up of some amount of variable sized messages. ReadKB and WriteKB identify
the KB involved in the traffic.
For host controller columns that provide overhead and management information, see
Chapter 8: ResUsageShst Table for details.
Column Name
Mode
Description
Data Type
HostBlockReads
count
Number of blocks read in from the host.
FLOAT
HostBlockWrites
count
Number of blocks written out to the host.
FLOAT
HostMessageReads
count
Number of messages read in from the host.
FLOAT
HostMessageWrites
count
Number of messages written out to the host.
FLOAT
HostReadKB
count
KB transferred in from the host.
FLOAT
HostWriteKB
count
KB transferred out to the host.
FLOAT
55

Statistics Columns
Memory Columns
Memory Allocation Column
Column Name
Mode
Description
Data Type
MemVprAllocKB
track
Change in memory.
FLOAT
MemVprAllocKB represents a delta from the previous reporting

period and will report negative values as less memory is used.
Note: The original meaning of this column was the total KB
attributed to allocations and size-increasing alters for vproc
memory types.
Memory Pages Column

Column Name
Mode
Description
Data Type
MemFreeKB
track
Approximate amount of memory that is available for use. The

Linux operating system uses most free memory for buffers and
caching to improve performance, but the operating system can
reclaim that memory if it is needed by programs.
FLOAT
The following formula is used by the RSS to calculate the

MemFreeKB value:
MemFreeKB = MemFree + Buffers + Cached + SwapCached fsgavailpgs*kbperpage - (active_slabs*pgsperslab*kbperpage)
where the values:
MemFree, Buffers, Cached, and SwapCached come from /proc/
meminfo.
fsgavailpgs come from the PDE FSG subsystem.
active_slabs and pgsperslab come from /proc/slabinfo.
Memory Availability Management Columns

These columns identify overhead to managing memory when memory availability is a
problem.
Column Name
Mode
Description
Data Type
MemCtxtPageReads
count
Number of pages swapped in.
FLOAT
MemCtxtPageWrites
count
Number of pages swapped out.
FLOAT
MemTextPageReads
count
Number of pages paged minus the pages swapped in.
FLOAT
56

Statistics Columns
Net Columns
Point-to-Point Net Traffic Columns
These columns identify the number (Reads, Writes) and amount (ReadKB, WriteKB) of input
and output messages passing through the Teradata Database nets through point-to-point
(1:1) methods (PtP). It excludes TCP/IP traffic.
Column Name
Mode
Description
Data Type
MsgPtPReadKB
count
Total KB of net point-to-point messages input to processes on the

node via the message subsystem.
FLOAT
MsgPtPReads
count
Number of net point-to-point messages input to processes on the

FLOAT
MsgPtPWriteKB
count
Total KB of net point-to-point messages output from processes on

the node via the message subsystem.
FLOAT
MsgPtPWrites
count
Number of net point-to-point messages output from processes on

the node via the message subsystem.
FLOAT
Broadcast Net Traffic Columns

and output messages passing through the Teradata Database nets through broadcast (1:many)
methods (Brd).
Note: If a single broadcast message is delivered to multiple processes in this node, the
NetBrdReads and NetBrdReadKB are only incremented once.
Column Name
Mode
Description
Data Type
MsgBrdReadKB
count
Total KB of net broadcast messages input to processes on the node

via the message subsystem.
FLOAT
MsgBrdReads
count
Number of net broadcast messages input to processes on the node

via the message subsystem.
FLOAT
MsgBrdWriteKB
count
Total KB of net broadcast messages output from processes on the

FLOAT
MsgBrdWrites
count
Number of net broadcast messages output from processes on the

FLOAT
Network Transport Data Columns

These columns identify the number (Reads, Writes) and amount of input and output passing
through the Teradata Database nets.
On a single-node (virtual network [vnet]) system, net-specific statistics are not meaningful
and are always zero.
57

Statistics Columns
Column Name
Mode
Description
Data Type
NetMsgPtpWriteKB
count
Amount of point-to-point message data in KB transmitted by both

Bynets.
FLOAT
NetMsgPtpWrites
count
Number of point-to-point messages transmitted by both Bynets.
FLOAT
NetMsgBrdWriteKB
count
Amount of broadcast message data in KB transmitted by both

Bynets.
FLOAT
NetMsgBrdWrites
count
Number of broadcast messages transmitted by both Bynets
FLOAT
NetMsgPtpReadKB
count
Amount of point-to-point message data in KB received by both

Bynets.
FLOAT
NetMsgPtpReads
count
Number of point-to-point messages received by both Bynets.
FLOAT
NetMsgBrdReadKB
count
Amount of broadcast message data in KB received by both Bynets.
FLOAT
NetMsgBrdReads
count
Number of broadcast messages received by both Bynets.
FLOAT
NetRxKBBrd
count
Total broadcast KB received over all Bynets.
FLOAT
NetRxKBPtP
count
Total point-to-point KB received over all Bynets.
FLOAT
NetTxKBBrd
count
Total broadcast KB transmitted over all Bynets.
FLOAT
NetTxKBPtP
count
Total point-to-point KB transmitted over all Bynets.
FLOAT
Net Controller Status and Miscellaneous Management

These columns provide utilization and other status information about the Teradata Database
net controllers.
On a single-node (VNET) system, net-specific statistics are not meaningful and always report
zero.
Column Name
Mode
Description
Data Type
NetTxConnected
count
Number of samples showing the transmitter connected on a Bynet.
FLOAT
NetRxConnected
count
Number of samples showing the receiver connected on a Bynet.
FLOAT
NetTxRouting
count
Number of samples showing the transmitter routing on a Bynet.
FLOAT
NetTxIdle
count
Number of samples showing the transmitter idle on a Bynet.
FLOAT
NetRxIdle
count
Number of samples showing the receiver idle on a Bynet.
FLOAT
Net Circuit Management Columns

The Net Circuit Management columns identify the management of Teradata net circuits
(Circ) and raw data traffic on the network (hardware) on all networks.
58

Statistics Columns
Column Name
Mode
Description
Data Type
NetBackoffs
count
Software backoffs, defined as BNS service blocked occurrences,

without regard for which net was involved.
FLOAT
NetHWBackoffs
count
Hardware backoffs reported by the BLM for all Bynets.
FLOAT
NetRxCircBrd
count
Total number (both normal and high priority) of broadcast circuits

received on all Bynets.
FLOAT
NetRxCircPtp
count
Total number (both normal and high priority) of point-to-point

circuits received on all Bynets.
FLOAT
NetTxCircBrd
count
Total number (both normal and high priority) of broadcast circuits

transmitted on all Bynets.
FLOAT
NetTxCircHPBrd
count
Number of high priority broadcast circuits transmitted on all

Bynets.
FLOAT
NetTxCircHPPtP
count
Number of high priority point-to-point circuits transmitted on all

Bynets.
FLOAT
NetTxCircPtp
count
Total number (both normal and high priority) of point-to-point

circuits transmitted on all Bynets.
FLOAT
Group Coordination Messages Columns

These columns identify messages that are communicated through the Teradata Database net
for coordination of a process among a group of vprocs. Coordination is handled either
through semaphores, groups, or channels.
Column Name
Mode
Description
Data Type
NetChanInUse
track
Number of channels in use at the current time.
FLOAT
NetChanInUseMax
max
Maximum number of channels in use.
FLOAT
MsgChnLastDone
count
Number of last done events that occurred on this node.
FLOAT
Note: The last AMP to finish an operation may send a last done
broadcast message indicating the work is done for this step. This is
used in tracking down the slowest node or AMP in the system. A
node or AMP that has more last done messages than the others
could be a bottleneck in the system performance.
NetGroupInUse
track
Number of groups in use at the current time. This number should

be same across all nodes.
FLOAT
NetGroupInUseMax
max
Maximum number of groups in use during each log interval.
FLOAT
NetSemInUse
track
Number of semaphores in use at the current time.
FLOAT
NetSemInUseMax
max
Maximum number of semaphores in use during each log interval.
FLOAT
59

Statistics Columns
Merge Services Columns

These columns identify activity occurring through merge (many:1) methods (Mrg) on
Teradata Database net.
Column Name
Mode
Description
Data Type
NetMrgRxKB
count
Number of KB received, without regard to which net, by merge

receive services for currently active merge operations.
FLOAT
NetMrgRxRows
count
Number of data rows received, without regard to which net, by

merge receive services for currently active merge operations.
FLOAT
NetMrgTxKB
count
Number of KB transmitted, without regard to which net, by merge

transmission services for currently active merge operations.
FLOAT
NetMrgTxRows
count
Number of data rows transmitted, without regard to which net, by

merge transmission services for currently active merge operations.
FLOAT
Process Scheduling Columns

Process Allocation Columns
These columns represent all currently allocated processes, subdivided into the possible process
states of running, ready, blocked or suspended.
Column Name
Mode
Description
Data Type
ProcBlocked
track
Number of threads blocked waiting for I/O at the current time.
FLOAT
ProcReady
track
Number of runnable or ready tasks, also called threads, able to

execute on CPUs when a CPU becomes available.
FLOAT
ProcReadyMax
max
Maximum number of ready tasks, also called threads, able to

execute on CPUs when a CPU becomes available.
FLOAT
Process Pending Snapshot Columns

These columns identify how many processes are blocked for each possible reason. These
columns total (minus ProcPendDBLock) approximately ProcBlocked, since we can only be
blocked on one blocking type at a time.
Note: In analyzing resource usage, a distinction should be made between the following two
kinds of process blocks:
Block involves a process that is logically idle, waiting to receive work on its primary
mailbox, or for a timer to elapse. This block does not affect throughput.
Block involves a process that has work to do but is being prevented from proceeding by
some circumstance like a segment lock or flow control. This kind of block does affect
throughput.
The first kind of block is represented by column ProcPendNetRead; the second kind is
represented by the remaining columns described here.
60

Statistics Columns
Column Name
Mode
Description
Data Type
ProcPendDBLock
track
Number of processes blocked pending database locks.
FLOAT
ProcPendFsgLock
track
Number of processes blocked pending an FSG lock.
FLOAT
ProcPendFsgRead
track
Number of processes blocked pending a File Segment (FSG) read

from disk.
FLOAT
ProcPendFsgWrite
track
Number of processes blocked pending an FSG write to disk.
FLOAT
ProcPendMemAlloc
track
Number of processes blocked pending memory allocations.
FLOAT
ProcPendMisc
track
Number of processes blocked pending miscellaneous events.
FLOAT
ProcPendMonitor
track
Number of processes blocked pending a user monitor.
FLOAT
ProcPendMonResume
track
Number of processes blocked pending a user monitor resume from

a yield.
FLOAT
ProcPendNetRead
track
Number of processes blocked pending non-step work, that is, the

number of processes blocked on any mailbox other than the work
mailbox.
FLOAT
Note: Non-step work is anticipated work the process spawned off

and is now waiting for some type of response from the spawned
process or processes. Non-step work is not unanticipated work such
as a new work request sent when a user initiates a request from the
host.
ProcPendNetThrottle
track
Number of processes blocked pending delivery of outstanding

outgoing messages.
FLOAT
ProcPendQnl
track
Number of processes blocked pending a TSKQNL lock.
FLOAT
ProcPendSegLock
track
Number of processes blocked pending a segment lock.
FLOAT
Process Block Count Columns

These columns identify how many times a process became blocked on which blocking type.
Average time blocked can be approximated by dividing the corresponding process pending
wait time column by the process block count column. For a list of these columns, see Process
Pending Wait Time Columns on page 62.
Column Name
Mode
Description
Data Type
ProcBlksFsgRead
count
Number of process blocks for an FSG read from disk.
FLOAT
ProcBlksFsgWrite
count
Number of process blocks for an FSG write to disk.
FLOAT
ProcBlksDBLock
count
Number of process blocks for database locks. The AMP Worker

Task can do other work while the lock is blocked.
FLOAT
ProcBlksFsgLock
count
Number of process blocks for an FSG lock.
FLOAT
ProcBlksTime
count
Number of process blocks waiting only for timer expiration.
FLOAT
61

Statistics Columns
Column Name
Mode
Description
Data Type
ProcBlksMemAlloc
count
Number of process blocks for memory allocations.
FLOAT
ProcBlksMisc
count
Number of process blocks for miscellaneous events.
FLOAT
ProcBlksMonitor
count
Number of process blocks for a user monitor.
FLOAT
ProcBlksMonResume
count
Number of process blocks for a user monitor resume from a yield.
FLOAT
ProcBlksMsgRead
count
Number of process blocks for non-step work.
FLOAT
ProcBlksNetThrottle
count
Number of process blocks for delivery of outstanding outgoing

messages.
FLOAT
ProcBlksQnl
count
Number of process blocks for a TSKQNL lock.
FLOAT
ProcBlksSegLock
count
Number of process blocks for a disk or task context (for example,

scratch, stack, and so on) segment lock.
FLOAT
Process Pending Wait Time Columns

These columns identify the how long the processes were blocked for each possible reason
listed below.
Note: Since this time is only accounted for when a blocked process leaves the blocked state, it
is possible for this statistic to be much larger than the amount of time available to all processes
in a single log period.
Column Name
Mode
Description
Data Type
ProcWaitDBLock
count
Total time in centiseconds processes were blocked pending database

locks.
FLOAT
ProcWaitFsgLock
count
Total time in centiseconds processes were blocked pending an FSG

lock.
FLOAT
ProcWaitFsgRead
count

read from disk.
FLOAT
ProcWaitFsgWrite
count

write to disk.
FLOAT
ProcWaitMemAlloc
count
Total time in centiseconds processes were blocked pending memory

allocations.
FLOAT
ProcWaitMisc
count
Total time in centiseconds processes were blocked pending

miscellaneous events.
FLOAT
ProcWaitMonitor
count
Total time in centiseconds processes were blocked pending a user

monitor.
FLOAT
ProcWaitMonResume
count
Total time in centiseconds processes were blocked pending a user

monitor resume from a yield.
FLOAT
ProcWaitMsgRead
count
Total time in centiseconds processes were blocked pending nonstep work.
FLOAT
62

Statistics Columns
Column Name
Mode
Description
Data Type
ProcWaitNetThrottle
count
Total time in centiseconds processes were blocked pending delivery

of outstanding outgoing messages.
FLOAT
ProcWaitPageRead
count
Total time in centiseconds processes were blocked pending a page

read from disk.
FLOAT
ProcWaitTime
count
Total time in centiseconds processes were blocked pending some

amount of elapsed time only.
FLOAT
ProcWaitQnl
count
Total time in centiseconds processes were blocked pending a

TSKQNL lock.
FLOAT
ProcWaitSegLock
count
Total time in centiseconds processes were blocked pending a disk or

task context (for example, scratch, stack, and so on) segment lock.
FLOAT
CPU Utilization Columns

The CPU utilization columns count all CPU activities, including activities performed for
virtual processors, and represent the sum of all CPUs on the node.
To obtain the average node CPU value for each column, CPU (Idle, IOWait, UServ, UExec),
divide the column data by the number of CPUs per node (the value in the NCPUs column)
and the number of centiseconds (CentiSecs column) in the logging interval.
CPU idle time = CPUIdle + CPUIoWait
CPU busy time = CPUUServ + CPUUExec
The NodeNormFactor is the per node normalization factor. This is related to the NodeType
value reported in this resource usage table. The normalization factor modifies the reported
CPU times to the equivalent time of a specified virtual processor. This does not add up to the
reported CPU time.
To calculate the non-normalized total CPU time, use the following formula:
CentiSecs x NCPUs = CPUIdle + CPUIoWait + CPUUServ + CPUUExec
Note: The CPU time returned in centiseconds is more accurate than those returned in
seconds.
Column Name
Mode
Description
Data Type
CPUIdle
count
Time in centiseconds CPUs are idle and not waiting for I/O.
FLOAT
CPUIoWait
count
Time in centiseconds CPUs are idle and waiting for I/O

completion.
FLOAT
Note: This time represents another variety of Idle, since a

CPU is only in this state if there are no processes eligible for
execution. If there was a process available, the CPU would
be immediately dispatched for that process.
63

Statistics Columns
Column Name
Mode
Description
Data Type
CPUUExec
count
Time in centiseconds CPUs are busy executing user

execution code, that is, time spent in a user state on behalf of
a process.
FLOAT
CPUUExec reports the CPU time not used in the system

call, or in the kernel.
CPUUServ
count
Time in centiseconds CPUs are busy executing user service

code, that is, privileged work performing system services on
behalf of user execution processes which do not have root
access.
FLOAT
CPUUServ reports if a task executing a step used CPU while

in the kernel.
TASM Columns
AMP Worker Task Columns
These columns report statistics about the AMP Worker Tasks.
For more information about the ResUsageSawt table and columns, see Chapter 7:
ResUsageSawt Table.
Column Name
Mode
Description
Data Type
AwtFlowControlled
track
Number of AMPs currently in flow control on the work input

mailbox.
FLOAT
AwtFlowCtlCnt
count
Number of times this log period that the node entered the flow
control state from a non-flow controlled state.
FLOAT
AwtInuse
track
Number of AMP Worker Tasks currently in use for this node.
FLOAT
AwtInuseMax
max
Peak number of AWTs for any one of the AMPs on the node.
FLOAT
Note: AwtInuseMax is not the peak number in use on the node for
all AMPs at any one point in time.
AwtInuseMax represents the largest number of AWTs in use on any
single AMP.
Priority Scheduler Columns

These columns provide data specific to the Priority Scheduler.
For more information about the ResUsageSps table and columns, see Chapter 11:
ResUsageSps.
64

Statistics Columns
Column Name
Mode
Description
Data Type
PSNumRequests
count
Number of work requests received for all Performance Groups on

this node.
FLOAT
PSQWaitTime
count
Time in centiseconds that work requests waited on an input queue

before being serviced.
FLOAT
To get an approximate average QWaitTime per request during this

period, divide QWaitTime by NumRequests.
PSServiceTime
count
Time in centiseconds that work requests required for service.
FLOAT
To get an approximate average ServiceTime per request during this

period, divide ServiceTime by NumRequests.
Teradata VS Columns
These columns identify pdisk I/O statistics that are reported by the Node Agent.
Column Name
Mode
Description
Data Type
TvsReadCnt
count
Number of logical device reads.
FLOAT
The TvsReadCnt column is a summary of the ReadCnt column in

the SPDSK table.
TvsReadRespMax
max
Maximum read response time value during the reporting period.
FLOAT
TvsReadRespTot
count
Total of individual read response time in centiseconds.
FLOAT
The TvsReadRespTot column is a summary of the ReadRespTot

column in the SPDSK table.
TvsWriteCnt
count
Number of logical device writes.
FLOAT
The TvsWriteCnt column is a summary of the WriteCnt column

in the SPDSK table.
TvsWriteRespMax
max
Maximum response time of logical device writes during the

reporting period.
FLOAT
TvsWriteRespTot
count
Total of individual write response time in centiseconds.
FLOAT
TvsWriteRespTot is a summary of the WriteRespTot column in

the SPDSK table.
User Command Columns

These columns summarize the type of statements given to Teradata Database by the user.
For more information, see Chapter 8: ResUsageShst Table.
Column Name
Mode
Description
Data Type
CmdDDLStmts
count
Number of alter, modify, drop, create, replace, grant or revoke

commands.
FLOAT
65

Spare Columns
Column Name
Mode
Description
Data Type
CmdDeleteStmts
count
Number of delete commands.
FLOAT
CmdInsertStmts
count
Number of insert commands.
FLOAT
CmdSelectStmts
count
Number of select commands.
FLOAT
CmdUpdateStmts
count
Number of update commands.
FLOAT
CmdUtilityStmts
count
Number of utility commands.
FLOAT
CmdOtherStmts
count
Number of other commands.
FLOAT
User Command Arrival and Departure Columns

These columns summarize the arrival and departure of user statements.
For more information, see Chapter 8: ResUsageShst Table.
Column Name
Mode
Description
Data Type
CmdStmtErrors
count
Number of statements that departed in error.
FLOAT
CmdStmtFailures
count
Number of statements that departed in failure or were aborted.
FLOAT
CmdStmtSuccesses
count
Number of statements that departed normally.
FLOAT
Column Name
Mode
Description
Data Type
Reserved
n/a
CHAR (2)
Reserved Column
Spare Columns
The ResUsageSpma table spare fields are named Spare00 through Spare19, and SpareInt.
The following table describes the Spare fields currently being used.
66

Spare Columns
Column Name
Description
Spare02
Number of I/Os completed on data blocks that were marked for contiguous
write.
You can get the average number of data blocks combined into a single I/O
by using the calculation:
FileContigWBlocks / FileContigWIOs
Note: This field will be converted to the FileContigWIOs column in
Teradata Database 15.0. You can access resource usage data for this field
using the FileContigWIOs column name in ResSpmaView view. For details,
see ResSpmaView on page 173.
Spare03
Number of data blocks marked for contiguous write.

Data blocks created using the contiguous write scheme are accumulated in
the cache until the optimal I/O size, the end of the cylinder, or the end of
the data created has been reached. At such a point, the blocks will be
written with a single I/O and removed from the cache.
Note: This field will be converted to the FileContigWBlocks column in
using the FileContigWBlocks column name in the ResSpmaView view. For
details, see ResSpmaView on page 173.
Spare04
Total size, in KB, of the data blocks that were marked for contiguous write.
You can get the average size of the writes that were candidates for being
combined into one single I/O by using the calculation:
FileContigWKB /FileContigWBlocks
Note: This field will be converted to the FileContigWKB column in
using the FileContigWKB column name in the ResSpmaView view. For
Spare05
Number of times the TDAT Control Group (cgroup) query is throttled due
to the WM CPU COD. A cgroup provides a mechanism for aggregating or
partitioning sets of tasks, and all its future children, into hierarchical
groups with a specialized behavior.
Note: This field will be converted to the CpuThrottleCount column in
using the CpuThrottleCount column name in the ResSpmaView view. For
Spare06
Time, in centiseconds, that the TDAT cgroup query is throttled due to the
WM CPU COD.
The data is reported to RSS in nanoseconds. During the data gathering, the
RSS converts the data to centiseconds.
Note: This field will be converted to the CpuThrottleTime column in
using the CpuThrottleTime column name in the ResSpmaView view. For
67

Spare Columns
Column Name
Description
Spare07
Sum of the full potential input or output token allocations from all devices
attached to the node.
The input or output token allocations are not limited by the IO_COD
setting.
The value is computed by summing the cod_full_potential_iota field for
each device in the /proc/tdmeter/disk_cod_stats file.
Note: This field will be converted to the FullPotentialIota column in
using the FullPotentialIota column name in the ResSpmaView view. For
Spare08
Sum of potential input or output token allocations from all devices

attached to the node.
The input or output token allocations are accumulated only when an I/O is
pending on a device and are limited by the IO_COD setting.
The value is computed by summing the cod_allocated_iota field for each
device in the /proc/tdmeter/disk_cod_stats file.
Note: This field will be converted to the CodPotentialIota column in
using the CodPotentialIota column name in the ResSpmaView view. For
Spare09
Sum of used input or output token allocations from all devices attached to
the node.
The value is computed by summing the cod_used_iota field for each device
in the /proc/tdmeter/disk_cod_stats file.
Note: This field will be converted to the UsedIota column in Teradata
Database 15.0. You can access resource usage data for this field using the
UsedIota column name in the ResSpmaView view. For details, see
ResSpmaView on page 173.
Spare10
Current size of the VH cache in KB. This field is populated by the FSG
subsytem.
Note: This field will be converted to the VHCacheKB column in Teradata
VHCacheKB column name in the ResSpmaView view. For details, see
ResSpmaView on page 173.
For more information about VH cache, see Glossary on page 261.
Spare11
WM CPU COD value in one tenths of a percent. For example, a value of

500 represents a WM CPU COD value of 50.0%.
using the WM_CPU_COD column name in the ResSpmaView view. For
Note: WM CPU COD is not supported on SLES 10. Its value is set to 1000
on SLES 10.
68

Spare Columns
Column Name
Description
Spare12
WM I/O COD value in whole percent. For example, a value of 50 represents

a WM I/O COD value of 50.0%.
The value is set to 100 if the WM I/O COD is disabled.
Note: This field will be converted to the WM_IO_COD column in
using the WM_IO_COD column name in the ResSpmaView view. For
Note: WM I/O COD is not supported on SLES 10. Its value is set to 100 on
SLES 10.
SpareInt
PM I/O COD value in whole percent values for the entire system. For
example, a value of 50 represents a PM I/O COD value of 50%.
The value is set to 100 if the PM I/O COD is disabled.
Note: This field will be converted to the PM_IO_COD column in Teradata
Database 15.0. You can access resource usage data for the SpareInt field
using the PM_IO_COD column name in the ResSpmaView view. For
Related Topics
For more information about the different type of data fields, see About the Mode Column
on page 42.
69

Spare Columns
70
CHAPTER 7
ResUsageSawt Table
The ResUsageSawt table contains:
The current and maximum number of AMP Worker Tasks in use by work type
Flow control information
You can use the ResUsageSawt table to:
Report the pattern in AMP Worker Task usage for standard or expedited work types.
Monitor the length of each message queue of the AMP, the queue which holds work
messages from the Dispatcher that are waiting to get an AMP Worker Task.
Identify if one or more AMPs have entered the state of flow control, and how often, during
the logging interval.
If you enable table logging, the data is written to the database once for each log period.
To consolidate and summarize the total number of rows written to the database, you can
enable Summary Mode. For details, see Summary Mode on page 76.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT

been logged. For more information, see the Secs and NominalSecs
columns.
71
Chapter 7: ResUsageSawt Table

Column Name
Mode
Description
Data Type
NodeID
n/a
Node ID on which the vproc resides. The Node ID is formatted as

CCC-MM, where CCC denotes the three-digit cabinet number and
MM denotes the two-digit chassis number of the node. For
example, a node in chassis 9 of cabinet 3 has a node ID of '003-09'.
INTEGER


Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
NodeType
n/a
CHAR (8)
VprId
n/a
Identifies the vproc number. All Vprocs in this table are AMPS so
there is no VprType column provided. In Summary Mode, this
column is -1.
INTEGER
Secs
n/a

cases:
SMALLINT

CentiSecs
n/a

INTEGER
NominalSecs
n/a
SMALLINT
SummaryFlag
n/a
Summarization status of this row. If the value is 'N,' the row is a

non-summary row. If the value is 'S,' the row is a summary row.
CHAR (1)
Active
max

FLOAT

usage tables.
72

Statistics Columns
Column Name
Mode
Description
Data Type
TheTimestamp
n/a
BIGINT
CODFactor
n/a
PM CPU COD value in one tenths of a percent. For example, a

value of 500 represents a PM CPU COD value of 50.0%.
SMALLINT
Statistics Columns
TASM Columns
These columns report statistics about the AMP Worker Tasks.
Column Name
Mode
Description
Data Type
Available
track
Number of unreserved AMP Worker Tasks from the pool that are
not being used at the end of the interval.
FLOAT
For example, if 12 of the normally 62 unreserved AMP Worker

Tasks are removed from the pool by reserving them for expedited
work, there could at most be 50 unreserved AMP Worker Tasks
available. If in this log period, 10 unreserved AMP Worker Tasks are
taken from the pool to service 10 queries that are still executing,
there would be only 40 available at the end of the log period.
AvailableMin
min
Minimum number of unreserved AMP Worker Tasks available in

the pool for each AMP for the logged period.
FLOAT
For example, a zero value means there were no unreserved AMP

Worker Tasks available in the pool at some point during the
reporting period.
AwtLimit
track
Current setting for AMP Worker Task (for example, 80, 100, or so
on) in the DBS Control MaxLoadAWT field. For more information,
see MaxLoadAWT in Utilities.
FLOAT
FlowControlled
track
Specifies if an AMP is in flow control. If the value is non-zero, the

AMP is in flow control.
FLOAT
Resource usage indicates flow control on work mailbox name 2-11

only. Work mailbox name 2-11 is the incoming mailbox for AMPs,
where 2 is the mailbox number and 11 is the AMP Worker Task
partition.
However, resource usage indicates flow control on work mailbox
name 2-11 for all work types. For example, if any work type is in
flow control, the system is in flow control.
FlowCtlCnt
count
Number of times during the log period that the system entered the
flow control state from a non-flow controlled state.
FLOAT
73

Statistics Columns
Column Name
Mode
Description
Data Type
FlowCtlTime
count
Total time, in milliseconds, that an AMP is in flow control.
FLOAT
InuseMax
max
Maximum number of AMP Worker Tasks in use at any one time

during the log period.
FLOAT
MailBoxDepth
track
Current depth of the AMP work mailbox at the end of the period.
FLOAT
WorkTypeInuse00 WorkTypeInuse15
track
Current number of AMP Worker Tasks in use during the log period
for each work type for the VprId vproc.
FLOAT
Note: The WorkTypeInuse04 column value is always at least 1 due

to the internal database design for the Control AMP rows in the
ResUsageSawt table.
WorkTypeMax00 WorkTypeMax15
max
Maximum number of AMP Worker Tasks in use at one time during

the log period for each work type for the VprId vproc.
FLOAT
In Summary Mode, the WorkTypeMax column values are the Max

of the values for all the AMPS.
Work Type Descriptions

The WorkTypeInuse and WorkTypeMax array data columns above each contain 16 Work Type
entries that are described here. For example, WorktypeInuse00 contains the number of in use
AMP Worker Tasks that are of Work Type MSGWORKNEW, and WorktypeInuse01 contains
the values for MSGWORKONE.
These columns allow you to monitor the usage of the AMP Worker Tasks of each work type.
This can be used to determine:
If the usage is close to the maximum values defined.
What type of work the AMP Worker Tasks are doing.
Characteristics of the system during skew conditions or when there are AMP Worker Task
shortages.
Use the tdntune utility to determine the settings for Flow Control. For information on
Expedited Allocation Groups, see Priority Scheduler (schmon) chapter of Utilities.
Column Name
Mode
Description
Data Type
MSGWORKNEW
n/a
Used for new work requests. This work type has the lowest number,
which means it is queued last. It also has the effect of honoring
secondary requests needed to complete existing work items before
any new ones are started.
n/a
A zero value is used for new work items.
74

Statistics Columns
Column Name
Mode
Description
Data Type
MSGWORKONE
n/a
First level secondary work items. Numbered work types are used for
secondary work items. For example, work type one
(MSGWORKONE) is used for secondary work requests spawned by
new work items; work type two (MSGWORKTWO) requests are
spawned from work type one requests and queued for delivery
before work type one requests; and so on. Each numbered work
type is queued for delivery just before the one from which it is
spawned.
n/a
MSGWORKTWO
n/a
Second level secondary work items.
n/a
MSGWORKTHREE
n/a
Special types of database work.
n/a
MSGWORKFOUR
n/a
Start System Recover.
n/a
MSGWORKFIVE
n/a
Reports new work for utilities, such as FastLoad, MultiLoad, and

FastExport if utilities are configured to use a separate pool of work
types.
n/a
Note: This column is not normally used and the MSGWORKNEW,

MSGWORKONE, and MSGWORKTWO columns report work
requests for utilities.
MSGWORKSIX
n/a
First level secondary work spawned work for utilities such as

FastLoad, MultiLoad, and FastExport. If the utilities are not
configured to use a separate pool of work types, they use
MSGWORKNEW, MSGWORKONE, and MSGWORKTWO.
n/a
MSGWORKSEVEN
n/a
Second level secondary work for utilities such as FastLoad,

MultiLoad, and FastExport. If the utilities are not configured to use
a separate pool of work types, they use MSGWORKNEW,
MSGWORKONE, and MSGWORKTWO.
n/a
MSGWORKEIGHT
n/a
New work for Expedited Allocation Groups.
n/a
MSGWORKNINE
n/a
First level spawned work for Expedited Allocation Groups.
n/a
MSGWORKTEN
n/a
Second level spawned work for Expedited Allocation Groups.
n/a
MSGWORKELEVEN
n/a
Not used.
n/a
MSGWORKABORT
n/a
Used for transaction abort requests. This work type has a higher
value than the numbered work types so that abort requests are
honored before beginning any additional work item for the
transactions being aborted.
n/a
The array number for MSGWORKABORT is 12.

MSGWORKSPAWN
n/a
Used for spawned abort requests and is delivered before normal

aborts.
n/a
The array number for MSGWORKSPAWN is 13.

MSGWORKNORMAL
n/a
Used for messages that do not fall within the standard work type
hierarchy. This work type is delivered before any of the work items
described above.
n/a
The array number for MSGWORKNORMAL is 14.
75

Summary Mode
Column Name
Mode
Description
Data Type
MSGWORKCONTROL
n/a
Used for system control messages. These are delivered before any
other kind of message.
n/a
The array number for MSGWORKCONTROL is 15.
Reserved Column
Column Name
Mode
Description
Data Type
Reserved
n/a
CHAR (1)
Summary Mode
When Summary Mode is active for the ResUsageSawt table, one row is written to the database
for each node in the system for each log interval. The AMP Worker Task data will be combined
for all the AMP vprocs on the node.
that row.
'S'
in Summary Mode.
'N'
normally.
Spare Columns
The ResUsageSawt table spare fields are named Spare00 through Spare19, and SpareInt.
76

Spare Columns
Column Name
Description
Spare00
WM CPU COD value in one tenths of a percent. For example, a value

of 500 represents a WM CPU COD value of 50.0%.
field using the WM_CPU_COD column name in the ResSawtView
view. For details, see ResSawtView on page 168
1000 on SLES 10.
Related Topics
77

Spare Columns
78
CHAPTER 8
ResUsageShst Table
The ResUsageShst table:
Contains resource usage information specific to the host channels and TCP/IP networks
communicating with Teradata Database.
Includes resource usage data for system-wide, host information.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT

columns.
NodeID
n/a

INTEGER


Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
79
Chapter 8: ResUsageShst Table

Column Name
Mode
Description
Data Type
NodeType
n/a
CHAR(8)
VprId
n/a
Identifies the vproc number.
INTEGER
For channel-connected hosts, the VprId value for:

IBMECS is the second 4 bytes from the IP address of the ECS
node (see ResShstView on page 170).
IBMMUX is zero.
For TCP/IP network-connected hosts, the VprId value for:
NETWORK is the Gateway vproc ID.
IBMNET is zero.
In Summary Mode, VprId is -1.
HstId
n/a
Identifies the host. The HstId value for:
INTEGER
IBMECS is cua/lcu/lchan/laddr, where:

cua = Channel unit address
lcu = Logical control unit address
lchan = Logical channel
laddr = Link address
IBMNET is the host group ID.
NETWORK and IBMMUX is BBMMPPHHH, where:
BB = Bus
MM = Module Number (or chassis number)
Note: The chassis number is always zero for networkconnected hosts.
PP = Port
HHH = Three digit host group ID
Note: Each of the fields above get two or three decimal digits of
the resulting nine digit value.
In Summary Mode, HstId is always 0.
HstType
n/a
Type of host connection:
CHAR(8)
IBMECS (Part TCP/IP and channel hardware connection)

IBMMUX (Channel hardware connection)
IBMNET (TCP/IP end-to-end, also called pure network,
connection)
NETWORK (Gateway channel connection)
80

Column Name
Mode
Description
Data Type
IPaddr
n/a
Identifies the IP address.
INTEGER
For channel-connected hosts, the IPaddr value for:

IBMECS is the first 4 bytes from the IP address of the ECS node.
For more information, see the HostId column.
IBMMUX is zero.
For TCP/IP network-connected hosts, the IPaddr value for:
NETWORK is zero.
IBMNET is zero.
Note: The IP address values (that is, the IPaddr and VprId
columns) for the IBMNET connection are displayed as a 32bit
integer, but to evaluate them as an IP address they need to be
considered one byte at a time. You can do this by using the SQL
commands listed in the ResShstView view. For details, see
ResShstView on page 170.
Secs
n/a

This value is useful for normalizing the statistics contained in this
row, for example, to a per-second measurement.
SMALLINT
CentiSecs
n/a

INTEGER
NominalSecs
n/a
SMALLINT
SummaryFlag
n/a

CHAR (1)
Active
max

FLOAT

usage tables.
TheTimestamp
n/a
BIGINT
CODFactor
n/a

SMALLINT
81

Statistics Columns
Statistics Columns
Host Controller Columns
Channel Management Columns
These columns identify overhead of channel management.
Column Name
Mode
Description
Data Type
HostReadFails
count
Number of failures transmitting from the host.
FLOAT
Note: This column is populated for Teradata Channel software

(TCHN) only.
HostWriteFails
count
Number of failures transmitting to the host.
FLOAT
Note: This column is populated for TCHN only.
Channel Traffic Columns

These columns identify the traffic between the host and the node in three levels of granularity:
Blocks
Messages
KB
Blocks are made up of some amount of variable sized messages. ReadKB and WriteKB identify
the KB involved in the traffic.
For information on the ResUsageSpma table channel and TCP/IP network traffic columns, see
Host Controller Channel and TCP/IP Network Traffic Columns on page 55.
Column Name
Mode
Description
Data Type
HostBlockReads
count
Number of blocks read in from the host.
FLOAT
HostBlockWrites
count
Number of blocks written out to the host.
FLOAT
HostMessageReads
count
Number of messages read in from the host.
FLOAT
HostMessageWrites
count
Number of messages written out to the host.
FLOAT
HostReadKB
count
KB transferred in from the host.
FLOAT
HostWriteKB
count
KB transferred out to the host.
FLOAT
User Command Columns

These columns identify the type of commands given to Teradata Database by the user. Three
levels of granularity are given:
82
Transactions. These columns consist of one or more requests.
Requests. These consist of one or more statements.


Statistics Columns
Statements. These columns are subdivided into the various statement types.
Column Name
Mode
Description
Data Type
CmdAlterStmts
count
Number of alter, modify, or drop statement commands.
FLOAT
CmdArchUtilityStmts
count
Number of archival utility commands (for example, restore, archive

and recovery).
FLOAT
CmdCreateStmts
count
Number of create or replace statement commands.
FLOAT
CmdDeleteStmts
count
Number of delete commands.
FLOAT
CmdGrantStmts
count
Number of grant or revoke commands.
FLOAT
CmdInsertStmts
count
Number of insert commands.
FLOAT
CmdLoadUtilityStmts
count
Number of FastLoad and MultiLoad utility commands. (Tpump

commands cannot be distinguished, and are therefore counted by
the INSERT, UPDATE and DELETE statements).
FLOAT
CmdOtherStmts
count
Number of other commands.
FLOAT
CmdRequests
count
Number of request commands.
FLOAT
CmdSelectStmts
count
Number of select commands.
FLOAT
CmdTransactions
count
Number of transaction commands.
FLOAT
CmdUpdateStmts
count
Number of update commands.
FLOAT
User Command Arrival and Departure Columns

These columns identify the arrival and departure times and status of user commands.
Column Name
Mode
Description
Data Type
CmdStmtErrors
count
Number of statements that departed in error.
FLOAT
CmdStmtFailures
count
Number of statements that departed in failure or abortion.
FLOAT
CmdStmtSuccesses
count
Number of statements that departed normally.
FLOAT
Column Name
Mode
Description
Data Type
Reserved
n/a
CHAR (5)
Reserved Column
83

Summary Mode
Summary Mode
When Summary Mode is active for the ResUsageShst table, one row is written to the database
for each type of host (network or channel-connected) on each node in the system,
summarizing the hosts of that type on that node.
that row.
'S'
in Summary Mode.
'N'
normally.
Spare Columns
The ResUsageShst table spare fields are named Spare00 through Spare09, and SpareInt.
Column Name
Description
Spare00
WM CPU COD value in one tenths of a percent. For example, a value of 500
represents a WM CPU COD value of 50.0%.
Note: This field will be converted to the WM_CPU_COD column in Teradata
WM_CPU_COD column name in the ResShstView view. For details, see
ResShstView on page 170
Note: WM CPU COD is not supported on SLES 10. Its value is set to 1000 on
SLES 10.
Related Topics
84
CHAPTER 9
ResUsageSldv Table
The ResUsageSldv table contains resource usage information for system-wide, logical device
information. Statistics from this table are collected from the storage devices.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT

columns.
NodeID
n/a

INTEGER


These columns provide statistics on current logging characteristics and storage device
information.
Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
NodeType
n/a
CHAR(8)
85
Chapter 9: ResUsageSldv Table

Column Name
Mode
Description
Data Type
CtlId
n/a
Controller number.
INTEGER
The value is the decimal equivalent of the three digit controller ID

in the LdvId (the HHC digits).
If the controller information is not available, its value is set to -1.
In Summary Mode, the CtlId is set to -1.
LdvId
n/a
Storage device in the Bus System where it resides.
BYTE(4)
If the device address information is available, the LdvId is derived

from the Host, Channel, Id, and Lun information of the device
(00HHCTLL).
If the device address information is not available, this column
contains the null ID which is 0xFFFFFFFF.
In Summary Mode, the value is set to 0xFFFFFFFF.
LdvType
n/a
Type of logical device. The value is either DISK for database disk or
SDSK for system disk.
CHAR(4)
Secs
n/a

cases:
SMALLINT

CentiSecs
n/a

INTEGER
NominalSecs
n/a
SMALLINT
SummaryFlag
n/a

CHAR (1)
Active
max

FLOAT

usage tables.
TheTimestamp
n/a
BIGINT
86

Statistics Columns
Column Name
Mode
Description
Data Type
CODFactor
n/a

SMALLINT
The value is set to 1000 if PM CPU COD is disabled.
Statistics Columns
Logical Device Columns
These columns identify individual logical device activities for storage components connected
through the buses.
The storage device statistics are calculated only on what can be derived from statistics
collected by the operating system since the disk array controllers do not provide us with any
useful data for resource usage.
The logical device columns are grouped into several subcategories as shown below.
Input and Output Traffic Columns
These columns represent the number and amount, in KB, of data read from or written to the
logical device.
Column Name
Mode
Description
Data Type
LdvReadKB
count
Number of KB read from the logical device.
FLOAT
LdvReads
count
Number of reads issued.
FLOAT
LdvWriteKB
count
Number of KB written to the logical device.
FLOAT
LdvWrites
count
Number of writes issued.
FLOAT
Response Time Columns

These columns represent the response time to requests given to the logical device.
Column Name
Mode
Description
Data Type
LdvReadRespMax
max
WM CPU COD value in one tenths of a percent. For example, a

value of 500 represents a WM CPU COD value of 50.0%.
FLOAT

Note: The LdvReadRespMax field is being used as a spare field to
represent the WM CPU COD and will be converted to the
WM_CPU_COD column in Teradata Database 15.0. For details, see
ResSldvView on page 171.
Note: WM CPU COD is not supported on SLES 10. Its value is set
to 1000 on SLES 10.
87

Statistics Columns
Column Name
Mode
Description
Data Type
LdvWriteRespMax
max
WM I/O COD value in whole percent. For example, a value of 50

represents a WM I/O COD value of 50.0%.
FLOAT

Note: The LdvWriteRespMax field is being used as a spare field to
represent the WM I/O COD and will be converted to the
WM_IO_COD column in Teradata Database 15.0. For details, see
Note: WM I/O COD is not supported on SLES 10. Its value is set to
100 on SLES 10.
LdvReadRespTot
count
Total of individual read response times in centiseconds.
FLOAT
LdvWriteRespTot
count
Total of individual write response times in centiseconds.
FLOAT
ReadActiveTotal
count
Total of read I/O active time, in centiseconds.
FLOAT

WriteActiveTotal
count
Total of write I/O active time, in centiseconds.
FLOAT
Outstanding Requests Columns

These columns represent the number of outstanding operation requests and the amount of
time with outstanding requests for the logical device.
Column Name
Mode
Description
Data Type
QReadLength
track
Current number of read operations in queue.
FLOAT

QWriteLength
track
Current number of write operations in queue.
FLOAT

LdvOutReqDiv
count
Average value calculated by the ResSldvView view as shown below.
FLOAT
LdvOutReqAvg = LdvOutReqSum / LdvOutReqDiv

LdvOutReqSum
count
Time-weighted total of samples in centiseconds spent doing I/O.

LdvOutReqDiv contains the divisor, which is the delta time in
centiseconds for the sample.
FLOAT
LdvOutReqTime
count
Total time in centiseconds with (any) outstanding requests. The

values in this column should be less than or equal to the reported
logging period.
FLOAT
88

Reserved Column
Reserved Column
Column Name
Mode
Description
Data Type
Reserved
n/a
Type of storage device.
CHAR (1)
If the value is 'S', this indicates the storage device is a database

disk and a solid state storage device (SSD).
If the value is 'H,' this indicates the storage device is a hard disk
drive (HDD).
In Summary mode, the Reserved column is set to an en dash
character symbol (-), where '-' indicates information about the
storage device is not available.
Note: The Reserved field will be converted to the LdvKind
column in Teradata Database 15.0. If using the ResSldvView
view, the resource usage data can be accessed using the LdvKind
column name instead of the Reserved name. For details, see
Summary Mode
When Summary Mode is active for the ResUsageSldv table, the following rows are written to
the database for each node in the system for each log interval:
One row summarizes the system logical devices
One row summarizes Teradata Database logical devices
that row.
'S'
in Summary Mode.
'N'
normally.
Spare Columns
The ResUsageSldv table spare fields are named Spare00 through Spare04, and SpareInt.
89

Spare Columns
Column Name
Description
Spare00
Major number of the device.

In Summary mode, this field contains the max of the major numbers of the
devices being summed. It is not paired with the minor number reported in
the Spare01 column.
Note: This field will be converted to the Major column in Teradata Database
15.0. You can access the resource usage data for this field using the Major
column name in the ResSldvView view. For details, see ResSldvView on
page 171.
Spare01
Minor number of the device.

In Summary mode, this field contains the max of the minor numbers of the
devices being summed. It is not paired with the major number reported in
the Spare00 column.
Note: This field will be converted to the Minor column in Teradata Database
15.0. You can access the resource usage data for this field using the Minor
column name in the ResSldvView view. For details, see ResSldvView on
page 171.
Spare02
Full potential input or output token allocations for a device.

The total input or output token allocations are not limited by the IO_COD
field setting.
Teradata Database 15.0. You can access the resource usage data for this field
using the FullPotentialIota column name in the ResSldvView view. For
details, see ResSldvView on page 171.
Spare03
Potential input or output token allocations for a device.

The potential input or output token allocations are accumulated only when
an I/O is pending on a device and are limited by the IO_COD column
setting.
Teradata Database 15.0. You can access the resource usage data for this field
using the CodPotentialIota column in the ResSldvView view. For details, see
Spare04
Used input or output token allocations for a device.

Database 15.0. You can access the resource usage data for this field using the
UsedIota column name in the ResSldvView view. For details, see
SpareInt
PM I/O COD value in whole percent. For example, a value of 50 represents a

PM I/OCOD value of 50.0%.
Note: This field will be converted to the PM_IO_COD column in Teradata
PM_IO_COD column name in the ResSldvView view. For details, see
90

Spare Columns
Related Topics
91

Spare Columns
92
CHAPTER 10
ResUsageSpdsk Table
The ResUsageSpdsk table:
Provides pdisk level statistics.
Includes resource usage logs on cylinder I/O, allocation, and migration.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT

field will still contain the time value when the entry should have
columns.
NodeID
n/a
Node ID on which the pdisk is connected. The Node ID is

formatted as CCC-MM, where CCC denotes the three-digit cabinet
number and MM denotes the two-digit chassis number of the node.
For example, a node in chassis 9 of cabinet 3 has a node ID of '00309'.
INTEGER


Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
93
Chapter 10: ResUsageSpdsk Table

Column Name
Mode
Description
Data Type
PdiskGlobalId
n/a
Identifies the pdisk in the system. Each pdisk in the system has a
global ID which uniquely identifies the pdisk in the system. If a
pdisk is connected to the nodes in a clique, all the nodes in that
clique see the same pdisk global ID associated with that pdisk.
INTEGER
In Summary Mode, the pdisk global ID is -1.

PdiskType
n/a
Type of pdisk. The pdisk can be one of the following:
CHAR(4)
DISK: A storage device.

FILE: A file.
SSD: A solid-state device.
PdiskDeviceId
n/a
Identifies the local pdisk device.
BYTE(4)
For both DISK and SSD pdisks, the pdisk device ID can be a
major/minor number. The major number bit positions are
20-31 and the minor number is in bits 0-19. The format is
similar to the one shown below.
(MMMM MMMM MMMM mmmm mmmm mmmm mmmm mmmm)
For FILE pdisk, the pdisk device ID is 0xFFFFFFFF.
In Summary Mode, the pdisk device ID is 0xFFFFFFFF.
NodeType
n/a
CHAR(8)
Secs
n/a

cases:
SMALLINT

CentiSecs
n/a
Number of centiseconds in the logging period. This field is useful

when performing data calculations with small elapsed times where
the difference between centisecond-based data and whole seconds
results in a percentage error.
INTEGER
NominalSecs
n/a
SMALLINT
SummaryFlag
n/a

CHAR (1)
Active
max

FLOAT

usage tables.
94

Statistics Columns
Column Name
Mode
Description
Data Type
TheTimestamp
n/a
BIGINT
CODFactor
n/a
PM CPU COD value in one tenths of a percent.For example, a value

of 500 represents a PM CPU COD value of 50.0%.
SMALLINT
Statistics Columns
Teradata VS Columns
Allocation Columns
These columns identify the allocation statistics reported by the Allocator process.
Column Name
Mode
Description
Data Type
ExtAllocHot
count
Number of hot allocations made in the current log period. A hot

allocation is an allocation whose estimated temperature falls within
the pre-defined hot temperature range. Each allocation is for a
cylinder size worth of data. The cylinder resides in some disk storage
location and holds some data. Temperature is the frequency of access
to the data by I/O independent of where the data resides.
FLOAT
ExtAllocNonPacing
count
Number of non-pacing allocations made in the current log period. A

non-pacing allocation is an allocation whose data access affects
neither system performance nor individual query performance.
FLOAT
ExtAllocStatic
count
Number of static allocations made in the current log period. A static

allocation is an allocation whose requested temperature is used and
the measured temperature is ignored during migration.
FLOAT
ExtAllocSystemPacing
count
Number of system pacing allocations made in the current log period.

A system pacing allocation is an allocation whose data access affects
system performance.
FLOAT
ExtAllocTotal
count
Total number of allocations made in the current log period. A number

of computations can be derived. For example:
FLOAT
Cold Allocation = ExtAllocTotal -ExtAllocHot -ExtAllocWarm

QueryPacing Allocation = ExtAllocTotal -ExtAllocNonPacing ExtAllocSystemPacing
Dynamic Allocation = ExtAllocTotal - ExtAllocStatic
ExtAllocWarm
count
Number of warm allocations made in the current log period. A warm

the pre-defined warm temperature range. Each allocation is for a
cylinder size worth of data. The cylinder resides in some disk storage
location and holds some data. Temperature is the frequency of access
to the data by I/O independent of where the data resides.
FLOAT
95

Statistics Columns
I/O Statistics Columns

Note: You can populate the ResUsageSpdsk table I/O statistics columns without having a
Teradata VS license.
These columns identify the I/O statistics reported by the Extent Driver.
Column Name
Mode
Description
Data Type
ConcurrentMax
count
Maximum number of concurrent I/O requests.
FLOAT
ConcurrentReadMax
max
Maximum number of concurrent read I/O requests.
FLOAT
ConcurrentWriteMax
max
Maximum number of concurrent write I/O requests.
FLOAT
MigrationBlockedIos
count
Number of inputs and outputs that are blocked due to migration

request.
FLOAT
ReadCnt
count
FLOAT
ReadKB
count
FLOAT
ReadRespMax
max
Maximum of individual read response time in centiseconds.
FLOAT
ReadRespSq
count
Total of squares of the individual read response time in

centiseconds.
FLOAT
ReadRespTot
count
FLOAT
WriteCnt
count
FLOAT
WriteKB
count
FLOAT
WriteRespMax
max
Maximum of individual write response time in centiseconds.
FLOAT
WriteRespSq
count
Total of squares of the individual write response time in

centiseconds.
FLOAT
WriteRespTot
count
FLOAT
Migration Columns
Note: The ResUsageSpdsk table migration columns are populated only in Teradata VS. For
more information, see Teradata Virtual Storage.
The migration columns identify the number of cylinders that migrated to a different location
on a device as well as the time, in centiseconds, of all migration I/Os used, incurred, or saved
Note: Each allocation is for a cylinder size worth of data, also known internally in the
allocator as an extent. Thus the column names begin with Ext for extent.
96

Statistics Columns
Column Name
Mode
Description
Data Type
ExtMigrateFaster
count
Number of cylinders migrated to a faster location on a device. This

count is for cylinders that were allocated on this device and
migrated to a different location within the same device or migrated
to a completely different device.
FLOAT
The following formula calculates a ExtMigrateSlower value, which

is the number of cylinders migrated to slower locations:
ExtMigrateSlower = ExMigrateTotal - ExMigrateFaster.
ExtMigrateIOTimeBenefit
count
Estimates the total I/O time savings achieved by migrations

completing in the log period. The I/O time savings include the
improvement in response time caused by the new data arrangement
up to the time horizon. ExtMigrateIOTimeBenefit does not include
the cost of the migration I/Os and is a gross benefit, not a net
benefit.
FLOAT
ExtMigrateIOTimeCost
count
Estimates the total cost, in centiseconds, incurred by migration

I/Os completing during the log period, where cost is the extra time
waited by all non-migration I/Os as a result of the migration
I/O.
FLOAT
ExtMigrateIOTimeImprove
count
Estimates the percent improvement in average I/O response time

due to migrations completing in the log interval. For example, if,
right before a particular log interval, the average I/O response time
was 10 milliseconds (ms), then the migration logs an
ExtMigrateIOTimeImprove value of 10% in this interval. The
average I/O response time after the log interval should be (100%10%)*10ms = 9ms. Migration then logs an
ExtMigrateIOTimeImprove of 1% in the next interval. The average
I/O response time in the new log interval is (100%-1%)*9ms =
8.91ms.
FLOAT
ExtMigrateIOTimeImprove is only an estimate. Its permanent

improvement remains in effect as long as the workload does not
change and newer migrations do not significantly alter the data
arrangement.
When the workload changes or new migrations affect data
arrangement, response time changes in an unquantifiable way.
Despite this, ExtMigrateIOTimeImprove is useful because it
predicts actual system performance at least for short periods of
time and can be used to understand why the migration algorithm is
doing what it is doing.
ExtMigrateReadRespTot
count
Migration read I/O response time.
FLOAT
ExtMigrateWriteRespTot
count
Migration write I/O response time.
FLOAT
ExtMigrateTotal
count
Total number of cylinders migrated to a different physical location.

For more information, see the ExtMigrateFaster field.
FLOAT
97

Summary Mode
Reserved Column
Column Name
Mode
Description
Data Type
Reserved
n/a
CHAR (1)
Summary Mode
When Summary Mode is active for the ResUsageSpdsk table, rows are summarized into a
single row for each pdisk type (for example, DISK, FILE, or SSD) for each node in the system
per log interval.
that row.
'S'
in Summary Mode.
'N'
normally.
Spare Columns
The ResUsageSpdsk table spare fields are named Spare00 through Spare19, and SpareInt.
Column Name
Description
Spare00

field using the WM_CPU_COD column name in the ResSpdskView
view. For details, see ResSpdskView on page 172.
1000 on SLES 10.
98

Spare Columns
Column Name
Description
Spare01

field using the WM_IO_COD column name in the ResSpdskView
100 on SLES 10.
SpareInt
PM I/O COD value in whole percent. For example, a value of 50

represents a PM I/O COD value of 50.0%.
Note: This field will be converted to the PM_IO_COD column in
field using the PM_IO_COD column name in the ResSpdskView
Related Topics
99

Spare Columns
100
CHAPTER 11
ResUsageSps
You can use the ResUsageSps table to:
Get a historical view of workload behavior for utilities and SQL operations.
Determine the number of workload requests that are using AMP Worker Task (see the
NumRequests column for details).
Examine queue wait and service time numbers to find backed up queries and allocation
groups.
Determine which workload is responsible for I/O skew.
Monitor CPU usage managed by the Priority Scheduler.
Identify the percent of CPU being used by different workloads.
If you are using the ResUsageSps table on SLES 10, this allows you to validate the relative
weights given to workloads.
For a complete description of the Priority Scheduler and its components, see Priority
Scheduler (schmon) chapter in Utilities.
If you are running SLES 10 or earlier ...
If you are running SLES 11 or later ...
and using TASM, each WD is equivalent of one

PGId.
and using TASM, each WD is equivalent of one

Priority Scheduler workload definition ID
(pWDid).
and table logging is enabled on ResUsageSps, a

row is written to the database once for every
triplet of Vproc Type (VprType), PG ID, and
Performance Period ID (VprType, PGId, PPId)
on each node in the system for each log interval.
and table logging is enabled on ResUsageSps, a

row is written to the database once for every
pWDid and VprType in the system for each log
interval.
For more information on the PGId and pWDid columns, see the RowIndex1 column.
Note: This table is created as a MULTISET table.
101
Chapter 11: ResUsageSps

Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT

columns.
NodeID
n/a

INTEGER


Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
NodeType
n/a
CHAR(8)
PPId
n/a
On SLES 10 or earlier systems, this column identifies the

performance period. The PPId is a mapping of the internal
performance period value (ranges 0 to 7) to a RSS value (ranges
0 to 1). A PPId of 0 maps to the value 0, and the PPId of 1 maps
to the values 1 through 7.
On SLES 11 or later systems, this column is not valid and returns
a zero value.
BYTEINT
VprType
n/a
Type of vproc (for example, AMP, PE, and MISC). Rows reported as
vproc type of MISC contain data for all vproc types other than the
AMP and PE vproc types.
CHAR(4)
AMPCount
n/a
Number of AMPs on the Node. AMPCount is used to divide

columns that are reporting data from all the AMPs. This allows the
ResSpsView view to report the data columns on a per AMP basis.
For an example of the view, see ResSpsView on page 175.
SMALLINT
102

Column Name
Mode
Description
Data Type
RowIndex1
n/a
On SLES 10 or earlier systems, this column contains the PGId.

The PGId identifies the PG. There is a one to one mapping
between a PG ID and a WD ID at any point in time.
SMALLINT
The mapping between PG ID and WD ID can be determined by

looking at the Teradata Viewpoint Workload Designer portlet.
The PG ID value ranges from 0 to 250, while the value of a
WD ID is not in a specific range (that is, the value is
incremented and not reused).
On SLES 11 or later systems, this column contains the pWDid.
The pWDid value ranges from 0 to 255.
Note: When running SLES 11 or later systems in the ResPs
macro outputs, the RowIndex1 column displays as PGid, instead
of pWDid. For sample outputs of the ResPs macros, see
Chapter 15: Resource Usage Macros.
WDId
track
WD ID.
FLOAT
The default PGs (System, L, M, H, R) have no associated WD and

will have a WDID of zero in the table. Only PGs with a name that
has PGWL in it have a nonzero WDID.
Note: This column is valid on SLES 11 or later systems only.
Secs
n/a

Normally, this value is the same as NominalSecs, but can be
different in three cases:
SMALLINT

CentiSecs
n/a

NominalSecs
n/a
INTEGER
SMALLINT
NCPUs
n/a
SMALLINT

Active
max

usage tables when Active Row Filter Mode is enabled.
FLOAT

103

Statistics Columns
Column Name
Mode
Description
Data Type
TheTimestamp
n/a
BIGINT
CODFactor
n/a

SMALLINT

VPId
track
Virtual partition ID.
FLOAT
Only one VpId is associated with a pWDid and VprType row at any
point in time. There can be multiple pWDid values associated with
a VPId.
Statistics Columns
TASM Columns
Monitor WD Columns
The Monitor WD columns provide RSS data to the following System PMPC APIs:
MONITOR WD request
MonitorWD function
For more information on these APIs, see Application Programming Reference.

The Control AMP WorkTimeInuse and WorkTimeInuseMax columns described in the
following table exclude uncompleted work time for the Control GDO AMP during recovery
work. This will occur for rows with the RowIndex1 column equal to zero, but will typically not
be noticed. Work time for recovery work that completes during the reporting period is
reported for the Control AMP.
For more information on the RowIndex1 column, see Miscellaneous Housekeeping
Columns on page 102.
Column Name
Mode
Description
Data Type
ActiveSessions
track
On SLES 10 or earlier systems, this is the number of Scheduling

Sets.
On SLES 11or later systems, this is the number of request-level
workload management objects created. There is one object
created per user request. For non-user work request, there is one
object created per WD.
FLOAT
AGId
track
The current Allocation Group (AG) for the PG ID that is being

reported. This value can be any number from 0 to 200.
FLOAT
Note: This column is valid on SLES 10 or earlier systems only.
104

Statistics Columns
Column Name
Mode
Description
Data Type
RelWgt
track
The Active Relative Weight. That is, the dynamically assigned

relative weight that considers, in its calculation, the activity of all
other Allocation Groups present on the system. The RelWgt
column constantly changes, unlike the relative weight assignment
the Database Administrator assigns in the Teradata Viewpoint
Workload Designer portlet.
FLOAT
RelWgt is the average relative weight of active online nodes (that is,
divide the sum of the non-zero RelWgt by the count of online
nodes with the non-zero RelWgt).
CPURunDelay
count
Number of milliseconds tasks in the WD sat in the CPU runqueue

waiting to run over the reporting period.
FLOAT
This data can be used in determining demand for the virtual

partition and WD Share workload management method. If the
CPU and I/O percentages for a virtual partition or WD are below
their relative share values and the CPURunDelay values are low,
there was insufficient demand to meet the share percentage. If the
CPURunDelay values are high, higher tier SQL requests were
allocated more resources so that there were insufficient resources
remaining to allocate to SQL requests in this WD to meet its relative
share.
For descriptions of the virtual partition and WD share workload
management method, see Glossary.
DecayLevel1IO
count
Number of times SQL requests in the WD hit decay level 1 due to

I/O.
FLOAT
The values are reported under VprType MISC because the summed
usage from PE and AMP go toward the usage that trips the decay.
The values reported from either PE or AMP usage alone would not
be accurate.
Note: DecayLevel1IO is used for Timeshare WDs only. For a
description of this workload management method, see
Glossary.
DecayLevel2IO
count
Number of times SQL requests in the WD decay level 2 due to I/O.
FLOAT
be accurate.
Note: DecayLevel2IO is used for Timeshare WDs only. For a
description of this workload management method, see Glossary.
105

Statistics Columns
Column Name
Mode
Description
Data Type
DecayLevel1CPU
count

CPU.
FLOAT
be accurate.
Note: DecayLevel1CPU is used for Timeshare WDs only. For a
DecayLevel2CPU
count

CPU.
FLOAT
be accurate.
Note: DecayLevel2CPU is used for Timeshare WDs. For a
IOBlks
count
Number of logical data blocks read or written.
FLOAT

IOCompleted
count
Number of I/Os completed on behalf of this WD.
FLOAT

IOCompletedKB
count
KB of I/O completed on behalf of this WD.
FLOAT

IOCriticalSubmitted
count
Number of I/Os submitted with a MI Lock Priority. These I/Os go

into a Teradata Scheduler queue and are not prioritized based on
the type of management method.
FLOAT

IOCriticalSubmittedKB
count
KB of I/O submitted with an MI Lock Priority. These I/Os go into a

Teradata Scheduler queue and are not prioritized based on the type
of management method.
FLOAT

IOSubmitted
count
Number of I/Os submitted on behalf of this WD.
FLOAT

IOSubmittedKB
count
KB of I/O submitted on behalf of this WD.
FLOAT

NumRequests
106
count
Number of AMP Worker Task messages or requests that got

assigned AMP Worker Tasks to them.
FLOAT

Statistics Columns
Column Name
Mode
Description
Data Type
NumTasks
track
Average number of tasks of online nodes. The column is the result

of:
FLOAT
NumTasks = SUM of (NumTasks-i) / N

where:
NumTasks-i is the number of tasks:
On SLES 10 or earlier systems, assigned to the PG at the end
of the reporting period.
On SLES 11 or later systems, assigned to the WD at the end of
the reporting period.
i varies from 1 to N.
N is the number of online nodes.
QWaitTime
count
Total time in milliseconds that work requests waited on an input

queue before being serviced. If the work requests are not delivered,
they are not counted.
FLOAT
To calculate the average QWaitTime per request, divide by

NumRequests, which is reported in DBC.ResSpsView as
QWaitTimeRequestAvg.
QWaitTimeMax
max
Maximum time in milliseconds that work requests waited on an

input queue before being serviced.
FLOAT
ServiceTime
count
Time in milliseconds that work requests required for service.
FLOAT
To calculate an approximate average ServiceTime for each request

during this period, divide the ServiceTime value by the
NumRequests value, which is reported in DBC.ResSpsView as
ServiceTimeRequestAvg.
The service time is the elapsed time from the time the message was
received to the time the AMP Worker Task was released. This is the
amount of time the AMP Worker Task was held through sleeps,
CPU, I/O, and so on until it is released.
ServiceTimeMax
max
Maximum time in milliseconds that work requests required for

service.
FLOAT
TacticalExceptionIO
count
Number of times SQL requests in the WD hit a tactical per-node

exception due to I/O.
FLOAT
usage from PE and AMP go toward the usage that trips the
exception. The value reported from either PE or AMP usage alone
would not be accurate.
An exception, used only for Tactical WDs, is created for each
Tactical WD. For a description of this WD Share workload
management method, see Glossary.
107

Statistics Columns
Column Name
Mode
Description
Data Type
TacticalExceptionCPU
count
Number of times SQL requests in the WD hit a tactical per-node

exception due to CPU.
FLOAT
usage from PE and AMP go toward the usage that trips the
exception. The value reported from either PE or AMP usage alone
would not be accurate.
Note: TacticalExceptionCPU is used for Tactical WDs only. For a
WaitIO
count
Number of milliseconds tasks in WD waited for I/O over the

reporting period.
FLOAT
WaitIO is updated when the wait for I/O is completed.

WorkMsgReceiveDelay
count
Time for all messages not yet delivered at the end of each reporting
period. This column is related to the QWaitTime column and
represents a running total of delays attributed to the tasks that still
have not been assigned an AMP Worker Task within this interval.
When the task does receive an AMP Worker Task in a later interval,
the time attributed here is counted again within QWaitTime of the
interval where it was assigned an AMP Worker Task.
FLOAT
WorkMsgReceiveDelayCnt
count
Number of messages that are still waiting for AMP Worker Tasks at
the end of each reporting period.
FLOAT
WorkMsgReceiveDelayMax
max
Maximum delay time in milliseconds for messages that are still in

the work box.
FLOAT
WorkMsgReceiveDelayCntMax
max
Maximum number of messages on the work mailbox wait to be

picked up by the AMP Worker Tasks at the end of each reporting
period.
FLOAT
WorkMsgSendDelay
count
Total time in milliseconds it takes to deliver work messages.
FLOAT
WorkMsgSendDelayCnt
count
Number of messages that are delivered to the work box.
FLOAT
WorkMsgSendDelayMax
max
Maximum time in milliseconds that it takes to deliver any single

work message.
FLOAT
WaitOther
count
Number of milliseconds tasks in WD waited for reasons other than

I/O over the reporting period (for example, a task waiting for a
message).
FLOAT
WaitOther is updated when wait is completed.

WorkTimeInuse
count
Service time consumed by a WD during the current reporting

period.
FLOAT
Note: This is not the running sum of a WD that exists over multiple
intervals.
WorkTimeInuseMax
108
max
Maximum service time of a single task in a WD that is running or

has finished in the current reporting period. This includes time
used during previous intervals for that task.
FLOAT

Statistics Columns

These columns report statistics about the AMP Worker Tasks. For more information about the
ResUsageSawt table and columns, see Chapter 7: ResUsageSawt Table.
The data is reporting the contribution of the respective WD to the column and the values are
not the same as the values reported in the ResUsageSawt table. The ResUsageSps table values
should add up to the ResUsageSawt table for columns like WorkTypeInuse. The Max columns
will not be able to be correlated to the ResUsageSawt table Max values in such a direct way
since the ResUsageSps Max columns report the Max value of the ResUsageSps table InUse
column for the WD and not the Max value of the ResUsageSawt table for all the WDs
combined.
Column Name
Mode
Description
Data Type
AwtReleases
count
Number of AMP Worker Tasks released (that is, completed

requests) while the NumRequests column reports the number of
AMP Worker Task requests that arrived. For details, see
NumRequests.
FLOAT
WorkTypeInuse00 WorkTypeInuse15
track
Current number of AMP Worker Tasks in use for each work type.
FLOAT
WorkTypeMax00 WorkTypeMax15
max
Maximum of the WorkTypeInuse values reported during the

logging period. When multiple data samples occur during the
reporting period the value is the maximum of the sampled values.
FLOAT
The true maximum number of in-use AMP Worker Tasks of a

WorkType may occur at a different time during the reporting
period and not be seen and, therefore, not be reported.

These columns represent CPU activities on the node associated with the AMP Worker Task,
Dispatcher, Parser, or miscellaneous activities.
Note: When an external routine (such as C, C++, Java UDF, or external stored procedure)
forks a child process or thread, the CPU time is not reported to these fields. As a result, the
resource usage table shows a lower CPU usage than shown in the ResUsageSpma table even if
the external routine consumes a large amount of CPU time. If there are child processes or
threads running on your system, this may account for the larger CPU times reported in the
ResUsageSpma table compared to this table. To confirm that the difference in CPU usage
reported by the ResUsageSpma table is caused by child processes or threads, contact your
Teradata Support Center personnel.
Column Name
Mode
Description
Data Type
CPUUServAWT
count
Time in milliseconds CPUs are busy in the AMP Worker Task

executing user service code. This is the system level time spent on a
process.
FLOAT
109

Statistics Columns
Column Name
Mode
Description
Data Type
CPUUServDisp
count
Time in milliseconds CPUs are busy in the Dispatcher or Parser

executing user service code. This is the system level time spent on a
process.
FLOAT
CPUUServMisc
count
Time in milliseconds CPUs are busy executing miscellaneous

activities for user service code. This is the system level time spent on
a process.
FLOAT
CPUUExecAWT
count
Time in milliseconds CPUs are busy in the AMP Worker Task

executing user execution code. This is the user level time spent on a
process.
FLOAT
CPUUExecDisp
count
Time in milliseconds CPUs are busy in the Dispatcher or Parser

executing user execution code. This is the user level time spent on a
process.
FLOAT
CPUUExecMisc
count
Time in milliseconds CPUs are busy executing miscellaneous

activities for user execution code. This is the user level time spent
on a process.
FLOAT
Process Block Counts Columns

These columns identify how many times a process became blocked.
For more information about process block counts columns, see Chapter 6: ResUsageSpma
Table.
Column Name
Mode
Description
Data Type
ProcBlksCPULimit
count
Number of times processes were blocked by delays due to the

Priority Scheduler CPU Limit (for example, System Limit, AG Limit,
or RP).
FLOAT
For information about the Priority Scheduler CPU Limit, see

Utilities.
ProcBlksDBLock
count
Number of times processes were blocked for a database lock.
FLOAT
ProcBlksFsgLock
count
Number of times processes were blocked for an FSG lock.
FLOAT
ProcBlksSegLock
count
Number of times processes were blocked for a disk or task context

(for example, scratch, stack, and so on).
FLOAT
ProcBlksTime
count
Number of times processes were blocked for a timer expiration.
FLOAT
ProcBlksFlowControl
count
Number of times processes were blocked by delays caused by the

flow control conditions.
FLOAT
ProcBlksFsgNIOs
count
Number of times processes were blocked waiting for task I/Os to

complete.
FLOAT
ProcBlksFsgRead
count
Number of times processes were blocked for an FSG read.
FLOAT
ProcBlksFsgWrite
count
Number of times processes were blocked for an FSG write from disk.
FLOAT
ProcBlksMisc
count
Number of times processes were blocked for miscellaneous events.
FLOAT
110

Statistics Columns
Column Name
Mode
Description
Data Type
ProcBlksMonitor
count
Number of times processes were blocked for a user monitor.
FLOAT
ProcBlksMonResume
count
Number of times processes were blocked for a user monitor resume

from a yield.
FLOAT
ProcBlksNetThrottle
count
Number of times processes were blocked for delivery of outstanding

outgoing messages.
FLOAT
ProcBlksSegMDL
count
Number of times processes were blocked waiting for an MDL

resource to become available. An MDL is an internal PDE data
structure needed by the operation of the segment subsystem.
FLOAT
ProcBlksSegNoVirtual
count
Number of times processes were blocked waiting for virtual memory

for a segment.
FLOAT
ProcBlksQnl
count
Number of times processes were blocked for a TSKQNL lock.
FLOAT

These columns identify the how long the processes were blocked for each possible reason
listed below.
The following column definition descriptions can also be found in Process Pending Wait
Time Columnsof Chapter 6: ResUsageSpma Table.
Column Name
Mode
Description
Data Type
ProcWaitCPULimit
count
Total time in milliseconds processes were blocked pending delays

due to the Priority Scheduler CPU Limits (for example, System
Limit, AG Limit, or RP).
FLOAT
For information about the Priority Scheduler CPU Limit, see

Utilities.
ProcWaitFlowControl
count
Total time in milliseconds processes were blocked pending the

delays caused by flow control conditions.
FLOAT
ProcWaitFsgLock
count
Total time in milliseconds processes were blocked pending an FSG

lock.
FLOAT
ProcWaitDBLock
count
Total time in milliseconds processes were blocked pending a

database lock.
FLOAT
ProcWaitSegLock
count
Total time in milliseconds processes were blocked pending a disk or

task context (for example, scratch, stack, and so on) lock.
FLOAT
ProcWaitFsgRead
count

read from disk.
FLOAT
ProcWaitFsgWrite
count

write from disk.
FLOAT
ProcWaitFsgNIOs
count
Total time in milliseconds processes were blocked waiting for task I/

Os to complete.
FLOAT
111

Statistics Columns
Column Name
Mode
Description
Data Type
ProcWaitMisc
count
Total time in milliseconds processes were blocked pending

miscellaneous events.
FLOAT
ProcWaitMonitor
count
Total time in milliseconds processes were blocked pending a user

monitor.
FLOAT
ProcWaitMonResume
count
Total time in milliseconds processes were blocked pending a user

monitor resume from a yield.
FLOAT
ProcWaitNetThrottle
count
Total time in milliseconds processes were blocked pending delivery

of outstanding outgoing messages.
FLOAT
ProcWaitSegMDL
count
Total time in milliseconds processes were blocked waiting for an

MDL resource to become available. An MDL is an internal PDE
data structure needed by the operation of the segment subsystem.
FLOAT
ProcWaitSegNoVirtual
count
Total time in milliseconds processes were blocked waiting for

virtual memory for a segment.
FLOAT
ProcWaitTime
count
Total time in milliseconds processes were blocked pending some

amount of elapsed time only.
FLOAT
ProcWaitQnl
count
Total time in milliseconds processes were blocked pending an

TSKQNL lock.
FLOAT
File System Columns

the log period. Logical acquires (Acqs) and the logical amount acquired (AcqKB) are
identified. Acquires causing physical reads (AcqReads) and the amount read (AcqReadKB) are
identified as a subset of logical acquires.
Column Name
Mode
Description
Data Type
FileAPtAcqs
count
Total number of append table or permanent journal table data

block or cylinder index disk segments acquired.
FLOAT
FilePCiAcqs
count
Total number of permanent cylinder index disk segments acquired.
FLOAT
FilePDbAcqs
count
Total number of permanent data block disk segments acquired.
FLOAT
FileSCiAcqs
count
Total number of regular or restartable spool index disk segments

acquired.
FLOAT
FileSDbAcqs
count
Total number of regular or restartable spool data block disk

segments acquired.
FLOAT
FileTJtAcqs
count
Total number of transient journal table disk segments acquired.
FLOAT
FileAPtAcqKB
count
Total KB acquired by FileAPtAcqs.
FLOAT
FilePCiAcqKB
count
Total KB acquired by FilePCiAcqs.
FLOAT
112

Statistics Columns
Column Name
Mode
Description
Data Type
FilePDbAcqKB
count
Total KB acquired by FilePDbAcqs.
FLOAT
FileSCiAcqKB
count
Total KB acquired by FileSCiAcqs.
FLOAT
FileSDbAcqKB
count
Total KB acquired by FileSDbAcqs.
FLOAT
FileTJtAcqKB
count
Total KB acquired by FileTJtAcqs.
FLOAT
FileAPtAcqReads
count
Number of append table or permanent journal table data block or

cylinder index disk segments that caused a physical read.
FLOAT
FilePCiAcqReads
count
Number of permanent cylinder index disk segment acquires that

caused a physical read.
FLOAT
FilePDbAcqReads
count
Number of permanent data block disk segment acquires that

FLOAT
FileSCiAcqReads
count
Number of regular or restartable spool index disk segments that

FLOAT
FileSDbAcqReads
count
Number of regular or restartable spool data block disk segments

that caused a physical read.
FLOAT
FileTJtAcqReads
count
Number of transient journal table disk segments that caused a

physical read.
FLOAT
FileAPtAcqReadKB
count
KB physically read by FileAPtAcqReads.
FLOAT
FilePCiAcqReadKB
count
KB physically read by FilePCiAcqReads.
FLOAT
FilePDbAcqReadKB
count
KB physically read by FilePDbAcqReads.
FLOAT
FileSCiAcqReadKB
count
KB physically read by FileSCiAcqReads.
FLOAT
FileSDbAcqReadKB
count
KB physically read by FileSDbAcqReads.
FLOAT
FileTJtAcqReadKB
count
KB physically read by FileTJtAcqReads.
FLOAT
Data Block Prefetch Columns

These columns identify File Segment Prefetch activities.
Note: A prefetch is either a cylinder read operation or an individual block read operation.
Either of these operations are generically called a prefetch.
Column Name
Mode
Description
Data Type
FileAPtPres
count

block or cylinder index disk segments prefetched.
FLOAT
FilePCiPres
count
Total number of permanent cylinder index disk segments

prefetched.
FLOAT
FilePDbPres
count
Total number of permanent data block disk segments prefetched.
FLOAT
113

Statistics Columns
Column Name
Mode
Description
Data Type
FileSCiPres
count

prefetched.
FLOAT
FileSDbPres
count

segments prefetched.
FLOAT
FileTJtPres
count
Total number of transient journal table disk segments prefetched.
FLOAT
FileAPtPresKB
count
Total number of KB prefetched by FileAPtPres.
FLOAT
FilePCiPresKB
count
Total number of KB prefetched by FilePCiPres.
FLOAT
FilePDbPresKB
count
Total number of KB prefetched by FilePDbPres.
FLOAT
FileSCiPresKB
count
Total number of KB prefetched by FileSCiPres.
FLOAT
FileSDbPresKB
count
Total number of KB prefetched by FileSDbPres.
FLOAT
FileTJtPresKB
count
Total number of KB prefetched by FileTJtPres.
FLOAT
FileAPtPreReads
count

block or cylinder index disk segment prefetches that caused a
physical read.
FLOAT
FilePCiPreReads
count
Total number of permanent cylinder index disk segment prefetches

FLOAT
FilePDbPreReads
count
Total number of permanent data block disk segment prefetches that

FLOAT
FileSCiPreReads
count
Total number of regular or restartable spool index disk segment

prefetches that caused a physical read.
FLOAT
FileSDbPreReads
count

segment prefetches that caused a physical read.
FLOAT
FileTJtPreReads
count
Total number of transient journal table disk segment prefetches

FLOAT
FileAPtPreReadKB
count
Total number of KB physical read by FileAPtPreReads.
FLOAT
FilePCiPreReadKB
count
Total number of KB physical read by FilePCiPreReads.
FLOAT
FilePDbPreReadKB
count
Total number of KB physically read by FilePDbPreReads.
FLOAT
FileSCiPreReadKB
count
Total number of KB physical read by FileSCiPreReads.
FLOAT
FileSDbPreReadKB
count
Total number of KB physical read by FileSDbPreReads.
FLOAT
FileTJtPreReadKB
count
Total number of KB physical read by FileTJtPreReads.
FLOAT

those segments that are dropped from memory during the log period. When a segment is
released, the segment is either:
114
Force out of memory (F)

Statistics Columns
Remains resident in memory (R)
Aged out of memory (A), from segments that remain resident
Both the number of segments (Rels, Writes, Drps) and the size of the segments (RelKB,
WriteKB, DrpKB) are counted. When a segment leaves memory, it must be written to disk
only if the segment is dirty, that is, modified (Dy). Otherwise, the clean or unmodified (Cn)
segment is simply dropped.
Most spool blocks are simply dropped from a task and put on the age queue. This may happen
multiple times. Each of these will be counted as a resident release. If the system is low on
memory and the age queue must be processed, this may also result in an age write or age drop.
Forced writes are always also counted as either clean resident releases or forced drops,
depending on whether age normal or age out now was specified.
Column Name
Mode
Description
Data Type
FileAPtDyRRels
count
Number of dirty append table and permanent journal data block or

cylinder index disk segment resident releases.
FLOAT
FilePCiDyRRels
count
Number of dirty permanent cylinder index disk segment resident

releases.
FLOAT
FilePDbDyRRels
count
Number of dirty permanent data block disk segment resident

releases.
FLOAT
FileSCiDyRRels
count
Number of dirty regular or restartable spool cylinder index disk

segment resident releases.
FLOAT
FileSDbDyRRels
count
Number of dirty regular or restartable spool data block disk

FLOAT
FileTJtDyRRels
count
Number of dirty transient journal table or WAL data block or WAL

FLOAT
FileAPtDyRRelKB
count
KB released by FileAPtDyRRels.
FLOAT
FilePCiDyRRelKB
count
KB released by FilePCiDyRRels.
FLOAT
FilePDbDyRRelKB
count
KB released by FilePDbDyRRels.
FLOAT
FileSCiDyRRelKB
count
KB released by FileSCiDyRRels.
FLOAT
FileSDbDyRRelKB
count
KB released by FileSDbDyRRels.
FLOAT
FileTJtDyRRelKB
count
KB released by FileTJtDyRRels.
FLOAT
FileAPtFWrites
count
Number of append table and permanent journal data block or

cylinder index disk segment forced releases or specific I/O requests
causing an immediate physical write.
FLOAT
FilePCiFWrites
count
Number of permanent cylinder index disk segment forced releases

or specific I/O requests causing an immediate physical write.
FLOAT
FilePDbFWrites
count
Number of permanent data block disk segment forced releases or

specific I/O requests causing an immediate physical write.
FLOAT
115

Statistics Columns
Column Name
Mode
Description
Data Type
FileSCiFWrites
count
Number of regular or restartable spool cylinder index disk segment

forced releases or specific I/O requests causing an immediate
physical write.
FLOAT
FileSDbFWrites
count
Number of regular or restartable spool data block disk segment

physical write.
FLOAT
FileTJtFWrites
count
Number of transient journal table or WAL data block or WAL

FLOAT
FileAPtFWriteKB
count
KB written by FileAPtFWrites.
FLOAT
FilePCiFWriteKB
count
KB written by FilePCiFWrites.
FLOAT
FilePDbFWriteKB
count
KB written by FilePDbFWrites.
FLOAT
FileSCiFWriteKB
count
KB written by FileSCiFWrites.
FLOAT
FileSDbFWriteKB
count
KB written by FileSDbFWrites.
FLOAT
FileTJtFWriteKB
count
KB written by FileTJtFWrites.
FLOAT
Memory Allocation Columns

These columns represent the number and amount of memory allocations, subdivided into
(the only applicable) generic node memory type.
Column Name
Mode
Description
Data Type
MemAllocs
count
Number of successful SEG memory allocations.
FLOAT
MemAllocKB
count
Total KB attributed to SEG memory allocations.
FLOAT
Net Columns
methods (Brd) per net.
Column Name
Mode
Description
Data Type
NetBrdReads
count
Number of broadcast messages input to the vproc.
FLOAT
NetBrdWrites
count
Number of broadcast messages output from the vproc.
FLOAT
116

Spare Columns

and output messages passing through either Teradata Database net through point-to-point
(1:1) methods (PtP).
Column Name
Mode
Description
Data Type
NetPtPReads
count
Number of point-to-point messages input to the vproc on behalf of

the WD.
FLOAT
NetPtPWrites
count
Number of point-to-point messages output from the vproc on

behalf of the WD.
FLOAT
NetPtPReadKB
count
Total KB of point-to-point messages input to the vproc on behalf of

the WD.
FLOAT
NetPtPWriteKB
count
Total KB of point-to-point messages output to the vproc on behalf

of the WD.
FLOAT
Column Name
Mode
Description
Data Type
Reserved
n/a
CHAR (3)
Reserved Column
Spare Columns
The ResUsageSps table spare fields are named Spare00 through Spare19, and SpareInt. Four of
those fields are being used below.
Column Name
Description
Spare00

Teradata Database 15.0. You can access the resource usage data for
this field using the WM_CPU_COD column name in the ResSpsView
view. For details, see ResSpsView on page 175.
1000 on SLES 10.
117

Spare Columns
Column Name
Description
Spare01

this field using the WM_IO_COD column name in the ResSpsView
100 on SLES 10.
Spare03
Number of times the parent Virtual Partition query is throttled due to

a set VP-level CPU hard limit. The hard limit at a Virtual Partition
level is a node-level maximum percent.
Note: This field will be converted to the CpuVpThrottleCount
column in Teradata Database 15.0. You can access the resource usage
data for this field using the CpuVpThrottleCount column name in
the ResSpsView view. For details, see ResSpsView on page 175.
Spare04
Time, in milliseconds, that the parent Virtual Partition query is

throttled due to a set Virtual Partition-level CPU hard limit.
Note: This field will be converted to the CpuVpThrottleTime column
in Teradata Database 15.0. You can access the resource usage data for
this field using the CpuVpThrottleTime column name in the
ResSpsView view. For details, see ResSpsView on page 175.
Spare05
Number of times the workload query is throttled due to a set

workload-level CPU hard limit. The hard limit at a workload level is a
node-level maximum percent.
Note: This field will be converted to the CpuThrottleCount column
this field using the CpuThrottleCount column name in the
Spare06
Time, in milliseconds, that the workload query is throttled due to a

set workload-level CPU hard limit. The hard limit at a workload level
is a node-level maximum percent.
Note: This field will be converted to the CpuThrottleTime column in
this field using the CpuThrottleTime column name in the
118

Spare Columns
Column Name
Description
Spare07
Sum of the full potential input or output token allocations from all
devices attached to the node. This field is reported only to the AMP
vproc rows.
The input or output token allocations are not limited by the IO_COD
setting.
The data source for this field is located in the
all_wd_full_potential_iota field in /proc/tdmeter/limit_stats file.
this field using the FullPotentialIota column name in the ResSpsView
Note: This field is available on SLES 11 only.
Spare08
Sum of the potential input or output token allocations from all

devices attached to the node for this workload. This field is reported
only to the AMP vproc rows.
The input or output token allocations are accumulated only when an
I/O is pending on a device and are limited by the IO_COD setting.
The data source for this field is located in the wd_potential_iota field
in /proc/tdmeter/limit_stats file.
this field using the CodPotentialIota column name in the ResSpsView
Spare09
Sum of the used input or output token allocations from all devices
attached to the node for this workload. This field is reported only to
AMP vproc rows.
The data source for this field is located in the following file:
wd_used_iota field in /proc/tdmeter/limit_stats
Database 15.0. You can access the resource usage data for this field
using the UsedIota column name in the ResSpsView view. For details,
see ResSpsView on page 175.
119

Spare Columns
Column Name
Description
Spare10
Number of times that an I/O was in a releasable position to disk. This

field is reported only to AMP vproc rows.
An I/O may not be in a releasable position because there are higher
priority I/Os.
The data source for this field is:
((tdsched_io_info *)TvsaIoBuf_p->io_info)->req_pct)
Note: This field will be converted to the IoThrottleCount column in
this field using the IoThrottleCount column name in the ResSpsView
Spare11
Number of logical reads from VH cache. This field is populated by the

FSG subsystem.
Note: This field will be converted to the VHLogicalDBRead column
this field using the VHLogicalDBRead column name in the
For information about VH cache, see Glossary.
Spare12
Volume of logical reads in KB from VH cache. This field is populated

by the FSG system.
Note: This field will be converted to the VHLogicalDBReadKB
data for this field using the VHLogicalDBReadKB column name in
Spare13
Number of very hot reads that were handled by physical disk I/O due
to a VH cache miss (that is, data not found in the VH cache). This
field is populated by the FSG system.
Note: This field will be converted to the VHPhysicalDBRead column
this field using the VHPhysicalDBRead column name in the
Spare14
Volume of very hot reads in KB that were handled by physical disk

I/O due to a VH cache miss. This field is populated by the FSG
system.
Note: This field will be converted to the VHPhysicalDBReadKB
data for this field using the VHPhysicalDBReadKB column name in
120

Spare Columns
Column Name
Description
SpareInt
PM I/O COD value in whole percent values for the entire system. For
example, a SpareInt value of 50 represents a
PM I/O COD value of 50%.
This field value is 100 if the PM I/O COD is disabled.
field using the PM_IO_COD column name in the ResSpsView view.
For details, see ResSpsView on page 175.
Related Topics
121

Spare Columns
122
CHAPTER 12
ResUsageSvdsk Table
The ResUsageSvdsk table:
Provides AMP-level storage statistics.
Includes resource usage logs on cylinder allocation, migration, and I/O statistics.
If you enable table logging on ResUsageSvdsk, a row is written to the database once for every
AMP vproc in the system for each log interval. To consolidate and summarize the total
number of rows written to the database, you can enable Summary Mode. For details, see
Summary Mode on page 128.
Note: This table is created as a MULTISET table.
The following table describes the ResUsageSvdsk table columns.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT

columns.
NodeID
n/a

INTEGER


These columns provide the general characteristics of the housekeeping columns.
123
Chapter 12: ResUsageSvdsk Table

Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
VprId
n/a
Identifies the AMP vproc.
INTEGER
In Summary Mode, the value of the AMP vproc ID is -1.

NodeType
n/a
CHAR(8)
Secs
n/a

cases:
SMALLINT

CentiSecs
n/a

INTEGER
NominalSecs
n/a
SMALLINT
SummaryFlag
n/a

CHAR (1)
Active
max

FLOAT

usage tables.
TheTimestamp
n/a
BIGINT
CODFactor
n/a

SMALLINT
124

Statistics Columns
Statistics Columns
Teradata VS Columns
Allocation Columns
These columns identify the allocation statistics reported by the Allocator process.
Column Name
Mode
Description
Data Type
ExtAllocHot
count
Number of hot allocations made in the current log period. A hot

the pre-defined hot temperature range.
FLOAT
The cylinder resides in some disk storage location and holds some
data. Temperature is the frequency of access to the data by I/O
independent of where the data resides.
ExtAllocNonPacing
count
Number of non-pacing allocations made in the current log period. A

non-pacing allocation is an allocation whose data access affects
neither system performance nor individual query performance.
FLOAT
ExtAllocStatic
count
Number of static allocations made in the current log period. A static

allocation is an allocation whose requested temperature is used and
the measured temperature is ignored during migration.
FLOAT
count
Number of system pacing allocations made in the current log period.

A system pacing allocation is an allocation whose data access affects
system performance.
FLOAT
ExtAllocTotal
count
Total number of allocations made in the current log period. A

number of computations can be derived. For example:
FLOAT
Cold Allocation = ExtAllocTotal ExtAllocHot ExtAllocWarm

QueryPacing Allocation = ExtAllocTotal ExtAllocNonPacing
Dynamic Allocation = ExtAllocTotal - ExtAllocStatic
ExtAllocWarm
count
Number of warm allocations made in the current log period. A

warm allocation is an allocation whose estimated temperature falls
within the pre-defined warm temperature range.
FLOAT
The cylinder resides in some disk storage location and holds some
data. Temperature is the frequency of access to the data by I/O
independent of where the data resides.
I/O Statistics Columns

These columns identify the I/O statistics reported by FSG for each AMP.
Column Name
Mode
Description
Data Type
ConcurrentMax
max
Maximum number of concurrent I/O requests.
FLOAT
125

Statistics Columns
Column Name
Mode
Description
Data Type
ConcurrentReadMax
max
Maximum number of concurrent read I/O requests.
FLOAT
ConcurrentWriteMax
max
Maximum number of concurrent write I/O requests.
FLOAT
ReadCnt
count
FLOAT
WriteCnt
count
FLOAT
ReadKB
count
FLOAT
WriteKB
count
FLOAT
ReadRespTot
count
FLOAT
WriteRespTot
count
FLOAT
ReadRespMax
max
Maximum of individual read response time in centiseconds.
FLOAT
WriteRespMax
max
Maximum of individual write response time in centiseconds.
FLOAT
ReadRespSq
count
Total of squares of the individual read response time in

centiseconds.
FLOAT
WriteRespSq
count
Total of squares of the individual write response time in

centiseconds.
FLOAT
OutReqTime
count
Time with outstanding requests (busy time), in centiseconds.
FLOAT
Migration Columns
Note: The ResUsageSvdsk table migration columns are populated only in Teradata VS. For
more information, see Teradata Virtual Storage.
The migration columns identify the number of cylinders that migrated to a different location
on a device as well as the time, in centiseconds, of all migration I/Os used, incurred, or saved
Note: Each allocation is for a cylinder size worth of data, also known internally in the
allocator as an extent. Therefore, the column names begin with Ext for extent.
Column Name
Mode
Description
Data Type
ExtMigrateFaster
count
Number of cylinders migrated to faster locations (that is,

migrations whose gross benefits are positive) for the associated
AMP.
FLOAT
The following formula calculates a Slower Migration value, which is

the number of cylinders migrated to slower locations:
SlowerMigration = ExtMigrateTotal - ExtMigrateFaster
Cylinders are migrated to slower locations to make room for hotter
cylinders to replace them.
126

Statistics Columns
Column Name
Mode
Description
Data Type
ExtMigrateIOTimeCost
count
Estimates the total cost, in centiseconds, incurred by migration

I/Os completing during the log period, where cost is the extra time
waited by all non-migration I/Os as a result of the migration
I/O.
FLOAT
ExtMigrateIOTimeBenefit
count
Estimates the total I/O time savings achieved by migrations

completing in the log period. The I/O time savings include the
improvement in response time caused by the new data arrangement
up to the time horizon.
FLOAT
This value does not include the cost of the migration I/Os and is a
gross benefit, not a net benefit.
ExtMigrateIOTimeImprove
count
Estimates the percent improvement in average I/O response time

due to migrations completing in the log interval.
FLOAT
For example, if, right before a particular log interval, the average
I/O response time was 10 milliseconds (ms), then the migration
logs an ExtMigrateIOTimeImprove value of 10% in this interval.
The average I/O response time after the log interval should be
(100%-10%)*10ms = 9ms. Migration then logs an
ExtMigrateIOTimeImprove of 1% in the next interval. The average
I/O response time in the new log interval is (100%-1%)*9ms =
8.91ms.
ExtMigrateIOTimeImprove is only an estimate. Its permanent
improvement remains in effect as long as the workload does not
change and newer migrations do not significantly alter the data
arrangement.
When the workload changes or new migrations affect data
arrangement, response time changes in a nonquantifiable way.
You can use this field to predict the actual system performance for
short periods of time and to understand why the migration
algorithm is doing what it is doing.
ExtMigrateReadRespTot
count
Migration read I/O response time.
FLOAT
ExtMigrateWriteRespTot
count
Migration write I/O response time.
FLOAT
ExtMigrateTotal
count
Total number of cylinders migrated to a different physical location.

For more information, see the ExtMigrateFaster column.
FLOAT
Column Name
Mode
Description
Data Type
Reserved
n/a
CHAR (1)
Reserved Column
127

Summary Mode
Summary Mode
When Summary Mode is active for the ResUsageSvdsk table, one row is written to the database
for each node in the system. This row summarizes all AMP vdisk data in each node per log
interval.
that row.
'S'
in Summary Mode.
'N'
normally.
Spare Columns
The ResUsageSvdsk table spare fields are named Spare00 through Spare19, and SpareInt.
Column Name
Description
Spare00

field using the WM_CPU_COD column name in the ResSvdskView
view. For details, see ResSvdskView on page 178.
1000 on SLES 10.
Spare01

field using the WM_IO_COD column name in the ResSvdskView
100 on SLES 10.
128

Spare Columns
Column Name
Description
SpareInt
PM I/O COD value in whole percent. For example, a value of 50

represents a PM I/O COD value of 50.0%.
field using the PM_IO_COD column name in the ResSvdskView
Related Topics
129

Spare Columns
130
CHAPTER 13
ResUsageSvpr Table
ResUsageSvpr logical table includes resource usage data for available system-wide, virtual
processor information.
Teradata recommends that you use the view ResSvprView on page 179 to access the data
rather than accessing the ResUsageSvpr table directly.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT
Note: Under conditions of heavy system load, entries might be

column will contain the time value when the entry should have been
logged. For more information, see the Secs and NominalSecs
columns.
NodeID
n/a

INTEGER


These columns provide the general characteristics of the housekeeping columns.
Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
131
Chapter 13: ResUsageSvpr Table

Column Name
Mode
Description
Data Type
NodeType
n/a
CHAR(8)
VprId
n/a
Identifies the vproc number (non-Summary Mode) or the vproc

type (Summary Mode; 0 = NODE, 1 = AMP, 2 = PE, 3=GTW,
4=RSG, 5=TVS).
INTEGER
VprType
n/a
Type of vproc. The values can be NODE, AMP, PE, GTW, RSG, or
TVS (see Teradata Virtual Storage).
CHAR(4)
Secs
n/a

cases:
SMALLINT

CentiSecs
n/a

INTEGER
NominalSecs
n/a
SMALLINT
NCPUs
n/a
SMALLINT

SummaryFlag
n/a
is a non-summary row and 'S' if it is a summary row.
CHAR (1)
Active
max

FLOAT

usage tables.
TheTimestamp
n/a
BIGINT
CODFactor
n/a

SMALLINT
132

Statistics Columns
Statistics Columns
File System Columns
Synchronized Full Table Scans Columns
These columns contain statistics relating to synchronized full table scans.
Note: The following columns are moved from the ResUsageIvpr table to the ResUsageSvpr
table to avoid costly joins.
Column Name
Mode
Description
Data Type
FileSyncGroups
track
Number of groups of scanners involved in full table scans. A group

consists of scanners who are able to use the same read I/O to obtain
data from disk.
FLOAT
FileSyncScanners
track
Number of tasks involved in full table scans who are willing to

synchronize with other scanners.
FLOAT
FileSyncScans
count
Number of attempts to synchronize a full table scan.
FLOAT
FileSyncSubtables
track
Number of subtables scanned by one or more full table scanners

who are willing to synchronize scans.
FLOAT

the log period.
Note: The FileXXxAcqs columns are the only columns counted as cache hits, where XXx
represents one of the following entries:
APt
PCi
PDb
SCi
SDb
TJt
Column Name
Mode
Description
Data Type
FileAPtAcqs
count

block or cylinder index disk segments.
FLOAT
FilePCiAcqs
count
Total number of permanent cylinder index disk segments acquires

that were logically acquired.
FLOAT
FilePDbAcqs
count
Total number of permanent data block disk segments acquires that

were logically acquired.
FLOAT
FileSCiAcqs
count
Total number of regular or restartable spool cylinder index disk

segments acquires that were logically acquired.
FLOAT
FileSDbAcqs
count

segments acquires that were logically acquired.
FLOAT
133

Statistics Columns
Column Name
Mode
Description
Data Type
FileTJtAcqs
count
Total number of transient journal or WAL data block or WAL

cylinder index disk segments acquires that were logically acquired.
FLOAT
FileAPtAcqKB
count
Total KB logically acquired by FileAPtAcqs.
FLOAT
FilePCiAcqKB
count
Total KB logically acquired by FilePCiAcqs.
FLOAT
FilePDbAcqKB
count
Total KB logically acquired by FilePDbAcqs.
FLOAT
FileSCiAcqKB
count
Total KB logically acquired by FileSCiAcqs.
FLOAT
FileSDbAcqKB
count
Total KB logically acquired by FileSDbAcqs.
FLOAT
FileTJtAcqKB
count
Total KB logically acquired by FileTJtAcqs.
FLOAT
FileAPtAcqReads
count

cylinder index disk segment acquires that caused a physical read.
FLOAT
FilePCiAcqReads
count
Number of permanent cylinder index disk segment acquires that

FLOAT
FilePDbAcqReads
count
Number of permanent data block disk segment acquires that

FLOAT
FileSCiAcqReads
count

acquires that caused a physical read.
FLOAT
FileSDbAcqReads
count

acquires that caused a physical read.
FLOAT
FileTJtAcqReads
count
Number of transient journal or WAL data block or WAL cylinder

index disk segment acquires that caused a physical read.
FLOAT
FileAPtAcqReadKB
count
Total KB physically read by FileAPtAcqReads.
FLOAT
FilePCiAcqReadKB
count
Total KB physically read by FilePCiAcqReads.
FLOAT
FilePDbAcqReadKB
count
Total KB physically read by FilePDbAcqReads.
FLOAT
FileSCiAcqReadKB
count
Total KB physically read by FileSCiAcqReads.
FLOAT
FileSDbAcqReadKB
count
Total KB physically read by FileSDbAcqReads.
FLOAT
FileTJtAcqReadKB
count
Total KB physically read by FileTJtAcqReads.
FLOAT
FileAPtAcqsOther
count

block or cylinder index scratch disk segments that were logically
acquired.
FLOAT
FilePCiAcqsOther
count
Total number of permanent cylinder index scratch disk segments

that were logically acquired.
FLOAT
FilePDbAcqsOther
count
Total number of permanent data block scratch disk segments that

were logically acquired.
FLOAT
FileSCiAcqsOther
count
Total number of regular or restartable spool cylinder index scratch

disk segments that were logically acquired.
FLOAT
134

Statistics Columns
Column Name
Mode
Description
Data Type
FileSDbAcqsOther
count
Total number of regular or restartable spool data block scratch disk

segments that were logically acquired.
FLOAT
FileTJtAcqsOther
count

cylinder index scratch disk segments that were logically acquired.
FLOAT
FileAPtAcqOtherKB
count

block or cylinder index scratch disk segments acquired in KB.
FLOAT
FilePCiAcqOtherKB
count
Total number of permanent cylinder index scratch disk segments

acquired in KB.
FLOAT
FilePDbAcqOtherKB
count
Total number of permanent data block scratch disk segments

acquired in KB.
FLOAT
FileSCiAcqOtherKB
count
Total number of regular or restartable spool cylinder index scratch

disk segments acquired in KB.
FLOAT
FileSDbAcqOtherKB
count
Total number of regular or restartable spool data block scratch disk

segments acquired in KB.
FLOAT
FileTJtAcqOtherKB
count

cylinder index scratch disk segments acquired in KB.
FLOAT
Data Block Prefetch Columns

These columns identify File Segment Prefetch activities.
Note: A prefetch is either a cylinder read operation or an individual block read operation.
Either of these operations are generically called a prefetch.
Column Name
Mode
Description
Data Type
FileAPtPres
count

block or cylinder index disk segments prefetched.
FLOAT
FilePCiPres
count
Total number of permanent cylinder index disk segments

prefetched.
FLOAT
FilePDbPres
count
Total number of permanent data block disk segments prefetched.
FLOAT
FileSCiPres
count

prefetched.
FLOAT
FileSDbPres
count

segments prefetched.
FLOAT
FileTJtPres
count
Total number of transient journal table disk segments prefetched.
FLOAT
FileAPtPresKB
count
Total number of KB prefetched by FileAPtPres.
FLOAT
FilePCiPresKB
count
Total number of KB prefetched by FilePCiPres.
FLOAT
FilePDbPresKB
count
Total number of KB prefetched by FilePDbPres.
FLOAT
FileSCiPresKB
count
Total number of KB prefetched by FileSCiPres.
FLOAT
135

Statistics Columns
Column Name
Mode
Description
Data Type
FileSDbPresKB
count
Total number of KB prefetched by FileSDbPres.
FLOAT
FileTJtPresKB
count
Total number of KB prefetched by FileTJtPres.
FLOAT
FileAPtPreReads
count

block or cylinder index disk segment prefetches that caused a
physical read.
FLOAT
FilePCiPreReads
count
Total number of permanent cylinder index disk segment prefetches

FLOAT
FilePDbPreReads
count
Total number of permanent data block disk segment prefetches that

FLOAT
FileSCiPreReads
count
Total number of regular or restartable spool index disk segment

prefetches that caused a physical read.
FLOAT
FileSDbPreReads
count

segment prefetches that caused a physical read.
FLOAT
FileTJtPreReads
count
Total number of transient journal table disk segment prefetches

FLOAT
FileAPtPreReadKB
count
Total number of KB physical read by FileAPtPreReads.
FLOAT
FilePCiPreReadKB
count
Total number of KB physical read by FilePCiPreReads.
FLOAT
FilePDbPreReadKB
count
Total number of KB physically read by FilePDbPreReads.
FLOAT
FileSCiPreReadKB
count
Total number of KB physical read by FileSCiPreReads.
FLOAT
FileSDbPreReadKB
count
Total number of KB physical read by FileSDbPreReads.
FLOAT
FileTJtPreReadKB
count
Total number of KB physical read by FileTJtPreReads.
FLOAT

those segments that are dropped from memory during the log period. When a segment is
released, the segment is either:
Forced (F)
Remains resident in memory (R)
Aged out of memory (A), from segments that are currently resident
Both the number of segments (Rels, Writes, Drps) and the size of the segments
(RelKB,WriteKB, DrpKB) are counted. When a segment leaves memory, it must be written to
disk only if the segment is dirty (Dy), that is, modified. Otherwise, the clean (Cn), that is,
unmodified segment is simply dropped.
Most spool blocks for a small table remain resident when they are created and age there. Each
of these will be counted as a dirty resident release (DyRRel columns). If a block survives in the
cache, it would be reacquired (whenever the system creates spool data, a subsequent step will
read it) and released again. The release will still be counted as a dirty resident release, since the
136

Statistics Columns
block survived in a modified state. On the other hand, if there is contention for room in the
FSG cache, the segment might be removed from memory. Because it is a modified segment, it
must be written out first. This is counted as a dirty age write (DyAWrite columns). When it is
reacquired it will no longer be modified, so the subsequent release will be counted as a clean
resident release (CnRRel columns).
If the segments are modified, only DyAWrites, DyAWriteKB, CnADrps, and CnADrpKB
columns are incremented.
If the segments are unmodified, only CnADrps and CnADrpKB columns are incremented.
Note: To determine the clean segments that aged out of memory for the CnADrps column,
subtract the DyAWrites value from the CnADrps value. To determine the clean segments that
aged out of memory for the CnADrpKB column, subtract the DyAWriteKB value from the
CnADrpKB value.
Full table modification operations make one pass on the table and modify each block only
once. Since these operations do not access a block multiple times, there is no point keeping
them in the cache. If a block that was examined did not contain any rows that qualify for the
modification, when the block is released, it will be dropped from memory immediately (FDrp
columns). However, if the block was modified, when it is released the system issues the write
as part of the release so it is counted as a forced write (FWrite columns). Since the system also
drops the block from memory as soon as the write is complete, this release is also counted as a
forced drop (FDrp columns).
Column Name
Mode
Description
Data Type
FileAPtDyRRels
count
Number of dirty append table or permanent journal table data

block or cylinder index disk segment resident releases.
FLOAT
FilePCiDyRRels
count
Number of dirty permanent cylinder index disk segment resident

releases.
FLOAT
FilePDbDyRRels
count
Number of dirty permanent data block disk segment resident

releases.
FLOAT
FileSCiDyRRels
count

FLOAT
FileSDbDyRRels
count

FLOAT
FileTJtDyRRels
count

FLOAT
FileAPtDyRRelsKB
count
KB released by FileAPtDyRRels.
FLOAT
FilePCiDyRRelsKB
count
KB released by FilePCiDyRRels.
FLOAT
FilePDbDyRRelsKB
count
KB released by FilePDbDyRRels.
FLOAT
FileSCiDyRRelsKB
count
KB released by FileSCiDyRRels.
FLOAT
FileSDbDyRRelsKB
count
KB released by FileSDbDyRRels.
FLOAT
137

Statistics Columns
Column Name
Mode
Description
Data Type
FileTJtDyRRelsKB
count
KB released by FileTJtDyRRels.
FLOAT
FileAPtFWrites
count

FLOAT
FilePCiFWrites
count

or specific I/O requests causing an immediate physical write.
FLOAT
FilePDbFWrites
count
Number of permanent data block disk segment forced releases or

specific I/O requests causing an immediate physical write.
FLOAT
FileSCiFWrites
count

physical write.
FLOAT
FileSDbFWrites
count

physical write.
FLOAT
FileTJtFWrites
count

FLOAT
FileAPtFWriteKB
count
KB written by FileAPtFWrites.
FLOAT
FilePCiFWriteKB
count
KB written by FilePCiFWrites.
FLOAT
FilePDbFWriteKB
count
KB written by FilePDbFWrites.
FLOAT
FileSCiFWriteKB
count
KB written by FileSCiWrites.
FLOAT
FileSDbFWriteKB
count
KB written by FileSDbFWrites.
FLOAT
FileTJtFWriteKB
count
KB written by FileTJtFWrites.
FLOAT
FileAPtDyAWrites
count
Number of dirty append table or permanent journal table data

block or cylinder index disk segments aged out of memory causing
a delayed physical write.
FLOAT
FilePCiDyAWrites
count
Number of dirty permanent cylinder index disk segments aged out

of memory causing a delayed physical write.
FLOAT
FilePDbDyAWrites
count
Number of dirty permanent data block disk segments aged out of

memory causing a delayed physical write.
FLOAT
FileSCiDyAWrites
count

segments aged out of memory causing a delayed physical write.
FLOAT
FileSDbDyAWrites
count

segments aged out of memory causing a delayed physical write.
FLOAT
FileTJtDyAWrites
count

cylinder index disk segments aged out of memory causing a delayed
physical write.
FLOAT
FileAPtDyAWriteKB
count
KB written by FileAPtDyAWrites.
FLOAT
138

Statistics Columns
Column Name
Mode
Description
Data Type
FilePCiDyAWriteKB
count
KB written by FilePCiDyAWrites.
FLOAT
FilePDbDyAWriteKB
count
KB written by FilePDbDyAWrites.
FLOAT
FileSCiDyAWriteKB
count
KB written by FileSCiDyAWrites.
FLOAT
FileSDbDyAWriteKB
count
KB written by FileSDbDyAWrites.
FLOAT
FileTJtDyAWriteKB
count
KB written by FileTJtDyAWrites.
FLOAT
FileAPtCnRRels
count
Number of clean append table or permanent journal table data

block or cylinder index disk segment resident releases.
FLOAT
FilePCiCnRRels
count
Number of clean permanent cylinder index disk segment resident

releases.
FLOAT
FilePDbCnRRels
count
Number of clean permanent data block disk segment resident

releases.
FLOAT
FileSCiCnRRels
count
Number of clean regular or restartable spool cylinder index disk

FLOAT
FileSDbCnRRels
count
Number of clean regular or restartable spool data block disk

FLOAT
FileTJtCnRRels
count
Number of clean transient journal table or WAL data block or WAL

FLOAT
FileAPtCnRRelKB
count
KB released by FileAptCnRRels.
FLOAT
FilePCiCnRRelKB
count
KB released by FilePCiCnRRels.
FLOAT
FilePDbCnRRelKB
count
KB released by FilePDbCnRRels.
FLOAT
FileSCiCnRRelKB
count
KB released by FileSCiCnRRels.
FLOAT
FileSDbCnRRelKB
count
KB released by FileSDbCnRRels.
FLOAT
FileTJtCnRRelKB
count
KB released by FileTJtCnRRels.
FLOAT
FileAPtFDrps
count

cylinder index disk segment forced releases causing an immediate
memory drop.
FLOAT
FilePCiFDrps
count

causing an immediate memory drop.
FLOAT
FilePDbFDrps
count
Number of permanent data block disk segment forced releases

causing an immediate memory drop.
FLOAT
FileSCiFDrps
count

forced releases causing an immediate memory drop.
FLOAT
FileSDbFDrps
count

forced releases causing an immediate memory drop.
FLOAT
FileTJtFDrps
count

cylinder index disk segment forced releases causing an immediate
memory drop.
FLOAT
139

Statistics Columns
Column Name
Mode
Description
Data Type
FileAPtFDrpKB
count
KB dropped by FileAPtFDrps.
FLOAT
FilePCiFDrpKB
count
KB dropped by FilePCiFDrps.
FLOAT
FilePDbFDrpKB
count
KB dropped by FilePDbFDrps.
FLOAT
FileSCiFDrpKB
count
KB dropped by FileSCiFDrps.
FLOAT
FileSDbFDrpKB
count
KB dropped by FileSDbFDrps.
FLOAT
FileTJtFDrpKB
count
KB dropped by FileTJtFDrps.
FLOAT
FileAPtCnADrps
count
Number of clean append table or permanent journal table data

block or cylinder index disk segments aged out of memory.
FLOAT
FilePCiCnADrps
count
Number of clean permanent cylinder index disk segments aged out

of memory.
FLOAT
FilePDbCnADrps
count
Number of clean permanent data block disk segments aged out of

memory.
FLOAT
FileSCiCnADrps
count
Number of clean regular or restartable spool cylinder index disk

segments aged out of memory.
FLOAT
FileSDbCnADrps
count
Number of clean regular or restartable spool data block disk

segments aged out of memory.
FLOAT
FileTJtCnADrps
count
Number of clean transient journal table or WAL data block or WAL

cylinder index disk segments aged out of memory.
FLOAT
FileAPtCnADrpKB
count
KB dropped by FileAPtCnADrps.
FLOAT
FilePCiCnADrpKB
count
KB dropped by FilePCiCnADrps.
FLOAT
FilePDbCnADrpKB
count
KB dropped by FilePDbCnADrps.
FLOAT
FileSCiCnADrpKB
count
KB dropped by FileSCiCnADrps.
FLOAT
FileSDbCnADrpKB
count
KB dropped by FileSDbCnADrps.
FLOAT
FileTJtCnADrpKB
count
KB dropped by FileTJtCnADrps.
FLOAT
FileAPtRelsOther
count
Total number of scratch append table or permanent journal table

data block or cylinder index disk segments released as part of them
being deleted.
FLOAT
FilePCiRelsOther
count
Total number of scratch permanent cylinder index disk segments

released as part of them being deleted.
FLOAT
FilePDbRelsOther
count
Total number of scratch permanent data block disk segments

released as part of them being deleted.
FLOAT
FileSCiRelsOther
count
Total number of scratch regular or restartable spool cylinder index

disk segments released as part of them being deleted.
FLOAT
FileSDbRelsOther
count
Total number of scratch regular or restartable spool data block disk

segments released as part of them being deleted.
FLOAT
140

Statistics Columns
Column Name
Mode
Description
Data Type
FileTJtRelsOther
count
Total number of scratch transient journal table or WAL data block

or WAL cylinder index disk segments released as part of them being
deleted.
FLOAT
FileAPtRelOtherKB
count
Total number of scratch append table or permanent journal table

data block or cylinder index disk segments released in KB.
FLOAT
FilePCiRelOtherKB
count
Total number of scratch permanent cylinder index disk segments

released in KB.
FLOAT
FilePDbRelOtherKB
count
Total number of scratch permanent data block disk segments

released in KB.
FLOAT
FileSCiRelOtherKB
count
Total number of scratch regular or restartable spool cylinder index

disk segments released in KB.
FLOAT
FileSDbRelOtherKB
count
Total number of scratch regular or restartable spool data block disk

segments released in KB.
FLOAT
FileTJtRelOtherKB
count
Total number of scratch transient journal table or WAL data block

or WAL cylinder index disk segments released in KB.
FLOAT
Data Segment Lock Requests Columns

These columns identify the number of lock requests, blocks, and deadlocks on a disk segment,
including those implied for segment acquires.
Column Name
Mode
Description
Data Type
FileLockBlocks
count
Number of lock requests that were blocked. This column is

calculated as follows:
FLOAT
Total locks - Locks blocked = Locks with immediate grants

FileLockDeadlocks
count
Number of deadlocks detected on lock requests.
FLOAT
FileLockEnters
count
Number of lock requests on disk segments.
FLOAT
Cylinder Management Overhead Events Columns

These columns identify the number of times the File System software performed a cylinder
management event. The table ResUsageIvpr further breaks down the I/Os associated with
these events. For more information, see Appendix C: ResUsageIvpr Table.
Column Name
Mode
Description
Data Type
FileCylAllocs
count
Number of new cylinders allocated.
FLOAT
FileCylDefrags
count
Number of cylinder defragments performed.
FLOAT
FileCylFrees
count
Number of logical or physical cylinders freed.
FLOAT
FileCylMigrs
count
Number of cylinder migrations.
FLOAT
141

Statistics Columns
Column Name
Mode
Description
Data Type
FileMCylPacks
count
Number of MiniCylPack operations performed.
FLOAT
Write Ahead Logging Columns

These columns identify the number of times the File System software performed a cylinder
management event associated with the WAL log.
Column Name
Mode
Description
Data Type
FileWCylAllocs
count
Number of new WAL cylinders allocated.
FLOAT
FileWCylFrees
count
Number of times the File System logically frees a cylinder.
FLOAT
FSG I/O Column

This column identifies the I/O statistics reported from the FSG.
Column Name
Mode
Description
Data Type
IoRespMax
max
Maximum I/O response time in milliseconds on an AMP.
FLOAT
FSG Cache Wait Columns

Column Name
Mode
Description
Data Type
FSGCacheWaits
count
Number of times the File System waits for memory to become

available in the file segment cache when trying to read data from
disk.
FLOAT
FSGCacheWaitTime
count
Total amount of time the File System waits for memory to become
available in the file segment cache when trying to read data from
disk.
FLOAT
FSGCacheWaitTimeMax
max
Maximum amount of time the File System waits for memory to

become available in the File System cache when trying to read data
from disk.
FLOAT
MI Columns
Column Name
Mode
Description
Data Type
MIWriteLocks
count
Number of write locks acquired on a MI.
FLOAT
MIWriteLockTime
count
Total MI write lock hold time in milliseconds.
FLOAT
MIWriteLockTimeMax
max
Maximum MI write lock hold time in milliseconds.
FLOAT
142

Statistics Columns
Column Name
Mode
Description
Data Type
MIWrites
count
Number of no modification (NOMOD) write locks acquired on

MI.
FLOAT
MIWriteTime
count
Total MI NOMOD write lock hold time in milliseconds.
FLOAT
MIWriteTimeMax
max
Maximum MI NOMOD write lock hold time in milliseconds.
FLOAT
MISleeps
count
Number of times spent waiting to get a lock on the MI.
FLOAT
MISleepTime
count
Total amount of time waited to get a lock on the MI.
FLOAT
MISleepTimeMax
max
Maximum amount of time waited to get a lock on the MI.
FLOAT
Data Block Merge Columns

These columns merge small data blocks into one large block so that subsequent merge
attempts will have less I/Os.
Column Name
Mode
Description
Data Type
DBMergeDone
count
Number of successful data block merges done. The number of

times the data block being modified successfully merged with some
number of adjacent data blocks as part of the modification.
FLOAT
Subtracting DBMergeDone from DBMergeTried will result in the

number of data block merge operations that were tried and failed.
DBMergeElim
count
Number of data blocks eliminated due to data block merges. If n

data blocks are merged into one large block, where n is the number
of data blocks, this number is incremented by n-1.
FLOAT
DBMergeExtraIO
count
Number of additional physical I/Os performed in the data block

merge process over and above those done if no data block merges
were attempted. This includes any extra physical I/Os that were
performed regardless of whether a particular merge attempt
succeeded or not.
FLOAT
DBMergeTried
count
Number of times the data block being modified tried to be merged

with adjacent data blocks as part of the modification.
FLOAT
AutoCylPack Columns
These columns track the number of cylinders used during AutoCylPack, a background task
that runs periodically to maintain the set levels of free space percent (FSP) on table cylinders.
For more information about AutoCylPack, see Database Administration.
Column Name
Mode
Description
Data Type
FileACPCylsMigr
count
Number of successful migrations performed by AutoCylPack.
FLOAT
143

Statistics Columns
Column Name
Mode
Description
Data Type
FileACPCylsPostponed
count
Number of cylinders AutoCylPack selected for performing a

migration, but could not be performed at the current time. This
can happen due to conflicts with foreground tasks modifying the
cylinder at around the same time as AutoCylPack. AutoCylPack,
therefore, postpones the work until the next time it scans the MI
from the beginning. When AutoCylPack comes across the cylinder
again, and it still qualifies, the cylinder is selected again for
processing.
FLOAT
FileACPCylsSkipped
count
Number of cylinders AutoCylPack skipped when scanning the MI

since nothing needed to be done.
FLOAT
FileACPCylsUnFSEOnly
count
Number of cylinders AutoCylPack selected for performing a

migration, but could not be performed at the current time. This
can happen because of a locking conflict or a cylinder being
modified or marked down. Instead, AutoCylPack cleans-up the
unfree sector entries (UnFSEs) on all cylinders, except for those
that are down.
FLOAT
BLC Columns
These columns collect various BLC statistics for use in performance analysis and debugging.
BLC enables data compression at the data block (DB) level of the Teradata Database file
system. Compression reduces the amount of storage required for a given amount of data. The
BlockLevelCompression field of DBS Control enables and disables BLC.
For more information on BLC and compression-related DBS Control settings, see Utilities.
Note: You must enable BLC to collect statistics from the columns below or zero will be
returned for each column.
Column Name
Mode
Description
Data Type
FileCompDBs
count
Total number of data blocks compressed.
FLOAT
FileCompCPU
count
Total compression time, including any overhead. The column

measures, in nanoseconds, the time it takes from the beginning
of the compression operation to the end.
FLOAT

FilePostCompMB
count
Total number of MBs of compressed data that result from

uncompressed blocks being compressed.
FLOAT
This column is used together with the FilePreCompMB column

to calculate the compression ratio.
FilePreCompMB
count
Total number of MBs for the data block that will be compressed
before any compression starts.
FLOAT
This column is used together with the FilePostCompMB column

to calculate the compression ratio.
144

Statistics Columns
Column Name
Mode
Description
Data Type
FileCompCylMigrs
count
Number of cylinder migrates done as part of a Ferret

COMPRESS command or by the AutoTempComp background
task.
FLOAT
The AutoTempComp background task is responsible for finding

the cylinders that should be compressed or decompressed based
on their temperature. For more information on this background
task, see Database Design.
For more information on the Ferret utility, COMPRESS
command, see Utilities.
FileCompFerretDBs
count
Number of blocks compressed as a result of the compression

operation called from the Ferret utility (see Utilities).
FLOAT
FileCompTempDBs
count
Number of data blocks compressed by the AutoTempComp

background task because they were colder than the DBS Control
TempBLCThresh field setting. See Database Design for more
information on the DBS Control TempBLCThresh field and
AutoTempComp background task.
FLOAT
Note: You must enable the temperature-based block-level

compression (TBBLC) feature when using this column or it will
return a zero. To determine if TBBLC is enabled, see
EnableTempBLC in Utilities.
FilePreUncompMB
count
Total number of MBs for the compressed data block that will be
uncompressed before uncompression starts.
FLOAT
This column is used together with the FilePostUncompMB

column to calculate the uncompression ratio.
FilePostUncompMB
count
Total number of MBs of uncompressed data that result from

compressed blocks being uncompressed.
FLOAT
This column is used together with the FilePreUncompMB

column to calculate the uncompression ratio.
FileTempCPU
count
CPU time spent by the AutoTempComp background task either

compressing or uncompressing data, including any overhead.
FLOAT
For more information on the AutoTempComp background task,

see Utilities.
Note: You must enable the temperature-based block-level
compression (TBBLC) feature when using this column or it will
return zero. To determine if TBBLC is enabled, see
EnableTempBLC in Utilities.
Also, any time counted against this column is not counted in
either the FileCompCPU or FileUncompCPU column.
Note: This column is valid on SLES 11 or later only.
FileUncompCPU
count
Total uncompression time, including any overhead. The column

measures, in nanoseconds, the time it takes from the beginning
of the uncompression operation to the end.
FLOAT
Note: This column is valid on SLES 11 or later only.
145

Statistics Columns
Column Name
Mode
Description
Data Type
FileUnCompCylMigrs
count
Number of cylinder migrates done as part of a Ferret

UNCOMPRESS command or by the AutoTempComp
background task.
FLOAT
For more information on the Ferret utility, UNCOMPRESS

command, or AutoTempComp background task, see Utilities.
FileUncompDBs
count
Total number of data blocks uncompressed.
FLOAT
FileUncompFerretDBs
count
Number of blocks uncompressed as a result of the

uncompression operation called from the Ferret utility (see
Utilities).
FLOAT
FileUnCompTempDBs
count
Number of data blocks uncompressed by the AutoTempComp

background task because they were warmer than the DBS
Control TempBLCThresh field setting. See Database Design for
more information on the DBS Control TempBLCThresh field
and AutoTempComp background task.
FLOAT
Note: You must enable the TBBLC feature when using this
column or it will return zero. To determine if TBBLC is enabled,
see EnableTempBLC in Utilities.
General Concurrency Control Database Locks Columns

These columns identify database locking activities.
Column Name
Mode
Description
Data Type
DBLockBlocks
count
Number of times a database lock was blocked.
FLOAT
DBLockEnters
count
Number of times a database lock was entered for holding.
FLOAT
DBLockDeadlocks
count
Number of times a database lock was deadlocked.
FLOAT
Memory Columns
Memory Resident Columns
These columns represent the amount of memory resident specific to virtual processor
activities, subdivided into memory types. The columns do not include any memory
allocations specific to the node the vproc is running under.
146

Statistics Columns
Disk memory segments can be in one of four states:
Clean (unmodified) and Unaccessed by any process (CU)
Dirty (modified) and Unaccessed (DU)
Clean and Accessed (CA)
Dirty and Accessed (DA)
Permanent segments for an entire table can be user-locked-in to memory. These are called
frozen segments (Frz), and no state subdivision is necessary because they cannot be aged or
forced out of memory.
Column Name
Mode
Description
Data Type
MemAPtKBResCA
track
Current KB resident in memory for append table or permanent

journal table data block or cylinder index disk segments that are
currently clean and accessed.
FLOAT

MemAPtKBResCU
track

currently clean and not accessed.
FLOAT

MemAPtKBResDA
track

currently dirty and accessed.
FLOAT

MemAPtKBResDU
track

currently dirty and unaccessed.
FLOAT

MemPCiKBResCA
track
Current KB resident in memory for permanent cylinder index disk

segments that are currently clean and accessed.
FLOAT

MemPCiKBResCU
track

segments that are currently clean and not accessed.
FLOAT

MemPCiKBResDA
track

segments that are currently dirty and accessed.
FLOAT

MemPCiKBResDU
track

segments that are currently dirty and unaccessed.
FLOAT
147

Statistics Columns
Column Name
Mode
Description
Data Type
MemPDbKBResCA
track
Current KB resident in memory for permanent data block segments

that are currently clean and accessed.
FLOAT

MemPDbKBResCU
track
Current KB resident in memory for permanent data block disk

segments that are currently clean and not accessed.
FLOAT

MemPDbKBResDA
track

segments that are currently dirty and accessed.
FLOAT

MemPDbKBResDU
track

segments that are currently dirty and unaccessed.
FLOAT

MemSCiKBResCA
track
Current KB resident in memory for regular or restartable spool

cylinder index disk segments that are currently clean and accessed.
FLOAT

MemSCiKBResCU
track

cylinder index disk segments that are currently clean and not
accessed.
FLOAT

MemSCiKBResDA
track

cylinder index disk segments that are currently dirty and accessed.
FLOAT

MemSCiKBResDU
track

cylinder index disk segments that are currently dirty and
unaccessed.
FLOAT

MemSDbKBResCA
track
Current KB resident in memory for regular or restartable spool data

block disk segments that are currently clean and accessed.
FLOAT

MemSDbKBResCU
track

block disk segments that are currently clean and not accessed.
FLOAT

MemSDbKBResDA
track

block disk segments that are currently dirty and accessed.
FLOAT

MemSDbKBResDU
track

block disk segments that are currently dirty and unaccessed.
FLOAT
148

Statistics Columns
Column Name
Mode
Description
Data Type
MemTJtKBResCA
track
Current KB resident in memory for transient journal table or WAL

data block or WAL cylinder index disk segments that are currently
clean and accessed.
FLOAT

MemTJtKBResCU
track

clean and not accessed.
FLOAT

MemTJtKBResDA
track

dirty and accessed.
FLOAT

MemTJtKBResDU
track

dirty and unaccessed.
FLOAT
Memory Allocation Column

Column Name
Mode
Description
Data Type
MemCtxtAllocs
count
Number of successful memory allocations and size-increasing

memory alters on task context pages.
FLOAT
Note: Only scratch pages are allocated. All other task context pages
appear resident at some point soon after component restart.
Task Context Segment Usage Columns

These columns identify the usage of task context segments and how they leave memory.
Column Name
Mode
Description
Data Type
MemCtxtAccesses
count
Number of segments accessed.
FLOAT
MemCtxtAccessKB
count
KB of segments accessed.
FLOAT
MemCtxtDeaccesses
count
Number of segments deaccessed.
FLOAT
Deaccessed segments remain in memory until paged out through

aging.
MemCtxtDeaccessKB
count
KB of segments deaccessed.
FLOAT
MemCtxtDestroyKB
count
KB of segments destroyed.
FLOAT
MemCtxtDestroys
count
Number of segments destroyed.
FLOAT
Destroyed segments are immediately dropped from memory.
149

Statistics Columns
Net Columns
and output messages passing through either Teradata Database net through point-to-point
(1:1) methods (PtP).
Column Name
Mode
Description
Data Type
NetPtPReadKB
count
Total KB of point-to-point messages input to the vproc.
FLOAT
NetPtPReads
count
Number of point-to-point messages input to the vproc.
FLOAT
NetPtPWriteKB
count
Total KB of point-to-point messages output to the vproc.
FLOAT
NetPtPWrites
count
Number of point-to-point messages output from the vproc.
FLOAT

methods (Brd) per net.
Column Name
Mode
Description
Data Type
NetBrdReadKB
count
Total KB of broadcast messages input to the vproc.
FLOAT
NetBrdReads
count
Number of broadcast messages input to the vproc.
FLOAT
NetBrdWriteKB
count
Total KB of broadcast messages output from the vproc.
FLOAT
NetBrdWrites
count
Number of broadcast messages output from the vproc.
FLOAT
Work Mailbox Queue Columns

These columns identify the virtual processor work mailbox queue length where requested
work awaits the allocation of a process to perform the work.
Column Name
Mode
Description
Data Type
MsgWorkQLen
track
Total number of work requests waiting at the current time.
FLOAT
MsgWorkQLenMax
max
Maximum number of work requests waiting during each log

interval.
FLOAT
The Max count, unlike the MsgWorkQLen field, tracks the

maximum count over the log period. Therefore, the MsgWorkQLen
field can be zero even though the Max count can be non-zero over
the log period.
150

Statistics Columns

ChnSignal Status Tracking Column
Column Name
Mode
Description
Data Type
MsgChnLastDone
count
Number of last done events that occurred for this vproc.
FLOAT
Note: The last AMP to finish an operation may send a last done
broadcast message indicating the work is done for this step. This is
used in tracking down the slowest AMP in the system. An AMP that
has more last done messages than the others could be a bottleneck
in the system performance.

These columns represent CPU activities associated with this virtual processor, subdivided into
48 partitions. Partition 0 is reserved for use by PDE processes. For more information on the
other partitions see Appendix E: Partition Assignments.
For definitions of user service and user execution, see Process Scheduling Columns in the
Chapter 5: ResUsageScpu Table.
Column Name
Mode
Description
Data Type
CPUUServPart00 CPUUServPart48
count
Time in centiseconds CPUs are busy in the partition doing user

service. This is the system level time spent on a process.
FLOAT
CPUUExecPart00 CPUUExecPart48
count
Time in centiseconds CPUs are busy in the partition doing user

execution. This is the user level time spent on a process.
FLOAT
Note: Some CPU times may not be reported to these fields if child
processes or threads are running on your system. For information
about the difference in CPU usage reported and possible cause, see
CPU Utilization Columnsin Chapter 11: ResUsageSps.
Cylinder Read Columns

These columns represent the File System resource usage statistics. The Cylinder Read feature
uses these statistics for tracking performance and utilization.
151

Statistics Columns
Column Name
Mode
Description
Data Type
FileFcrRequests
count
Total number of requests for the File System to use Cylinder

Read.
FLOAT
This column records the number of attempts to use Cylinder

Read independent of whether the request will be issued to FSG or
not. A request can be denied due to insufficient data blocks or
because there is insufficient space in the FSG cache. Requests can
also be denied at both the user and kernel level. Each of these
items is counted in other FileFcr resource usage columns.
This column can perform a number of calculations:
Requests issued to FSG =
FileFcrRequests - FileFcrDeniedUser
Successful Cylinder Reads =
FileFcrRequests - FileFcrDeniedUser - FileFcrDeniedKern
FileFcrRequestsAdaptive
count
Number of adaptive requests from File System.
FLOAT
This column records the number of requests for adaptive-style

Cylinder Reads.
Note: This column is not currently used.
FileFcrBlocksDeniedCache
count
Number of data blocks contained in Cylinder Read requests

rejected by the FSG subsystem due to insufficient cache.
FLOAT
This column records the number of data blocks that were part of
attempts to use Cylinder read that were denied by the FSG
subsystem due to insufficient cache space; therefore, also
incremented the FileFcrDeniedCache column.
FileFcrBlocksDeniedKern
count
Number of data blocks contained in the FSG subsystem rejected

requests for Cylinder Read.
FLOAT
attempts to use Cylinder Read that were denied by the FSG
subsystem, and therefore, also incremented the
FileFcrDeniedKern column.
FileFcrBlocksDeniedUser
count
Number of data blocks contained in the File System rejected

requests for Cylinder Read.
FLOAT
attempts to use Cylinder Read that were denied by the File
System.
FileFcrBlocksDeniedThreshKern
count

rejected for threshold by the FSG subsystem.
FLOAT
attempts to use Cylinder read that were denied by the FSG
subsystem due to the number of blocks being below the
threshold, and therefore, also incremented the
FileFcrDeniedThreshKern column.
152

Statistics Columns
Column Name
Mode
Description
Data Type
FileFcrBlocksDeniedThreshUser
count

rejected for threshold by the File System.
FLOAT
attempts to use Cylinder Read that were denied by the File System
due to the number of blocks being below the threshold, and
therefore, also incremented the FileFcrDeniedThreshUser
column.
FileFcrBlocksRead
count
Number of data blocks read in using Cylinder Read.
FLOAT
This column records the total number of data blocks read in by

successful Cylinder Read operations.
The average number of data blocks in a successful Cylinder read
can be calculated as follows:
Average data blocks/ Cylinder Read = FileFcrBlocksRead /
(FileFcrRequests - FileFcrDeniedUser - FileFcrDeniedKern)
FileFcrDeniedCache
count
Number of Cylinder Read requests denied by FSG due to

insufficient cache.
FLOAT
This column records the number of Cylinder Read requests

denied due to insufficient FSG cache space for a cylinders worth
of data.
FileFcrDeniedKern
count
Number of Cylinder Read requests denied by the FSG subsystem.
FLOAT

issued to the FSG subsystem which, for any reason, have been
denied. A request can be denied due to insufficient data blocks
(for example, the FileFcrDeniedThreshKern column) or because
there is insufficient space in the FSG cache (for example, the
FileFcrDeniedCache column). The FSG subsystem can reject a
request containing insufficient data blocks that the File System
thought had enough blocks because the FSG subsystem reduces
the count by the number of data blocks that are already resident
in the cache.
FileFcrDeniedUser
count
Number of Cylinder Read requests denied by the File System.
FLOAT
This column records the number of Cylinder Read attempts

denied by the File System. A request can be denied by the File
System due to insufficient number of data blocks being requested
(for example, the FileFcrDeniedThreshUser column). For
information, see the FileFcrDeniedThreshUser column.
153

Statistics Columns
Column Name
Mode
Description
Data Type
FileFcrDeniedThreshKern
count
Number of Cylinder Read requests denied by the FSG subsystem

due to insufficient data blocks.
FLOAT

denied due to the data block threshold criteria. There is a
minimum threshold of data blocks for an individual Cylinder
Read request. If the number of data blocks is below this
threshold, the overhead of the Cylinder Read operation is
considered too large and issuing individual data block reads is
considered more efficient. Therefore, the Cylinder Read request
is denied. FSG must reevaluate the threshold for a request that
the File System considered valid since FSG eliminates any data
blocks from the request list that are already resident in the cache.
This could reduce the count that the File System thought was
above the threshold to one that is now below.
FileFcrDeniedThreshUser
count
Number of Cylinder Read requests denied by the File System due

to insufficient data blocks.
FLOAT

denied due to the data block threshold criteria. There is a
minimum threshold of data blocks for an individual Cylinder
Read request. If the number of data blocks is below this
threshold, the overhead of the Cylinder Read operation is
considered too large and issuing individual data block reads is
considered more efficient. Therefore, the Cylinder Read request
is denied.
PE and AMP UDF CPU Columns

These columns report the system level and user execution UDF CPU time value under the
AMP and PE vprocs. They also provide information about whether the UDFs were doing
work for the AMP or PE vprocs.
The data reported by CPUUExecPart00 for the NODE vproc includes the UDF CPU time for
all UDFs running on the node.
All UDFs are invoked by either a PE or an AMP, but PDE reports UDF CPU usage to the Node
partition by design, not the associated AMP and PE partitions. Therefore, the
CPUUExecPart00 and CPUUServPart00 columns will report the UDF CPU usage and the
CPUUServOrExecPart11 and 13 will not, where ServOrExec is either the user service (Serv) or
user execution (Exec) partition. RSS code reports the UDF CPU usage by AMP and PE in the
following UDF CPU columns:
UDFAMPExec
UDFAMPServ
UDFPEExec
UDFPEServ
Note: The total of these UDF CPU columns can exceed 100% of elapsed centiseconds due to
the CPU data gathering operation. The UDF CPU reporting for the ResUsageSvpr table is
reported when the UDF completes an operation. A long running UDF will report all the CPU
154

Statistics Columns
time at one time and that may be significantly larger than the current reporting period. This
can cause spikes in the UDF CPU reporting columns.
The UDF CPU time value over multiple periods averages 100% or less.
Note: When an external routine (such as C, C++, Java UDF, or external stored procedure)
forks a child process or thread, the CPU time is not reported to these fields. As a result, the
resource usage table shows a lower CPU usage than shown in the ResUsageSpma table even if
the external routine consumes a large amount of CPU time. If there are child processes or
threads running on your system, this may account for the larger CPU times reported in the
ResUsageSpma table compared to this table. To confirm that the difference in CPU usage
reported by the ResUsageSpma table is caused by child processes or threads, contact your
Teradata Support Center personnel.
Column Name
Mode
Description
Data Type
UDFAMPServ
count
Reported system-level UDF CPU time value under the AMP vproc.
FLOAT
UDFAMPExec
count
Reported user execution UDF CPU time value under the AMP
vproc.
FLOAT
UDFPEServ
count
Reported system-level UDF CPU time value under the PE vproc.
FLOAT
UDFPEExec
count
Reported user execution UDF CPU time value under the PE vproc.
FLOAT
Process Block Count Columns

These columns identify how many times a process became blocked on which blocking type.
Column Name
Mode
Description
Data Type
ProcBlksDBLock
count
Number of process blocks for database locks.
FLOAT
ProcPendDBLock
track
Number of process blocks pending database locks.
FLOAT

Column Name
Mode
Description
Data Type
ProcWaitDBLock
count
Total time in centiseconds processes were blocked pending database

locks.
FLOAT
Teradata VS Columns
Allocation Columns
These columns identify the allocation statistics reported by the Allocator.
155

Statistics Columns
Column Name
Mode
Description
Data Type
AllocatorExtentAllocReqs
count
Number of cylinder allocation requests received by the allocator.
FLOAT
AllocatorExtentFreeReqs
count
Number of cylinder free requests received by the allocator.
FLOAT
AllocatorMapIOsStarted
count
Number of map I/Os initiated by the allocator.
FLOAT
AllocatorMapIOsDone
count
Number of map I/Os completed by the allocator.
FLOAT
Extent Driver I/O Columns

These columns identify the I/O statistics reported from the extent driver.
Column Name
Mode
Description
Data
Type
ReadsCold
count
Total number of reads issued to all cylinders that are considered COLD.
FLOAT
A cylinder is considered to be COLD if the response time of its

associated physical storage is less than TVS Percentile Cold Lower
Bound configuration setting.
ReadsHot
count
Total number of reads issued to all cylinders that are considered HOT.
FLOAT
A cylinder is considered HOT if the response time of its associated

physical storage is greater than the Teradata VS Percentile Hot Lower
Bound configuration setting.
ReadsWarm
count
Total number of reads issued to all cylinders that are considered

WARM.
FLOAT
A cylinder is considered to be WARM if the response time of its

associated physical storage is between TVS Percentile Hot Lower
Bound and TVS Percentile Cold Upper Bound configuration settings.
WritesCold
count
Total number of writes issued to all cylinders that are considered

COLD.
FLOAT

WritesHot
count
Total number of writes issued to all cylinders that are considered HOT.
FLOAT

WritesWarm
count
Total number of writes issued to all cylinders that are considered

WARM.
FLOAT

ReadResponseColdTotal
count
Total read response time of all cylinders that are considered COLD
FLOAT
156

Statistics Columns
Column Name
Mode
Description
Data
Type
ReadResponseHotTotal
count
Total read response time of all cylinders that are considered HOT.
FLOAT

ReadResponseWarmTotal
count
Total read response time of all cylinders that are considered WARM.
FLOAT

WriteResponseColdTotal
count
Total write response time of all cylinders that are considered COLD.
FLOAT

WriteResponseHotTotal
count
Total write response time of all cylinders that are considered HOT.
FLOAT

WriteResponseWarmTotal
count
Total write response time of all cylinders that are considered WARM.
FLOAT

ReadResponseColdMax
max
Maximum read response time of all cylinders that are considered

COLD.
FLOAT

ReadResponseHotMax
max
Maximum read response time of all cylinders that are considered HOT.
FLOAT

ReadResponseWarmMax
max
Maximum read response time of all cylinders that are considered

WARM.
FLOAT

ReadResponseColdMin
min
Minimum read response time of all cylinders that are considered

COLD.
FLOAT
This column has a default value of the largest value of a 64-bit integer
column.
ReadResponseHotMin
min
Minimum read response time of all cylinders that are considered HOT.
FLOAT
column.
ReadResponseWarmMin
min
Minimum read response time of all cylinders that are considered

WARM.
FLOAT
column.
WriteResponseColdMax
max
Maximum write response time of all cylinders that are considered

COLD.
FLOAT
157

Statistics Columns
Data
Type
Column Name
Mode
Description
WriteResponseHotMax
max

HOT.
FLOAT

WriteResponseWarmMax
max

WARM.
FLOAT

WriteResponseColdMin
min
Minimum write response time of all cylinders that are considered

COLD.
FLOAT
column.
WriteResponseHotMin
min

HOT.
FLOAT
column.
WriteResponseWarmMin
min

WARM.
FLOAT
column.
Node Agent Columns

Note: The NodeAgentMigrationsStarted and NodeAgentMigrationsDone columns are
populated only in Teradata VS. For more information on these columns and on TVS, see
Teradata Virtual Storage.
Column Name
Mode
Description
Data Type
NodeAgentMigrationsDone
count
Number of migration requests completed by the Node Agent.
FLOAT
NodeAgentMigrationsStarted
count
Number of migration requests started by the Node Agent.
FLOAT
Reserved Column
Column Name
Mode
Description
Data Type
Reserved
n/a
CHAR (3)
158

Summary Mode
Summary Mode
When Summary Mode is active for the ResUsageSvpr table, one row is written to the database
for each type of vproc on each node in the system, summarizing the vprocs of that type on
that node, for each log interval.
that row.
'S'
in Summary Mode.
'N'
normally.
Spare Columns
The ResUsageSvpr table spare fields are named Spare00 through Spare19, and SpareInt.
Column Name
Description
Spare00
Number of data segments that were aged out of VH cache. This field
is populated by the FSG subsystem.
Note: This field will be converted to the VHAgedOut column in
field using the VHAgedOut column name in the ResSvprView view.
For details, see ResSvprView on page 179.
Spare01
Volume of data segments in KB that were aged out of VH cache. This

field is populated by the FSG subsystem.
Note: This field will be converted to the VHAgedOutKB column in
field using the VHAgedOutKB column name in the ResSvprView
view. For details, see ResSvprView on page 179.
159

Spare Columns
Column Name
Description
Spare02
Number of logical reads from VH cache. This field is populated by the

FSG subsystem.
Note: This field will be converted to the VHLogicalDBRead column
in Teradata Database 15.0. You can access resource usage data for this
field using the VHLogicalDBRead column name in the ResSvprView
Spare03
Volume of logical reads in KB from VH cache. This field is populated

by the FSG subsystem.
Note: This field will be converted to the VHLogicalDBReadKB
column in Teradata Database 15.0. You can access resource usage data
for this field using the VHLogicalDBReadKB column name in the
ResSvprView view. For details, see ResSvprView on page 179.
Spare04
Number of very hot reads that were handled by physical disk I/O due
to a VH cache miss (that is, data not found in the VH cache). This
field is populated by the FSG subsystem.
Note: This field will be converted to the VHPhysicalDBRead column
in Teradata Database 15.0. You can access resource usage data for this
field using the VHPhysicalDBRead column name in the ResSvprView
Spare05
Volume of very hot reads in KB that were handled by physical disk

I/O due to a VH cache miss. This field is populated by the FSG
subsystem.
Note: This field will be converted to the VHPhysicalDBReadKB
column in Teradata Database 15.0. You can access resource usage data
for this field using the VHPhysicalDBReadKB column name in the
ResSvprView view. For details, see ResSvprView on page 179.
Spare06

field using the WM_CPU_COD column name in the ResSvprView
1000 on SLES 10.
160

Spare Columns
Column Name
Description
Spare07
Size of VH cache in KB currently in use. This field is populated by the

FSG subsystem.
Note: This field will be converted to the VHCacheInuseKB column in
field using the VHCacheInuseKB column name in the ResSvprView
Related Topics
161

Spare Columns
162
CHAPTER 14
Resource Usage Views
This chapter provides the definitions of the resource usage views.

Note: Views are the recommended method to access the resource usage table data.
To see the view definitions, execute SHOW VIEW viewname, where viewname is the name of
the view whose most recent SQL create text is to be reported. For details on using the SHOW
VIEW statement, see SQL Data Definition Language.
The following views report the table column, GroupId. A homogenous system requires no
changes to use this macro because all the nodes will be assigned to group A. For a coexisting
system, however, the values need to be set up when the system is installed or reconfigured so
that each type of node is assigned to a specific group ID.
Since each of the views include all the columns for the represented table, the views will be
changed whenever the base table columns are changed.
Notice:
Do not change or delete columns in these views. If the columns are modified, the resource
usage macros that use these views may not work properly. You can, however, safely add
columns.
163
Chapter 14: Resource Usage Views

ResCPUUsageByAMPView
ResCPUUsageByAMPView describes CPU usage per AMP.
REPLACE VIEW DBC.ResCPUUsageByAMPView
AS SELECT
TheDate,
TheTime,
VprId,
VprId
as Vproc,
NodeID,
Secs,
NCPUs,
/* GroupId */
/* Changes in GroupId definition affects the displayed grouping in
* the Res*ByGroup macros. The default setting below shows how to
* select different node families, but does not differentiate the
* resulting groups. If a coexistence system had 5550, 5600 and 5650 nodes
* the CASE expression might look like:
* WHEN NodeType
IN ('5650H') THEN '5650Nodes'
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
/*
/*
/*
/*
164
CPUUExecPart11 AS AMPWorkTaskExec,
CPUUServPart11 AS AMPWorkTaskServ,
AMPMiscUserExec*/
(
CPUUExecPart01+CPUUExecPart02+CPUUExecPart03 +
CPUUExecPart04+CPUUExecPart05+CPUUExecPart06+CPUUExecPart07 +
CPUUExecPart44+CPUUExecPart45+CPUUExecPart46+CPUUExecPart47) AS AMPMiscUserExec,
AMPMiscUserServ */
(
CPUUServPart01+CPUUServPart02+CPUUServPart03 +
CPUUServPart04+CPUUServPart05+CPUUServPart06+CPUUServPart07 +
CPUUServPart44+CPUUServPart45+CPUUServPart46+CPUUServPart47) AS AMPMiscUserServ,
AMPTotalUserExec */
(CPUUExecPart00+CPUUExecPart01+CPUUExecPart02+CPUUExecPart03 +
CPUUExecPart44+CPUUExecPart45+CPUUExecPart46+CPUUExecPart47) AS AMPTotalUserExec,
AMPTotalUserServ */
(CPUUServPart00+CPUUServPart01+CPUUServPart02+CPUUServPart03 +

CPUUServPart44+CPUUServPart45+CPUUServPart46+CPUUServPart47)
AS AMPTotalUserServ
FROM DBC.ResUsageSvpr WHERE VprType like 'AMP%' WITH CHECK OPTION;
165

ResCPUUsageByPEView
ResCPUUsageByPEView
ResCPUUsageByPEView describes CPU usage by each PE.
EPLACE VIEW DBC.ResCPUUsageByPEView
AS SELECT
TheDate,
TheTime,
VprId,
VprId
AS Vproc,
NodeID (FORMAT'999-99') AS NodeId,
Secs, NCPUs,
/*
*
*
*
Changes in GroupId definition affects the displayed grouping in

the Res*ByGroup macros. The default setting below shows how to
select different node families, but does not differentiate the
resulting groups. If a coexistence system had 5550, 5600 and
5650 nodes
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
CPUUExecPart13
CPUUServPart13
CPUUExecPart12
CPUUServPart12
AS
AS
AS
AS
PEDispExec,
PEDispServ,
PESessExec,
PESessServ,
/* PEMiscUserExec */
(
CPUUExecPart14+CPUUExecPart15 +
CPUUExecPart44+CPUUExecPart45+CPUUExecPart46+CPUUExecPart47) AS PEMiscUserExec,
/* PEMiscUserServ */
(
CPUUServPart14+CPUUServPart15 +
CPUUServPart44+CPUUServPart45+CPUUServPart46+CPUUServPart47) AS PEMiscUserServ,
/* PETotalUserExec */
(CPUUExecPart00+CPUUExecPart01+CPUUExecPart02+CPUUExecPart03 +
CPUUExecPart44+CPUUExecPart45+CPUUExecPart46+CPUUExecPart47) AS PETotalUserExec,
/* PETotalUserServ */
(CPUUServPart00+CPUUServPart01+CPUUServPart02+CPUUServPart03 +
166

ResCPUUsageByPEView
CPUUServPart44+CPUUServPart45+CPUUServPart46+CPUUServPart47)
AS PETotalUserServ
FROM DBC.ResUsageSvpr WHERE VprType like 'PE%' WITH CHECK OPTION;
167

ResSawtView
ResSawtView
ResSawtView is based on the ResUsageSawt table.
/* housekeeping fields */
TheDate,
NodeID (FORMAT '999-99') AS NodeID,
TheTime,
GmtTime,
NodeType,
TheTimestamp,
CentiSecs,
Secs,
NominalSecs,
CodFactor,
SummaryFlag,
Reserved,
VprId,
/* Aliased fields */
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,
/* PM CPU Capacity On Demand factor */
/* Spare Field usage */

/* PM/WM CODs */
Spare00
AS WM_CPU_COD,
/* WM CPU Capacity On Demand factor */
/* transformed fields */
/* PM/WM CODs */
( PM_CPU_COD * WM_CPU_COD / 1000 ) (FORMAT '----9') AS CPU_COD, /* effective CPU COD */
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
( WorkTypeInuse00
WorkTypeInuse04
WorkTypeInuse08
WorkTypeInuse12
+
+
+
+
WorkTypeInuse01
WorkTypeInuse05
WorkTypeInuse09
WorkTypeInuse13
AS
/* Remaining table fields */
+ WorkTypeInuse02
+ WorkTypeInuse06
+ WorkTypeInuse10
+ WorkTypeInuse14
WorkTypeInuse,
+
+
+
+
WorkTypeInuse03
WorkTypeInuse07
WorkTypeInuse11
WorkTypeInuse15
+
+
+
)
FROM DBC.ResUsageSawt WITH CHECK OPTION;
Note: The ResUsageSawt table fields have been removed from this sample output.
168

ResScpuView
ResScpuView
ResScpuView is based on the ResUsageScpu table.
thedate,
NodeID
(FORMAT
thetime,
GmtTime,
NodeType
(FORMAT
TheTimestamp,
CentiSecs
(FORMAT
Secs
(FORMAT
NominalSecs
(FORMAT
CodFactor
(FORMAT
SummaryFlag
(FORMAT
Reserved
(FORMAT
cpuid,
Reserved2
(FORMAT
'999-99')
AS NodeID,
'X(8)')
AS NodeType,
'-------9')
'----9')
'ZZZ9')
'-----9')
'X(1)')
'X(1)')
AS
AS
AS
AS
AS
AS
'X(2)')
AS Reserved2,
CentiSecs,
Secs,
NominalSecs,
CodFactor,
SummaryFlag,
Reserved,
/* Aliased Fields */
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,

/* PM/WM CODs */
Spare00
AS WM_CPU_COD,
/* PM/WM CODs */
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
FROM DBC.ResUsageScpu WITH CHECK OPTION;
Note: The ResUsageScpu table fields have been removed from this sample output.
169

ResShstView
ResShstView
ResShstView is based on the ResUsageShst table.
REPLACE VIEW DBC.ResShstView
AS SELECT
TheDate,
TheTime,
GmtTime,
NodeType,
TheTimestamp,
CentiSecs,
Secs,
NominalSecs,
CodFactor,
SummaryFlag,
Reserved,
IPaddr,
HstType,
vprid,
hstid (FORMAT '----------9'),
/* PM/WM CODs */
CodFactor
AS
PM_CPU_COD,

/* PM/WM CODs */
Spare00
AS WM_CPU_COD,
/* PM/WM CODs */
/*
*
*
*
*
*
*
*
*
*
ECS connections have IP address values.

IPv4 has 4 byte addresses and IPv6 allows 16 byte addresses
First 4 bytes are stored in IPaddr integer field
Second 4 bytes are stored in VprId integer field
Convert to standard 256.256.256.256 byte representation
by displaying each byte as a separate integer.
BITAND not available until DIPSYSFNC runs, which is after DIPRUM.
Request moving DIPRUM to after DIPSYSFNC in TD14.10 release.
SHIFTRIGHT(
SHIFTRIGHT(
SHIFTRIGHT(
SHIFTRIGHT(
SHIFTRIGHT(
SHIFTRIGHT(
SHIFTRIGHT(
SHIFTRIGHT(
*/
BITAND(
BITAND(
BITAND(
BITAND(
BITAND(
BITAND(
BITAND(
BITAND(
IPaddr,
IPaddr,
IPaddr,
IPaddr,
VprId,
VprId,
VprId,
VprId,
'000000FF'XB
'0000FF00'XB
'00FF0000'XB
'FF000000'XB
'000000FF'XB
'0000FF00'XB
'00FF0000'XB
'FF000000'XB
), 0)
), 8)
),16)
),24)
), 0)
), 8)
),16)
),24)
(Format
(Format
(Format
(Format
(Format
(Format
(Format
(Format
'ZZ9')
'ZZ9')
'ZZ9')
'ZZ9')
'ZZ9')
'ZZ9')
'ZZ9')
'ZZ9')
AS
AS
AS
AS
AS
AS
AS
AS
IP0,
IP1,
IP2,
IP3,
IP4,
IP5,
IP6,
IP7,

* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
FROM
170
ResUsageShst
WITH CHECK OPTION;

ResSldvView
Note: The ResUsageShst table fields have been removed from this sample output.
ResSldvView
ResSldvView is based on the ResUsageSldv table.
REPLACE VIEW DBC.ResSldvView
AS SELECT
thedate,
NodeID,
thetime,
GmtTime,
NodeType
(FORMAT 'X(8)')
TheTimestamp,
CentiSecs
(FORMAT '-------9')
Secs
(FORMAT '----9')
NominalSecs
(FORMAT 'ZZZ9')
CodFactor
(FORMAT '-----9')
SummaryFlag
(FORMAT 'X(1)')
Reserved
(FORMAT 'X(1)')
ctlid,
ldvid,
LdvType
(FORMAT 'X(4)')
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,
AS NodeType,
AS
AS
AS
AS
AS
AS
CentiSecs,
Secs,
NominalSecs,
CodFactor,
SummaryFlag,
Reserved,
AS LdvType,
CASE
WHEN Reserved IN ('S') THEN 'SSD'
WHEN Reserved IN ('H') THEN 'HDD'
ELSE '---'
END AS LdvKind,
/* PM/WM CODs */
LdvReadRespMax
AS WM_CPU_COD, /* WM CPU Capacity On Demand factor, was invalid field */
LdvWriteRespMax AS WM_IO_COD, /* WM IO Capacity On Demand factor, was invalid field */
Spare00
AS Major, /* device major number */
Spare01
AS Minor, /* device minor number */
/* I/O Hard Limits and I/O Capacity On Demand (COD) */
Spare02
AS FullPotentialIota, /* Device Full Pontential iotas
*/
Spare03
AS CodPotentialIota,
/* Device Pontential iotas (after COD) */
Spare04
AS UsedIota,
/* Device used iotas
*/
/* PM/WM CODs */
SpareInt
AS PM_IO_COD,
/* PM IO Capacity On Demand factor
*/
/* PM/WM CODs */
( PM_IO_COD * WM_IO_COD / 100 ) (FORMAT '----9') AS IO_COD, /* effective IO COD */
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
LdvOutReqSum / NULLIFZERO(LdvOutReqDiv) AS LdvOutReqAvg,
171

ResSpdskView
FROM
DBC.ResUsageSldv
WITH CHECK OPTION;
Note: The ResUsageSldv table fields have been removed from this sample output.
ResSpdskView
ResSpdskView is based on the ResUsageSpdsk table.
REPLACE VIEW DBC.ResSpdskView
AS SELECT
thedate,
NodeID
thetime,
GmtTime,
NodeType
TheTimestamp,
CentiSecs
Secs
NominalSecs
CodFactor
SummaryFlag
Reserved
PdiskGlobalId,
PdiskDeviceId,
PdiskType
(FORMAT '999-99')
AS NodeID,
(FORMAT 'X(8)')
AS NodeType,
(FORMAT
(FORMAT
(FORMAT
(FORMAT
(FORMAT
(FORMAT
AS
AS
AS
AS
AS
AS
'-------9')
'----9')
'ZZZ9')
'-----9')
'X(1)')
'X(1)')
(FORMAT 'X(4)')
CentiSecs,
Secs,
NominalSecs,
CodFactor,
SummaryFlag,
Reserved,
AS PdiskType,
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,

/* PM/WM CODs */
Spare00
AS WM_CPU_COD,
Spare01
AS WM_IO_COD,
/* WM IO
Capacity On Demand factor */
SpareInt
AS PM_IO_COD,
/* PM IO
Capacity On Demand factor */
/* PM/WM CODs */
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
FROM DBC.ResSpdskView WITH CHECK OPTION;
Note: The ResUsageSpdsk table fields have been removed from this sample output.
172

ResSpmaView
ResSpmaView
ResSpmaView provides an overview of system operation.
Note: The GroupId column in this view provides a grouping for PE-only nodes.
REPLACE VIEW DBC.ResSpmaView
AS SELECT
TheDate,
TheTime,
GmtTime,
NodeType,
TheTimestamp,
CentiSecs,
Secs,
NominalSecs,
CodFactor,
NCPUs,
Reserved,
Vproc1,
VprocType1,
Vproc2,
VprocType2,
Vproc3,
VprocType3,
Vproc4,
VprocType4,
Vproc5,
VprocType5,
Vproc6,
VprocType6,
Vproc7,
VprocType7,
MemSize,
NodeNormFactor,
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,

Spare00
AS FilePreReadKB, /* long until TD15.0 */
Spare01
AS FileWriteKB,
/* long until TD15.0 */
Spare02
AS FileContigWIos,
Spare03
AS FileContigWBlocks,/* long until TD15.0 */
Spare04
AS FileContigWKB,
/* SLES11 Hard Limits Capacity On Demand (COD) */
Spare05
AS CpuThrottleCount,
/* Control Group throttle count */
Spare06
AS CpuThrottleTime,
/* Control Group: centi-seconds */
Spare07
AS FullPotentialIota, /* Total Full Pontential iotas
Spare08
/* Total Pontential iotas (after COD)
Spare09
AS UsedIota,
/* Total used iotas
/* TIM stats */
Spare10
As VHCacheKB, /* VH Cache size */
/* PM/WM CODs */
Spare11
AS WM_CPU_COD,
/* WM CPU Capacity On Demand factor
Spare12
AS WM_IO_COD,
/* WM IO Capacity On Demand factor
SpareInt
AS PM_IO_COD,
*/
*/
*/
*/
*/
*/
/*PM/WM CODs */
/*
*
*
*
*
*
*
*
*

the ResNodeByGroup macro. The default setting below shows how to
resulting groups. If a coexistence system had 5550, 5600 and 5650 nodes
as well as some PE-only nodes the CASE expression might look like:
WHEN (VPROCTYPE1='AMP' AND NodeType
IN ('5650H')) THEN '5650Nodes'
WHEN (VPROCTYPE1='AMP' AND NodeType NOT IN ('5600H', '5650')) THEN '5550Nodes'
ELSE 'PEonly'
173

ResSpmaView
*/
CASE
IN ('5600H')) THEN 'AMPNodes'
WHEN (VPROCTYPE1='AMP' AND NodeType NOT IN ('5600H')) THEN 'AMPNodes'
ELSE 'PEonly'
END AS GroupId,
(
(
(
(
(
(
(
(
(CPUUServ + CPUUExec) / NULLIFZERO(NCPUs) ) AS CPUBusy,

CPUUServ / NULLIFZERO(NCPUs) )
AS CPUOpSys,
CPUUExec / NULLIFZERO(NCPUs) )
AS CPUUser,
CPUIoWait / NULLIFZERO(NCPUs) )
AS CPUWaitIO,
(CPUUServ + CPUUExec) * (NodeNormFactor / 100)/ NULLIFZERO(NCPUs) ) AS CPUBusyNorm,
CPUUServ * (NodeNormFactor / 100) / NULLIFZERO(NCPUs) )
AS CPUOpSysNorm,
CPUUExec * (NodeNormFactor / 100) / NULLIFZERO(NCPUs) )
AS CPUUserNorm,
CPUIoWait* (NodeNormFactor / 100) / NULLIFZERO(NCPUs) )
AS CPUWaitIONorm,
( FileAcqReads + FilePreReads + FileWrites ) AS DiskSegmentIO,

( FileAcqReadKB + FilePreReadKB +
/* paging or swapping count times pagesize (= 4K) */
(MemTextPageReads + MemCtxtPageReads ) * 4 ) AS LogicalDeviceReadKB,
( FileWriteKB +
/* paging or swapping count times pagesize (= 4K) */
MemCtxtPageWrites * 4 ) AS LogicalDeviceWriteKB,
( LogicalDeviceReadKB + LogicalDeviceWriteKB ) AS LogicalDeviceIOKB,
( MemTextPageReads + MemCtxtPageReads + MemCtxtPageWrites )*4 AS PageOrSwapIOKB,
( NetTxCircPtP
+ NetTxCircBrd ) (FORMAT '------9') AS NetAttempts,
( NetMsgBrdReads
+ NetMsgBrdWrites )
AS NetMultiIO,
( NetMsgPtPReads
+ NetMsgPtPWrites )
AS NetPtoPIO,
( NetMsgBrdReadKB + NetMsgPtPReadKB )
AS NetReadKB,
( NetMsgBrdReads
+ NetMsgPtPReads )
AS NetReads,
( NetMsgBrdWriteKB + NetMsgPtPWriteKB )
AS NetWriteKB,
( NetMsgBrdWrites + NetMsgPtPWrites )
AS NetWrites,
( ProcBlocked
+ ProcReady )
AS ProcActive,
( ProcBlksDBLock + ProcBlksMemAlloc
+ ProcBlksMisc
+
ProcBlksMonitor + ProcBlksMonResume
+
ProcBlksNetThrottle
+
ProcBlksSegLock + ProcBlksFsgLock
+
ProcBlksFsgRead + ProcBlksFsgWrite ) (FORMAT '------9')
AS ProcBlocks,
( ProcWaitDBLock + ProcWaitMemAlloc
+ ProcWaitMisc
+
ProcWaitMonitor + ProcWaitMonResume
+
ProcWaitNetThrottle + ProcWaitPageRead +
ProcWaitSegLock + ProcWaitFsgLock
+
ProcWaitFsgRead + ProcWaitFsgWrite ) (FORMAT '------9')
AS ProcWaits,
( CmdDDLStmts
+ CmdDeleteStmts + CmdInsertStmts + CmdSelectStmts +
CmdUpdateStmts + CmdUtilityStmts + CmdOtherStmts )
AS UserStmtsArriving,
/* iota limit per device */
( UsedIota*100 / NULLIFZERO(FullPotentialIota) ) AS Used_FullPotential,
/* Spare Field Usage */
/* SPMA table fields */
FROM DBC.ResUsageSpma WITH CHECK OPTION;
Note: The ResUsageSpma table fields have been removed from this sample output.
174

ResSpsView
ResSpsView
ResSpsView is based on the ResUsageSps table.
REPLACE VIEW DBC.ResSpsView
AS SELECT
TheDate,
TheTime,
GmtTime,
NodeType,
TheTimestamp,
CentiSecs,
Secs,
NominalSecs,
CodFactor,
/* Capacity On Demand: 500 = 50 percent */
NCPUs,
AMPcount,
RowIndex1,
/* SLES10 = PGid, SLES11 = pWDid */
VprType,
PPid,
/* always zero for SLES11
*/
Reserved,
/* Aliased table fields */
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,/*
NumTasks
AS NumProcs, /*
RowIndex1
AS PGid,
/*
RowIndex1
AS pWDid,
/*
ActiveSessions AS NumSets,
/*
PM CPU Capacity On Demand factor */

renamed NumProcs to NumTasks in 14.0 */
SLES10 */
SLES11 */
renamed NumSets in 14.0 */

/* PM/WM CODs */
Spare00
AS WM_CPU_COD,
Spare01
AS WM_IO_COD,
/* WM IO Capacity On Demand factor */
/* SLES11 Hard Limits Capacity On Demand (COD) */
Spare03
AS CpuVpThrottleCount, /* VP-level CPU hard limits */
Spare04
AS CpuVpThrottleTime, /* VP-level milli-seconds */
Spare05
AS CpuThrottleCount,
/* WD level throttle count */
Spare06
AS CpuThrottleTime,
/* WD level milli-seconds */
Spare07
AS FullPotentialIota, /* node level total Full Pontential iotas
*/
Spare08
/* WD level device Pontential iotas (after COD) */
Spare09
AS UsedIota,
/* WD level device used iotas
*/
Spare10
AS IoThrottleCount,
/* WD level IoThrottleCount
*/
/* TIM stats */
Spare11
As VHLogicalDBRead,
/* WD level logical db reads from VH Cache
*/
Spare12
As VHLogicalDBReadKB, /* WD level KB of logical db reads from VH Cache */
Spare13
As VHPhysicalDBRead,
/* WD level physical db reads for VH data
*/
Spare14
As VHPhysicalDBReadKB, /* WD level KB of physical db reads for VH data */
/* PM/WM CODs */
SpareInt
AS PM_IO_COD,
*/
/* PM/WM CODs */
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
/* report average wait time per request for messages */
/* already delivered to AWTs
*/
175

ResSpsView
QWaitTime / NULLIFZERO(NumRequests)
AS QWaitTimeRequestAvg,
/* Avg delay for msgs delivered in period per Request

*/
WorkMsgSendDelay / NULLIFZERO(WorkMsgSendDelayCnt)
AS WorkMsgSendDelayRequestAvg,
/* avg delay per request on receive side for messages not */
/* yet delivered to AWTs
*/
WorkMsgReceiveDelay / NULLIFZERO(WorkMsgReceiveDelayCnt)
AS WorkMsgReceiveDelayRequestAvg,
/* report average service time per request */
ServiceTime / NULLIFZERO(AwtReleases)
AS ServiceTimeRequestAvg,
(CPUUServAWT + CPUUServDisp + CPUUServMisc)
(CPUUExecAWT + CPUUExecDisp + CPUUExecMisc)
(CPUUServ + CPUUExec)
(CpuTime * 10)/(CentiSecs * NCPUs)
AS
AS
AS
AS
CPUUServ,
CPUUExec,
CpuTime,
CpuPct,
/* All WorkTypes in use, averaged per AMP */

(WorkTypeInuse00 + WorkTypeInuse01 + WorkTypeInuse02 + WorkTypeInuse03 +
WorkTypeInuse04 + WorkTypeInuse05 + WorkTypeInuse06 + WorkTypeInuse07 +
WorkTypeInuse08 + WorkTypeInuse09 + WorkTypeInuse10 + WorkTypeInuse11 +
WorkTypeInuse12 + WorkTypeInuse13 + WorkTypeInuse14 + WorkTypeInuse15)
/ NULLIFZERO(AMPcount)
AS WorkTypeInuseAmp,
/* Max of (WorkTypesInuse for all AMPs (NOT per AMP)
*/
/* Can NOT divide by AMPs. MAX(SUM(AWT[AMP]))
*/
(WorkTypeMax00 + WorkTypeMax01 + WorkTypeMax02 + WorkTypeMax03 +
WorkTypeMax04 + WorkTypeMax05 + WorkTypeMax06 + WorkTypeMax07 +
WorkTypeMax08 + WorkTypeMax09 + WorkTypeMax10 + WorkTypeMax11 +
WorkTypeMax12 + WorkTypeMax13 + WorkTypeMax14 + WorkTypeMax15)
AS WorkTypeInuseMax,
(ProcBlksFsgRead + ProcBlksFsgWrite + ProcBlksFsgNIOs) AS IODelay,
(ProcWaitFsgRead + ProcWaitFsgWrite + ProcWaitFsgNIOs) AS IODelayTime,
(FilePDbAcqs + FilePDbPres)
AS LogicalReadPerm,
FilePDbDyRRels
AS LogicalWritePerm,
(FilePDbAcqReads + FilePDbPreReads)
AS PhysicalReadPerm,
FilePDbFWrites
AS PhysicalWritePerm,
FilePDbFWriteKB
AS PhysicalWritePermKB,
(NetPtPReads + NetBrdReads)
AS NetReads,
(NetPtPWrites + NetBrdWrites)
AS NetWrites,
(FilePCiAcqs + FileSDbAcqs + FileSCiAcqs + FileTJtAcqs + FileAPtAcqs +
FilePCiPres + FileSDbPres + FileSCiPres + FileTJtPres + FileAPtPres)
AS LogicalReadOther,
(FilePCiAcqReads + FileSDbAcqReads + FileSCiAcqReads + FileTJtAcqReads
FilePCiPreReads + FileSDbPreReads + FileSCiPreReads + FileTJtPreReads
AS PhysicalReadOther,
(FilePCiDyRRels + FileSDbDyRRels + FileSCiDyRRels + FileTJtDyRRels
AS LogicalWriteOther,
(FilePCiFWrites + FileSDbFWrites + FileSCiFWrites + FileTJtFWrites
AS PhysicalWriteOther,
(FilePCiFWriteKB + FileSDbFWriteKB + FileSCiFWriteKB + FileTJtFWriteKB
AS PhysicalWriteOtherKB,
( FilePDbAcqKB
( FileSDbAcqKB
FilePCiAcqKB
FilePCiPresKB
+
+
+
+
+ FileAPtAcqReads +
+ FileAPtPreReads)
+ FileAPtDyRRels)
+ FileAPtFWrites)
+ FileAPtFWriteKB)
FilePDbPresKB
FileSDbPresKB
FileSCiAcqKB
FileSCiPresKB
)
AS LogicalReadPermKB,
+
+ FileTJtAcqKB + FileAPtAcqKB +
+ FileTJtPresKB + FileAPtPresKB)
AS LogicalReadOtherKB,
( FilePDbPreReadKB + FilePDbAcqReadKB )
AS PhysicalReadPermKB,
( FilePCiAcqReadKB + FileSDbAcqReadKB + FileSCiAcqReadKB + FileTJtAcqReadKB + FileAPtAcqReadKB +
FilePCiPreReadKB + FileSDbPreReadKB + FileSCiPreReadKB + FileTJtPreReadKB + FileAPtPreReadKB)
AS PhysicalReadOtherKB,
FilePDbDyRRelKB
AS LogicalWritePermKB,
( FileSDbDyRRelKB +
FilePCiDyRRelKB + FileSCiDyRRelKB + FileTJtDyRRelKB + FileAPtDyRRelKB)
AS LogicalWriteOtherKB,
(ProcBlksSegNoVirtual + ProcBlksSegMDL + ProcBlksSegLock) AS ProcBlksSeg,
(ProcWaitSegNoVirtual + ProcWaitSegMDL + ProcWaitSegLock) AS ProcWaitSeg,
(ProcBlksMisc + ProcBlksNetThrottle + ProcBlksQnl +ProcBlksTime + ProcBlksFlowControl)
AS ProcBlksOther,
/* Donald.P: ProcBlksMonResume = covered by DBLocks */
(ProcWaitMisc + ProcWaitMonResume + ProcWaitNetThrottle +ProcWaitQnl +
ProcWaitTime + ProcWaitFlowControl)
AS ProcWaitOther,
/* Average number of AWTs used based on WorkTimeInuse
WorkTimeInuse/(Centisecs*10)
AS AwtUsedAvg,
*/
/* iota limit per WD */

( UsedIota*100 / NULLIFZERO(FullPotentialIota) ) AS Used_FullPotential_ByWd,
176

ResSpsView
/* iota limit by WD */
( UsedIota*100 / NULLIFZERO(CodPotentialIota ) ) AS Used_CodPotential_ByWd,
FROM DBC.ResUsageSps WITH CHECK OPTION;
Note: The ResUsageSps table fields have been removed from this sample output.
177

ResSvdskView
ResSvdskView
ResSvdskView is based on the ResUsageSvdsk table.
REPLACE VIEW DBC.ResSvdskView
AS SELECT
thedate,
NodeID
thetime,
GmtTime,
NodeType
TheTimestamp,
CentiSecs
Secs
NominalSecs
CodFactor
SummaryFlag
Reserved
VprId,
(FORMAT '999-99')
AS NodeID,
(FORMAT 'X(8)')
AS NodeType,
(FORMAT
(FORMAT
(FORMAT
(FORMAT
(FORMAT
(FORMAT
AS
AS
AS
AS
AS
AS
'-------9')
'----9')
'ZZZ9')
'-----9')
'X(1)')
'X(1)')
CentiSecs,
Secs,
NominalSecs,
CodFactor,
SummaryFlag,
Reserved,
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,

/* PM/WM CODs */
Spare00
AS WM_CPU_COD,
Spare01
AS WM_IO_COD,
SpareInt
AS PM_IO_COD,

/* WM IO Capacity On Demand factor */
/* PM IO Capacity On Demand factor */
/* PM/WM CODs */
/*
*
*
*

resulting groups. If a coexistence system had 5550, 5600 and 5650 n
odes
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
FROM DBC.ResUsageSvdsk WITH CHECK OPTION;
Note: The ResUsageSvdsk table fields have been removed from this sample output.
178

ResSvprView
ResSvprView
This view allows data to be properly presented and reports all the columns available from the
ResUsageSvpr table.
Note: The data columns in this view will change as the columns in the ResUsageSvpr table
change.
REPLACE VIEW DBC.ResSvprView
AS SELECT
thedate,
NodeID,
thetime,
GmtTime,
NodeType,
TheTimestamp,
CentiSecs,
Secs,
NominalSecs,
CodFactor,
SummaryFlag,
Reserved,
NCPUs,
vprid,
VprType,
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,
/* Spare field usage */

/* TIM stats */
Spare00
As VHAgedOut,
/* data blocks aged out of VH Cache */
Spare01
As VHAgedOutKB, /* KB aged out of VH Cache
*/
Spare02
As VHLogicalDBRead,
/* logical db reads from VH Cache
*/
Spare03
As VHLogicalDBReadKB, /* KB of logical db reads from VH Cache */
Spare04
As VHPhysicalDBRead,
/* physical db reads for VH data
*/
Spare05
As VHPhysicalDBReadKB, /* KB of physical db reads for VH data */
/* PM/WM CODs */
Spare06
AS WM_CPU_COD, /* WM CPU Capacity On Demand factor */
/* DTIM inuse cache size */
Spare07
AS VHCacheInuseKB, /* current inuse VH cache KB */
/*PM/WM CODs */
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
/* SVpr table fields (remaining) */
FROM DBC.ResUsageSvpr WITH CHECK OPTION;
Note: The ResUsageSvpr table fields have been removed from this sample output.
179

ResSvprView
180
CHAPTER 15
This chapter describes the output format of the resource usage macros and each macro.
Macro Output Format

Resource usage macros provide output in the following general format.
<Report Date>
<Title of Report>
--------------1st
2nd
1st
Date
Time
Type
Id
Id
Stat
-------- -------- ---- ------ ------ ------99/99/99 99:99:99 AAAA 999-99 999-99 999.99%
999.99%
...........
AAAA 999-99 999-99 999.99%
999-99 999-99 999.99%
99:99:99 AAAA 999-99 999-99 999.99%
...........
Page <num>
2nd
3rd
Stat
Stat ...
-------- -------99999.99 99999.99
99999.99 99999.99
99999.99 99999.99
99999.99 99999.99
99999.99 99999.99
where:
Column
Description
Date
The date at the end of a log interval.
Time
The time at the end of a log interval. Statistics on each line cover the time
period ending at the indicated time.
Type
A virtual processor type, logical device type, host type, or a special type
for certain reports.
1st ID, 2nd ID, etc.
The appropriate identifier, which varies, depending on the macro. It is

one or more of the following:
1st Stat, 2nd Stat, etc.
NodeID
VprocID
HostID
GroupID
The appropriate statistics. Details are given with the descriptions of each
macro in this chapter.
Numbers are generally displayed with the appropriate fixed format (for
example, 'zzzz9.99') unless the number represents a percentage or sum of
percentages.
Percentages are displayed with the appropriate format (for example,
'zz9.9%', 'zz9' or 'zz9.99').
Unless otherwise specified, all statistical numbers are expressed as either:
181
Chapter 15: Resource Usage Macros

Macro Output Format
Percentage
Millisecond (ms)
Kilobyte (KB)
Columns whose values depend on the logging rate are never reported as raw data. Instead,
they are converted to a normalized value, such as per second.
All reports are ordered by date, time, type, 1st ID, 2nd ID, and so forth. Repeated date, time,
type, and ID column values are suppressed until a new row presents a different value.
Question Marks
Question marks used as values in the output examples are generated when a division by zero is
made. It represents data that is not available. The numbers in the columns are calculated, for
example, by dividing KB by number of blocks read. When there are no blocks read, KB is
divided by zero. A question mark does not mean there is an error, but indicates that there is no
information to report for this time period.
Usage Notes
To get current data, logging must be enabled on the resource usage table used by the view or
macro.
182

ResAWT Macros
ResAWT Macros
Function
Macro...
Reports the average AMP Worker Task...
ResAWT
in use for all AMPs in the system.
ResAWTByAMP
in use for each AMP.
ResAWTByNode
on all AMPs in each node.
Input Format Examples

The input forms of these three macros are described below.
EXEC ResAWT
(FromDate,ToDate,FromTime,ToTime);
EXEC ResAWTByAMP
(FromDate,ToDate,FromTime,ToTime
EXEC ResAWTByNode
(FromDate,ToDate,FromTime,ToTime,FromNode,ToNode);
See Macro Execution on page 31 for a description of the FromDate, ToDate, FromTime,
ToTime, FromNode, ToNode and Node parameters.
Usage Notes
You must have logging enabled on the ResUsageSawt table.
Output Examples
The reports in the following sections are sample output reports from the ResAWT,
ResAWTByAMP, and ResAWTByNode macros.
In the ResAWT output report, the statistics columns, after the Date and Time columns,
provide a summary of the AMP Worker Tasks resource usage.
The following table describes the statistics columns, after the Date and Time columns, in the
ResAWTByAMP output.
Statistics columns
Description
Node ID.
AMP ID.
3 through 24
Summary of AMP Worker Tasks resource usage.
183

ResAWT Macros
ResAWTByNode output report.
Statistics columns
Description
Node ID.
2 through 23
Summary of AMP Worker Tasks resource usage.
The following table describes the columns in all output reports (with the exception of
ResAWTByNode, which has the NodeID column, and ResAWTByAMP, which has the Node
ID and AMP ID columns).
184
Column...
Reports the...
Mail box Depth
current depth of the AMP work mailbox.
In Flow Ctl
AMP that is or is not in flow control.
Flow Ctls Per Sec
number of times during the log period that the system entered the
flow control state from a non-flow controlled state.
Work New AWTs
current number of AMP Worker Tasks in use during the log

period for each new first-level secondary work type for the VprId
vproc.
Work One AWTs

period for each first-level secondary work type for the VprId
vproc.
New + One AWTs
summary of the previous two columns, Work New AWTs and

Work One AWTs.
Work Two AWTs

period for each second-level secondary work type for the VprId
vproc.
Work 3 AWTs

period for each third-level secondary work type for the VprId
vproc.
Work Abrt AWTs

period for each transaction abort request for the VprId vproc.
Work Spwn AWTs

period for each spawned abort request for the VprId vproc.
Work Norm AWTs

period for each message that does not fall within the standard
work type hierarchy for the VprId vproc.
Work Ctl AWTs

ResAWT Macros
Column...
Reports the...
Work Exp AWTs

period for each expedited Allocation Groups for the VprId vproc.
(Expedited Allocation Groups exist when using the reserved AMP
Worker Task pool. See the Priority Scheduler (schmon) chapter
in Utilities for details.)
Max Work New AWTs
the maximum number of AMP Worker Tasks in use at one time

during the log period for each new work type for the VprId vproc.
Max Work One AWTs

during the log period for each first-level secondary work type for
the VprId vproc.
Max Work Two AWTs

during the log period for each second-level secondary work type
for the VprId vproc.
Max Work 3 AWTs

during the log period for each third-level secondary work type for
the VprId vproc.
Max Work Abrt AWTs

during the log period for each transaction abort request for the
VprId vproc.
Max Work Spwn AWTs

during the log period for each spawned abort request for the
VprId vproc.
Max Work Norm AWTs

during the log period for each message that does not fall within
the standard work type hierarchy for the VprId vproc.
Max Work Ctl AWTs
Max Work Exp AWTs

during the log period for each expedited Allocation Groups for the
VprId vproc. (Expedited Allocation Groups exist when using the
reserved AMP Worker Task pool. See the Priority Scheduler
(schmon) chapter in Utilities for details.)
For a complete description of the columns above, see Chapter 7: ResUsageSawt Table.
185

ResAWT Macros
ResAWT Sample Output

07/08/17
Page
1
Date
Time
-------- -------07/08/12 23:00:00
23:01:00
23:02:00
23:03:00
23:04:00
23:05:00
23:06:00
23:07:00
23:08:00
23:09:00
23:10:00
23:11:00
23:12:00
23:13:00
23:14:00
23:15:00
23:16:00
23:17:00
23:18:00
23:19:00
23:20:00
23:21:00
23:22:00
AMP Worker Task Summary Average Usage per AMP Across System
Mail
In
Flow Work Work New+ Work Work
Box Flow
Ctls New One One Two
3
Depth Ctl? PerSec AWTs AWTs AWTs AWTs AWTs
------ ----- ------ ---- ---- ---- ---- ---8 1.00
0.01
31
22
54
0
0
9 0.00
0.02
28
25
54
0
0
6 0.00
0.00
31
22
54
0
0
3 0.00
0.00
31
22
53
0
0
2 0.00
0.00
31
22
52
0
0
5 0.00
0.00
31
22
53
0
0
3 0.00
0.00
27
24
51
0
0
1 0.00
0.00
21
19
40
0
0
1 0.00
0.00
26
20
46
0
0
2 0.00
0.00
30
20
50
0
0
1 0.00
0.00
29
16
46
0
0
2 0.00
0.00
30
19
49
0
0
2 0.00
0.00
31
21
52
0
0
1 0.00
0.00
29
19
49
0
0
1 0.00
0.00
29
18
47
0
0
2 0.00
0.00
29
19
48
0
0
3 0.00
0.00
34
18
52
0
0
6 0.00
0.00
35
19
54
0
0
8 0.00
0.00
30
23
53
0
0
1 0.00
0.01
25
24
49
0
0
1 0.00
0.00
28
18
46
0
0
7 0.00
0.01
34
20
54
0
0
2 0.00
0.01
30
22
53
0
0
Work
Abrt
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Spwn
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Norm
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Ctl
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Exp
AWTs
---0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Max Max Max Max Max Max Max Max Max

Work Work Work Work Work Work Work Work Work
New One Two
3 Abrt Spwn Norm Ctl Exp
AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs
---- ---- ---- ---- ---- ---- ---- ---- --35
25
2
1
1
0
0
0
3
33
27
1
1
1
0
0
0
2
35
26
1
1
1
0
0
0
2
35
26
1
1
1
0
0
0
3
36
27
1
1
1
0
0
0
3
36
26
1
1
1
0
0
0
4
34
27
1
1
1
0
0
0
3
35
29
1
1
1
0
0
0
3
35
29
1
1
1
0
0
0
3
37
26
1
1
1
0
0
0
4
38
25
2
1
1
0
0
0
3
38
27
1
1
1
0
0
0
3
37
26
1
1
1
0
0
0
3
36
26
1
1
1
0
0
0
4
36
25
1
1
1
0
0
0
3
37
25
2
1
1
0
0
0
3
37
25
1
1
1
0
0
0
4
40
22
1
1
1
0
0
0
4
37
24
1
1
1
0
0
0
4
36
30
1
1
1
0
0
0
3
36
30
2
2
1
0
0
0
4
36
26
1
1
1
0
0
0
3
36
27
1
1
1
0
0
0
3
ResAWTByAMP Sample Output

07/08/17
Page
1
Mail In
AMP Box
Flow
Date
Time
ID
Depth Ctl?
-------- -------- ------ ----- ----07/08/12 23:00:00
0
3
0.00
1
1
0.00
2
12 0.00
3
14 0.00
4
11 0.00
5
12 0.00
6
17 0.00
7
5
0.00
8
4
0.00
9
1
0.00
10
0.00
15
7
0.00
12
1
0.00
13
1
0.00
14
4
0.00
15
27 0.00
186
AMP Worker Task Summary Usage per AMP
Flow
Ctls
PerSec
-----0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Work
New
AWTs
---31
32
31
32
32
30
32
32
32
32
32
32
30
32
32
32
Work
One
AWTs
---22
22
22
22
22
24
22
22
22
22
22
22
23
22
23
22
New+
One
AWTs
---53
54
54
54
54
53
54
54
54
54
54
54
54
54
52
54
Work
Two
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
3
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Abrt
AWTs
---1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Spwn
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Norm
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Ctl
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Work
Exp
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Max
Work
New
AWTs
---33
34
34
34
34
35
34
34
34
34
35
34
34
34
33
34
Max
Work
One
AWTs
---24
24
24
24
25
23
24
23
23
24
24
24
24
24
25
24
Max
Max Max Max Max Max Max
Work Work Work Work Work Work Work
Two
3
Abrt Spwn Norm Ctl Exp
AWTs AWTs AWTs AWTs AWTs AWTs AWTs
---- ---- ---- ---- ---- ---- ---2
1
1
0
0
0
3
1
0
0
0
0
0
2
1
0
0
0
0
0
2
1
0
0
0
0
0
3
1
0
0
0
0
0
3
1
0
0
0
0
0
3
1
0
0
0
0
0
4
1
0
0
0
0
0
3
1
0
0
0
0
0
3
1
0
0
0
0
0
3
1
0
0
0
0
0
4
1
0
0
0
0
0
3
1
0
0
0
0
0
3
1
0
0
0
0
0
3
1
0
0
0
0
0
3
1
0
0
0
0
0
3

ResAWT Macros
ResAWTByNode Sample Output

07/08/17
AMP Worker Task Summary Average Usage per AMP By Node
Max Max Max Max

Mail
In
Flow Work Work New+ Work Work Work Work Work Work Work Work Work Work Work
Node
Box Flow
Ctls New One One Two
3 Abrt Spwn Norm Ctl Exp New One Two
3
Date
Time
ID Depth Ctl? PerSec AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs AWTs
-------- -------- ------ ------ ----- ------ ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---07/08/12 23:00:00 1-04
8 1.00
0.01
31
22
54
0
0
0
0
0
0
0
35
25
2
1
23:01:00 1-04
9 0.00
0.02
28
25
54
0
0
0
0
0
0
0
33
27
1
1
23:02:00 1-04
6 0.00
0.00
31
22
54
0
0
0
0
0
0
0
35
26
1
1
23:03:00 1-04
3 0.00
0.00
31
22
53
0
0
0
0
0
0
0
35
26
1
1
23:04:00 1-04
2 0.00
0.00
31
22
52
0
0
0
0
0
0
0
36
27
1
1
23:05:00 1-04
5 0.00
0.00
31
22
53
0
0
0
0
0
0
0
36
26
1
1
23:06:00 1-04
3 0.00
0.00
27
24
51
0
0
0
0
0
0
0
34
27
1
1
23:07:00 1-04
1 0.00
0.00
21
19
40
0
0
0
0
0
0
1
35
29
1
1
23:08:00 1-04
1 0.00
0.00
26
20
46
0
0
0
0
0
0
0
35
29
1
1
23:09:00 1-04
2 0.00
0.00
30
20
50
0
0
0
0
0
0
0
37
26
1
1
23:10:00 1-04
1 0.00
0.00
29
16
46
0
0
0
0
0
0
0
38
25
2
1
23:11:00 1-04
2 0.00
0.00
30
19
49
0
0
0
0
0
0
0
38
27
1
1
23:12:00 1-04
2 0.00
0.00
31
21
52
0
0
0
0
0
0
0
37
26
1
1
23:13:00 1-04
1 0.00
0.00
29
19
49
0
0
0
0
0
0
0
36
26
1
1
23:14:00 1-04
1 0.00
0.00
29
18
47
0
0
0
0
0
0
0
36
25
1
1
23:15:00 1-04
2 0.00
0.00
29
19
48
0
0
0
0
0
0
0
37
25
2
1
23:16:00 1-04
3 0.00
0.00
34
18
52
0
0
0
0
0
0
0
37
25
1
1
23:17:00 1-04
6 0.00
0.00
35
19
54
0
0
0
0
0
0
0
40
22
1
1
23:18:00 1-04
8 0.00
0.00
30
23
53
0
0
0
0
0
0
0
37
24
1
1
23:19:00 1-04
1 0.00
0.01
25
24
49
0
0
0
0
0
0
0
36
30
1
1
23:20:00 1-04
1 0.00
0.00
28
18
46
0
0
0
0
0
0
0
36
30
2
2
Page
Max
Work
Abrt
AWTs
---1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Max
Work
Spwn
AWTs
---0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
Max Max Max
Work Work Work
Norm Ctl Exp
AWTs AWTs AWTs
---- ---- ---0
0
3
0
0
2
0
0
2
0
0
3
0
0
3
0
0
4
0
0
3
0
0
3
0
0
3
0
0
4
0
0
3
0
0
3
0
0
3
0
0
4
0
0
3
0
0
3
0
0
4
0
0
4
0
0
4
0
0
3
0
0
4
187

ResCPUByAMP Macros
ResCPUByAMP Macros
Function
Macro...
Reports the following...
ResCPUByAMP
how each AMP on each node utilizes the CPUs.
ResCPUByAMPOneNode
how each AMP on a specific node utilizes the CPUs.
ResAmpCpuByGroup
the summary of AMP CPU usage by node grouping.

EXECUTE ResCPUByAMP
EXECUTE ResCPUByAMPOneNode
(FromDate,ToDate,FromTime,ToTime,Node);
EXECUTE ResCAMPCpuByGroup
Usage Notes
You must have logging enabled on the ResUsageSvpr table.
Output Examples
The reports in the following sections are sample output reports from the ResCPUByAMP, the
ResCPUByAMPOneNode, and the ResAmpCpuByGroup macros, respectively, where:
Column...
Reports the percent of time AMPs were busy doing user...
Awt User Serv%
service for the AMP Worker Task (Awt) partition.
Misc User Serv%
service for miscellaneous (all other except Partition 0) AMP partitions.
Awt User Exec%
execution within the AMP Worker Task (Awt) partition.
Misc User Exec%
execution within miscellaneous (all other except Partition 0) AMP

partitions.
Total User Serv%
service work. This is the sum of the Awt User Serv%, the Misc User
Serv%, and AMP Partition 0 user service%.
Note: User service is the time that a CPU is busy executing kernel system
calls or servicing I/O and timer hardware interrupts.
188

ResCPUByAMP Macros
Column...
Reports the percent of time AMPs were busy doing user...
Total User Exec%
execution work. This is the sum of the Awt User Exec%, Misc User
Exec%, and AMP Partition 0 user execution.
Note: User execution is the time a CPU is busy executing user execution
code, which is the time spent in a user state on behalf of a process.
Total Busy%
service and execution work. This is the sum of the Total User Serv% and
the Total User Exec% columns.
Note: The above CPU statistics represent the aggregate of all time spent in the indicated way
by all CPUs on the node. Because there are multiple CPUs, the Total Busy % should be
compared to a theoretical maximum of 100% times the number of CPUs on the node.
The Node CPU column in the following sample outputs reports the number of CPUs
(NCPUs).
For more information on how to monitor busy AMP Worker Tasks, see AWT Monitor
(awtmon) in Utilities.
ResCPUByAMP Sample Output

Note: The NodeID column only appears in the ResCPUByAMP output report.
01/07/12
CPU USAGE BY AMP
Date
-------01/07/12
Page
Vproc
Id
----0
1
Node
Id
-----001-01
001-01
Node
CPUs
-------4
4
Awt
User
Serv%
------0.36%
0.26%
Misc
User
Serv%
------0.00%
0.00%
Awt
User
Exec%
------0.05%
0.12%
Misc
User
Exec%
------0.00%
0.00%
Total
User
Serv%
------0.36%
0.30%
Total
User
Exec%
------0.05%
0.12%
Total
Busy
%
------0.41%
0.42%
09:57:20
0
1
001-01
001-01
4
4
0.41%
0.34%
0.00%
0.00%
0.12%
0.05%
0.00%
0.00%
0.45%
0.38%
0.12%
0.05%
0.58%
0.42%
09:57:40
0
1
001-01
001-01
4
4
0.25%
0.19%
0.00%
0.00%
0.18%
0.06%
0.00%
0.00%
0.28%
0.29%
0.18%
0.06%
0.45%
0.35%
09:58:00
0
1
001-01
001-01
4
4
0.38%
0.31%
0.00%
0.00%
0.08%
0.09%
0.00%
0.00%
0.45%
0.34%
0.08%
0.09%
0.52%
0.42%
09:58:20
0
1
001-01
001-01
4
4
0.31%
0.36%
0.00%
0.00%
0.08%
0.09%
0.00%
0.00%
0.34%
0.40%
0.08%
0.09%
0.41%
0.49%
09:58:40
0
1
001-01
001-01
4
4
0.39%
0.32%
0.00%
0.00%
0.11%
0.12%
0.00%
0.00%
0.41%
0.36%
0.11%
0.12%
0.52%
0.49%
09:59:00
0
1
001-01
001-01
4
4
0.29%
0.21%
0.00%
0.00%
0.11%
0.09%
0.00%
0.00%
0.30%
0.22%
0.11%
0.09%
0.41%
0.31%
09:59:20
0
1
001-01
001-01
4
4
0.30%
0.30%
0.00%
0.00%
0.06%
0.19%
0.00%
0.00%
0.31%
0.32%
0.06%
0.19%
0.38%
0.51%
09:59:40
0
1
001-01
001-01
4
4
0.40%
0.26%
0.00%
0.00%
0.09%
0.08%
0.00%
0.00%
0.46%
0.38%
0.09%
0.08%
0.55%
0.45%
10:00:00
0
1
001-01
001-01
4
4
0.32%
0.28%
0.00%
0.00%
0.08%
0.09%
0.00%
0.00%
0.34%
0.31%
0.08%
0.09%
0.41%
0.40%
Time
-------09:57:00
189

ResCPUByAMP Macros
ResCPUByAMPOneNode Sample Output

01/07/12
CPU Usage by AMP for Node 001-01 (4 CPUs)
Date
-------01/07/12
Page
68
Vproc
Id
----0
1
NCPUs
----4
4
Awt
User
Serv%
------0.36%
0.26%
09:57:20
0
1
4
4
0.41%
0.34%
0.00%
0.00%
0.12%
0.05%
0.00%
0.00%
0.45%
0.38%
0.12%
0.05%
0.58%
0.42%
09:57:40
0
1
4
4
0.25%
0.19%
0.00%
0.00%
0.18%
0.06%
0.00%
0.00%
0.28%
0.29%
0.18%
0.06%
0.45%
0.35%
09:58:00
0
1
4
4
0.38%
0.31%
0.00%
0.00%
0.08%
0.09%
0.00%
0.00%
0.45%
0.34%
0.08%
0.09%
0.52%
0.42%
09:58:20
0
1
4
4
0.31%
0.36%
0.00%
0.00%
0.08%
0.09%
0.00%
0.00%
0.34%
0.40%
0.08%
0.09%
0.41%
0.49%
09:58:40
0
1
4
4
0.39%
0.32%
0.00%
0.00%
0.11%
0.12%
0.00%
0.00%
0.41%
0.36%
0.11%
0.12%
0.52%
0.49%
09:59:00
0
1
4
4
0.29%
0.21%
0.00%
0.00%
0.11%
0.09%
0.00%
0.00%
0.30%
0.22%
0.11%
0.09%
0.41%
0.31%
Time
-------09:57:00
Misc
User
Serv%
------0.00%
0.00%
Awt
User
Exec%
------0.05%
0.12%
Misc
User
Exec%
------0.00%
0.00%
Total
User
Serv%
------0.36%
0.30%
Total
User
Exec%
------0.05%
0.12%
Total
Busy
%
------0.41%
0.42%
ResAmpCpuByGroup Sample Output

Note: The GroupID column only appears in the ResAmpCpuByGroup output report.
01/07/12
45
AMP CPU USAGE BY GROUP
Date
-------01/07/12
190
Page
Node
CPUs
-------4
Awt
User
Serv%
------0.32%
Misc
User
Serv%
------0.00%
Awt
User
Exec%
------0.07%
Misc
User
Exec%
------0.00%
Total
User
Serv%
------0.36%
Total
User
Exec%
------0.07%
Total
Busy
%
------0.43%
0.33%
0.00%
0.08%
0.00%
0.36%
0.08%
0.44%
09:52:20
0.35%
0.00%
0.07%
0.00%
0.37%
0.07%
0.44%
09:52:40
0.36%
0.00%
0.09%
0.00%
0.39%
0.09%
0.48%
09:53:00
0.27%
0.00%
0.09%
0.00%
0.28%
0.09%
0.37%
09:53:20
0.29%
0.00%
0.06%
0.00%
0.34%
0.06%
0.40%
09:53:40
0.36%
0.00%
0.06%
0.00%
0.40%
0.06%
0.46%
09:54:00
0.35%
0.00%
0.11%
0.00%
0.38%
0.11%
0.49%
09:54:20
0.34%
0.00%
0.07%
0.00%
0.36%
0.07%
0.43%
09:54:40
0.41%
0.00%
0.04%
0.00%
0.43%
0.04%
0.47%
09:55:00
0.28%
0.00%
0.09%
0.00%
0.28%
0.09%
0.37%
09:55:20
0.35%
0.00%
0.09%
0.00%
0.43%
0.09%
0.53%
09:55:40
0.34%
0.00%
0.06%
0.00%
0.42%
0.06%
0.48%
09:56:00
0.26%
0.00%
0.08%
0.00%
0.29%
0.08%
0.37%
Time
-------09:51:40
Group
Id
----A
09:52:00

ResCPUByAMP Macros
Normalized Viewing of CPU Usage by AMP

Some users may prefer to view CPU usage by AMP in a normalized fashion. Conceptually, this
restates each of the above statistics in terms of percentage of total CPU capacity of the node.
The following SQL example shows how to perform this normalization for the Total Busy %
statistic.
SEL TheDate, TheTime, Vproc, NodeID,
(AmpTotalUserExec+AmpTotalUserServ)
/Secs/NCPUs
(FORMAT zz9%,TITLE Total// Busy// %)
FROM ResCpuUsageByAMPView
WHERE TheDate = CURRENT_DATE AND TheTime>080000
ORDER BY 1,2,3;
191

ResCPUByPE Macros
ResCPUByPE Macros
Function
Macro...
Reports...
ResCPUByPE
how each PE on each node is utilizing the CPUs.
ResCPUByPEOneNode
how each PE on a specific node is utilizing the CPUs.
ResPeCpuByGroup
the PE CPU utilization summarized by a node grouping.

The input forms of the these three macros are described below.
EXEC ResCPUByPE
EXEC ResCPUByPEOneNode
EXEC ResPeCpuByGroup
Usage Notes
You must have logging enabled on the ResUsageSvpr table.
Output Examples
The reports in the following sections are sample output reports from the ResCPUByPE,
ResCPUByPEOneNode, and ResPeCPUByGroup macros, respectively, where:
192
Column...
Reports the percent of time PEs are busy doing user...
Disp User Serv%
service for the Dispatcher partition of the PE.
Ses User Serv%
service for the Session Control partition of the PE.
Misc User Serv%
service for miscellaneous (all other, except Partition 0) PE partitions.
Disp User Exec%
execution within the Dispatcher partition of the PE.
Ses User Exec%
execution within the Session Control partition of the PE.
Misc User Exec%
execution within miscellaneous (all other, except Partition 0) PE partitions.
Total User Serv%
service work. This is the sum of the four user service columns above plus PE
Partition 0 user service.

ResCPUByPE Macros
Column...
Reports the percent of time PEs are busy doing user...
Total User Exec%
execution work. This is the sum of the four user execution columns above
plus PE Partition 0 user execution.
Total Busy%
service and execution work. This is the sum of the Total User Serv% and the
Total User Exec% columns.
Note: The above CPU statistics represent the aggregate of all time spent in the indicated way
by all CPUs on the node. Because there are multiple CPUs, the Total Busy % should be
compared to a theoretical maximum of 100% times the number of CPUs on the node.
The Node CPU column in the following sample outputs reports the number of CPUs
(NCPUs).
ResCPUByPE Sample Output

Note: The NodeID column only appears in the ResCPUByPE output report.
01/07/12
Page 1
CPU USAGE BY PE
Vproc
Node Node
Date
Time
Id
Id CPUs
-------- -------- ----- ------ ---01/07/12 09:57:00 16382 001-01
4
16383 001-01
4
Disp
Ses
Misc
User
User
User
Serv%
Serv%
Serv%
------- ------- ------0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
Disp
User
Exec%
------0.00%
0.00%
Ses
User
Exec%
------0.00%
0.00%
Misc
User
Exec%
------0.00%
0.00%
Total
Total
Total
User
User
Busy
Serv%
Exec%
%
------ ------ ------0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
09:57:20 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
09:57:40 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
09:58:00 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
09:58:20 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
09:58:40 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
09:59:00 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
09:59:20 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
09:59:40 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
10:00:00 16382 001-01

16383 001-01
4
4
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
ResCPUByPEOneNode Sample Output

01/09/13
Page 4
CPU Usage by PE for Node 001-01 (4 CPUs)
Vproc
Date
Time
Id
------------ ----01/08/21 15:41:00 2
Node
CPUs
---4
Disp
User
Serv%
-----0.05%
Ses
User
Serv%
-----0.00%
Misc
User
Serv%
-----0.00%
Disp
User
Exec%
-----0.01%
Ses
User
Exec%
-----0.00%
Misc
User
Exec%
-----0.00%
Total
User
Serv%
-----0.08%
Total
User
Exec%
-----0.58%
Total
Busy
%
----0.7%
15:42:00 2
0.01%
0.00%
0.00%
0.00%
0.00%
0.00%
0.02%
0.18%
0.2%
15:43:00 2
0.02%
0.00%
0.00%
0.00%
0.00%
0.00%
0.04%
0.20%
0.2%
15:44:00 2
0.02%
0.00%
0.00%
0.00%
0.00%
0.00%
0.05%
0.56%
0.6%
193

ResCPUByPE Macros
01/08/27
15:45:00 2
0.01%
0.00%
0.00%
0.00%
0.00%
0.00%
0.02%
0.18%
0.2%
16:21:00 2
0.08%
0.00%
0.00%
0.00%
0.00%
0.00%
0.13%
0.70%
0.8%
ResPeCpuByGroup Sample Output

Note: The GroupID column only appears in the ResPeCpuByGroup output report.
01/07/12
Page 8
PE CPU USAGE BY GROUP
Group Node
Date
Time
Id CPUs
----------- ----- ---01/07/12 04:55:40 A
4
Disp
User
Serv%
-----0.00%
Ses
User
Serv%
------0.00%
Misc
User
Serv%
-----0.00%
Disp
User
Exec%
-----0.00
0.00
Ses
User
Exec%
-----0.00%
Misc
Total
User
User
Exec%
Serv%
------- -----0.00%
0.00%
0.00%
0.00%
0.00%
Total
User
Exec%
----0.00%
Total
Busy
%
----0.00%
0.00%
0.00%
04:56:00
0.00%
0.00%
0.00%
04:56:20
0.00%
0.00%
0.00%
0.00
0.00%
0.00%
0.00%
0.00%
0.00%
04:56:40
0.00%
0.00%
0.00%
0.00
0.00%
0.00%
0.00%
0.00%
0.00%
04:57:00
0.00%
0.00%
0.00%
0.00
0.00%
0.00%
0.00%
0.00%
0.00%
04:57:20
0.00%
0.00%
0.00%
0.00
0.00%
0.00%
0.00%
0.00%
0.00%
04:57:40
0.00%
0.00%
0.00%
0.00
0.00%
0.00%
0.00%
0.00%
0.00%
04:58:00
0.00%
0.00%
0.00%
0.00
0.00%
0.00%
0.00%
0.00%
0.00%
04:58:20
0.00%
0.00%
0.00%
0.00
0.00%
0.00%
0.00%
0.00%
0.00%
Normalized Viewing of CPU Usage by PE

Some users may prefer to view CPU usage by PEs in a normalized fashion. Conceptually, this
restates each of the above statistics in terms of percentage of total CPU capacity of the node.
The following SQL example shows how to perform this normalization for the Total Busy %
statistic.
SEL TheDate, TheTime,Vproc,NodeID,
(PETotalUserExec+PETotalUserServ)
/Secs/NCPUs
(FORMAT zz9%,TITLE Total// Busy// %)
FROM ResCpuUsageByPEView
WHERE TheDate = CURRENT_DATE AND TheTime>080000
ORDER BY 1,2,3;
194

ResCPUByNode Macros
ResCPUByNode Macros
Function
Macro...
Reports how...
ResCPUByNode
each individual node is utilizing its CPUs.
ResCPUOneNode
a specific node is utilizing its CPUs.
ResCPUByGroup
a specified Node Group is utilizing the system CPUs.

EXEC ResCPUByNode
EXEC ResCPUOneNode
EXEC ResCPUByGroup
Usage Notes
You must have logging enabled on the ResUsageSpma table.
Output Examples
The reports in the following sections are sample output reports from the ResCPUByNode, the
ResCPUOneNode, and the ResCPUByGroup macros.
The following columns are the averages for all CPUs on the node.
This column ...
Lists percentage of time spent ...
I/O Wait %
idle and waiting for I/O completion.
Total User Serv %
busy doing user service work.
Total User Exec %
busy doing user execution work.
Total Busy %
busy doing user service and execution work.

This is the sum of Total User Serv % and the Total User Exec %
columns.
195

ResCPUByNode Macros
where:
This variable
Describes the time a CPU is busy executing
User service
kernel system calls or servicing I/O and timer hardware interrupts.
User execution
user execution code, which is the time spent in a user state on behalf of a
process.
ResCPUByNode Sample Output

Note: The NodeID column only appears in the ResCPUByNode output report.
01/07/12
CPU USAGE BY NODE Page
Date
-------01/07/12
Time
-------09:51:40
09:52:00
09:52:20
09:52:40
09:53:00
09:53:20
Node
Id
-----001-01
001-01
001-01
001-01
001-01
001-01
I/O
Wait
%
-----16.2%
17.2%
15.5%
16.1%
15.8%
15.5%
Total
User
Serv%
-----1.4%
1.3%
1.6%
1.5%
1.0%
1.5%
45
Total
User
Exec%
-----0.1%
0.2%
0.2%
0.2%
0.2%
0.2%
Total
Busy
%
-----1.5%
1.5%
1.8%
1.7%
1.2%
1.7%
ResCPUOneNode Sample Output

01/07/12
CPU Usage for Node 001-01
Date
-------01/07/12
196
Time
-------09:44:20
09:44:40
09:45:00
09:45:20
09:45:40
09:46:00
09:46:20
I/O
Wait
%
-----16.2%
16.9%
16.5%
17.0%
17.4%
16.6%
16.2%
Page
01
Total
Total
Total
User
User
Busy
Serv%
Exec%
%
------ ------ -----1.6%
0.2%
1.9%
1.3%
0.2%
1.5%
1.1%
0.1%
1.2%
1.7%
0.2%
1.9%
1.1%
0.2%
1.3%
1.3%
0.2%
1.5%
1.6%
0.2%
1.8%

ResCPUByNode Macros
ResCPUByGroup Sample Output

Note: The GroupID column only appears in the ResCpuByGroup output report.
00/10/16
CPU USAGE BY Group
Date
-------00/10/16
Time
-------11:25:00
Group
Id
----A
B
Page 2
I/O
Wait
%
-----0.0%
0.0%
Total
User
Serv%
-----0.0%
0.0%
Total
User
Exec%
-----0.0%
0.0%
Total
Busy
%
-----0.0%
0.0%
11:30:00
A
B
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
11:35:00
A
B
0.0%
0.0%
0.6%
0.3%
0.6%
0.4%
1.1%
0.7%
11:40:00
A
B
0.0%
0.0%
1.3%
1.1%
0.9%
0.9%
2.2%
2.0%
11:45:00
A
B
0.0%
0.0%
0.6%
0.3%
0.9%
1.0%
1.5%
1.3%
11:50:00
A
B
0.0%
0.0%
0.6%
0.6%
0.6%
0.8%
1.2%
1.3%
11:55:00
A
B
0.0%
0.0%
1.5%
1.6%
1.1%
1.0%
2.6%
2.6%
12:00:00
A
B
0.0%
0.0%
0.5%
0.7%
0.8%
0.9%
1.3%
1.6%
12:05:00
A
B
0.0%
0.0%
1.2%
0.6%
0.7%
0.5%
1.8%
1.1%
12:10:00
A
B
0.0%
0.0%
0.6%
1.1%
0.9%
1.2%
1.6%
2.2%
12:15:00
A
B
0.0%
0.0%
0.6%
0.5%
0.8%
0.7%
1.4%
1.2%
12:20:00
A
B
0.0%
0.0%
1.4%
1.1%
0.8%
0.8%
2.2%
1.9%
12:25:00
A
B
0.0%
0.0%
0.9%
0.9%
1.0%
0.9%
1.9%
1.8%
12:30:00
A
B
0.0%
0.0%
0.6%
0.6%
0.6%
0.8%
1.2%
1.4%
12:35:00
A
B
0.0%
0.0%
1.6%
1.3%
1.1%
0.9%
2.7%
2.2%
197

ResHostByLink Macros
Function
Macro...
Reports the host traffic for...
ResHostByLink
every communication link in the system.
ResHostOneNode
the communication links of a specific node.
ResHostByGroup
the communication links of a node grouping.

EXEC ResHostByLink
Note: The ResHostByLink macro syntax does not include the FromNode and ToNode
parameters to specify a range of nodes.
EXEC ResHostOneNode
EXEC ResHostByGroup
See Macro Execution on page 31 for a description of the FromDate, FromTime, ToDate,
ToTime, and Node parameters.
Usage Notes
The ResHostByLink macros help you answer the following questions:
Is my set up correct?
Am I making good use of the channels? If not, how high are they? If not high, then there
may not be enough host resources.
Study the incoming traffic. Problems with incoming traffic may be simply caused by an
incorrect configuration. Once configured correctly, if there is still a traffic problem, consider
studying the TCP/IP network traffic, for example, when doing an export, the ResUsageSpma
table may show 30 million rows/log period.
There must be ResUsageShst table rows logged for the period of time that the macro is going
to produce a report for.
Output Examples
The reports in the following sections are sample output reports from the ResHostByLink, the
ResHostOneNode, and the ResHostByGroup macros, respectively, where:
198

Column...
Reports the...
Host Type
type of host connection:
IBMECS
IBMMUX
IBMNET
NETWORK
Host Id
Logical ID of host (or client) with sessions logged on.
KBs Read/ Sec
number of KB read per second.
KBs Write/ Sec
number of KB written per second.
Blks Read/ Sec
number of successful blocks read per second.
Blks Write/ Sec
number of successful blocks written per second.
Blk Read Fail %
percentage of block read attempts that failed.
Blk Write Fail %
percentage of block write attempts that failed.
KBs/Blk Read
average number of KB per block read.
KBs/Blk Write
average number of KB per block written.
Msgs/Blk Read
average number of messages per block read.
Msgs/Blk Write
average number of messages per block written.
Avg ReqQ Len
average number of messages queued for output to the host.
Max ReqQ Len
maximum number of messages queued for output to the host.
ResHostByLink Sample Output

Note: The NodeID column only appears in the ResHostByLink output report.
00/10/16
HOST COMMUNICATIONS BY COMMUNICATION LINK
Page
KBs
KBs
Blks Blks
Blk
Blk
KBs
KBs Msgs
Msgs
Avg
Max
Node Vproc
Host Host Read
Write Read Write Read Write /Blk
/Blk /Blk
/Blk ReqQ ReqQ
Date
Time
Id
Id
Type
Id /Sec
/Sec
/Sec
/Sec Fail% Fail% Read
Write Read Write Len
Len
------- ------- ------ ----- ------- ----- ----- ------ ----- ----- ----- ----- ------ ------ ----- ----- ----- ---00/10/16 11:07:00 105-04 65535 NETWORK
0 24.0
13.3
0.1
0.1
0.0%
0.0% 350.5
186.2 0.8
0.8
0.0
0.0
IBMMUX 101 0.0
0.0
0.0
0.0
?
?
?
?
?
0.0
0.0
0.0
105-05 65535 NETWORK
0
IBMMUX 202
IBMMUX 304
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
IBMMUX 101
105-05 65535 NETWORK
0
IBMMUX 202
IBMMUX 304
0.0
4.1
0.0
0.0
0.0
22.6
0.0
0.0
0.0
0.1
0.0
0.0
0.0
0.1
0.0
0.0
?
?
?
?
0.0%
?
?
?
?
?
?
0.0%
?
?
?
?
?
?
?
?
?
?
?
?
?
?
0.0
0.0
0.0
0.0
0.0
0.0
?
391.8
?
?
?
206.5
?
?
?
0.9
?
?
?
0.9
?
?
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
199

ResLdvByNode Macros
ResHostOneNode Sample Output

00/10/16
Host Communications for Node 105-05

KBs
KBs
Read
Write
/Sec
/Sec
------- -------0.0
0.0
0.0
0.0
0.0
0.0
Blks Blks
Blk
Blk
Read Write Read Write
/Sec /Sec Fail% Fail%
----- ----- ----- ----0.0
0.0
?
?
0.0
0.0
?
?
0.0
0.0
?
?
Page
KBs
KBs
/Blk
/Blk
Read
Write
------ ------?
?
?
?
?
?
Vproc Host
Date
Time
Id Type
-------- -------- ----- -------00/10/16 11:07:00 65535 NETWORK
IBMMUX
IBMMUX
Host
Id
----0
202
304
11:22:42 65535 NETWORK

IBMMUX
IBMMUX
0
202
304
44.1
0.0
0.0
22.6
0.0
0.0
0.1
0.0
0.0
0.1
0.0
0.0
0.0%
?
?
0.0%
?
?
391.8
?
?
11:32:42 65535 NETWORK

IBMMUX
IBMMUX
0
202
304
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
?
?
?
?
?
?
?
?
?
Msgs Msgs
Avg
Max
/Blk /Blk ReqQ ReqQ
Read Write
Len
Len
----- ----- ----- --?
?
0.0
0.0
?
?
0.0
0.0
?
?
0.0
0.0
206.5
?
?
?
?
?
0.9
?
?
0.9 0.0
?
0.0
?
0.0
0.0
0.0
0.0
?
?
?
?
?
?
0.0
0.0
0.0
0.0
0.0
0.0
ResHostByGroup Sample Output

Note: The GroupID column only appears in the ResHostByGroup output report.
Date
-------00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
00/10/16
Time
-------11:30:00
11:30:00
11:35:00
11:35:00
11:40:00
11:40:00
11:45:00
11:45:00
11:50:00
11:50:00
11:55:00
11:55:00
12:00:00
12:00:00
12:05:00
12:05:00
12:10:00
12:10:00
12:15:00
12:15:00
12:20:00
12:20:00
12:25:00
12:25:00
12:30:00
Group
Id
-----A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
Host
Type
-------NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
NETWORK
KBs
Read
/Sec
-------0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
KBs
Write
/Sec
-------0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Blks
Read
/Sec
----0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Blks
Write
/Sec
----0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Blk
Read
Fail%
----?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Blk
Write
Fail%
----?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
KBs
/Blk
Read
------?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
KBs
/Blk
Write
------?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Msgs
/Blk
Read
----?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Msgs
/Blk
Write
----?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Avg
ReqQ
Len
----0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Max
ReqQ
Len
----0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
ResLdvByNode Macros
Function
Macro...
Reports the logical device traffic channeled through...
ResLdvByNode
each node by totaling its controller links into one summarized node
output line.
ResLdvOneNode
a specific node by totaling all its controller links into one summarized
node output line.
ResLdvByGroup
a node grouping.

200

ResLdvByNode Macros
EXEC ResLdvByNode
EXEC ResLdvOneNode
EXEC ResLdvByGroup
Usage Notes
You must have logging enabled on the ResUsageSldv table.
Output Examples
The reports in the following sections are sample output reports from the ResLdvByNode, the
ResLdvOneNode, and the ResLdvByGroup macros, respectively, where:
Column...
Reports the...
Reads / Sec
average number of logical device reads per second.
Writes / Sec
average number of logical device writes per second.
Rd KB/ I/O
average number of KB per logical device read.
Wrt KB / I/O
average number of KB per logical device write.
Avg I/O Resp
average response time for a logical device read or write in seconds.
Max Concur Rqsts
maximum number of concurrent requests during the log period.
Avg Out Rqsts
average number of outstanding requests.
Out Rqst Time %
percent of time there are outstanding requests.
ResLdvByNode Sample Output

Note: The NodeID column only appears in the ResLdvByNode output report.
06/09/26
LOGICAL DEVICE TRAFFIC BY NODE

Ldv
Date Type Time
-------- ---- -------06/09/26 DISK 10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
10:12:15
10:12:30
10:12:45
Node
Id
-----001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
Reads
/ Sec
-------0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Writes
/ Sec
-------2.00
1.27
2.20
1.20
3.53
1.33
2.00
1.33
1.87
40.67
3.40
5.40
1.87
KB
/ I/O
-----4.20
5.89
5.88
6.22
3.96
5.85
4.30
5.65
8.71
31.27
16.57
7.44
14.29
Avg
I/O
Resp
------0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
Page
1
Avg
Out
Out
Rqst
Rqsts Time %
----- ----0.0
1.3%
0.0
0.6%
0.0
1.3%
0.0
0.8%
0.0
2.5%
0.0
1.0%
0.0
1.3%
0.0
0.7%
0.0
1.0%
4.0 100.0%
0.0
1.4%
0.0
5.2%
0.0
0.9%
201

ResLdvByNode Macros
SDSK 10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
0.13
0.07
0.00
0.44
0.50
0.48
0.51
0.88
0.97
0.98
1.00
7.98
9.21
8.73
9.01
8.45
8.85
2.72
0.34
0.28
55.42
109.53
111.57
107.32
98.02
100.64
100.83
410.24
******
******
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.0
1.2
1.3
1.1
1.1
1.0
1.0
0.3
0.1
0.0
3.2%
9.1%
9.1%
9.1%
11.2%
9.1%
9.1%
9.1%
9.2%
8.6%
ResLdvOneNode Sample Output

06/09/26
LOGICAL DEVICE TRAFFIC FOR NODE 001-01
Ldv
Date Type
-------- ---06/09/26 DISK
SDSK
Time
-------10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
10:12:15
10:12:30
10:12:45
Reads
/ Sec
-------0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Writes
/ Sec
-------2.00
1.27
2.20
1.20
3.53
1.33
2.00
1.33
1.87
40.67
3.40
5.40
1.87
KB
/ I/O
-----4.20
5.89
5.88
6.22
3.96
5.85
4.30
5.65
8.71
31.27
16.57
7.44
14.29
Avg
I/O
Resp
------0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
Avg
Out
Rqsts
----0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
4.0
0.0
0.0
0.0
10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
10:12:15
10:12:30
10:12:45
0.13
0.07
0.00
0.44
0.50
0.48
0.51
0.88
0.97
0.98
0.59
0.08
0.00
1.00
7.98
9.21
8.73
9.01
8.45
8.85
2.72
0.34
0.28
3.14
4.15
0.39
55.42
109.53
111.57
107.32
98.02
100.64
100.83
410.24
******
******
285.78
16.99
32.94
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.0
1.2
1.3
1.1
1.1
1.0
1.0
0.3
0.1
0.0
0.1
0.1
0.0
Out
Rqst
Time %
-----1.3%
0.6%
1.3%
0.8%
2.5%
1.0%
1.3%
0.7%
1.0%
100.0%
1.4%
5.2%
0.9%
3.2%
9.1%
9.1%
9.1%
11.2%
9.1%
9.1%
9.1%
9.2%
8.6%
9.1%
9.2%
1.0%
ResLdvByGroup Sample Output

Note: The GroupID column only appears in the ResLdvByGroup output report.
06/09/26
Grp
Date Id
-------- --06/09/26 A
LOGICAL DEVICE TRAFFIC BY GROUP Page
Ldv
Type
---DISK
SDSK
202
Time
--------10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
10:12:15
10:12:30
10:12:45
Reads
/ Sec
-------0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10:09:45
10:10:00
10:10:15
0.13
0.07
0.00
Writes
Rd KB
Wrt KB
/ Sec
/ I/O
/ I/O
-------- ------- -----2.00
?
4.20
1.27
?
5.89
2.20
?
5.88
1.20
?
6.22
3.53
?
3.96
1.33
?
5.85
2.00
?
4.30
1.33
?
5.65
1.87
?
8.71
40.67
?
31.27
3.40
?
16.57
5.40
?
7.44
1.87
?
14.29
1.00
7.98
9.21
121.27
139.00
?
46.64
109.28
111.57
Avg
I/O
Resp
------0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
Max
Concur
Rqsts
----0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Out
Rqst
Time %
-----1.3%
0.6%
1.3%
0.8%
2.5%
1.0%
1.3%
0.7%
1.0%
100.0%
1.4%
5.2%
0.9%
3.2%
9.1%
9.1%

ResLdvByNode Macros
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
10:12:15
10:12:30
10:12:45
0.44
0.50
0.48
0.51
0.88
0.97
0.98
0.59
0.08
0.00
8.73
9.01
8.45
8.85
2.72
0.34
0.28
3.14
4.15
0.39
113.33
12.12
12.01
12.08
******
******
******
******
295.14
?
107.02
102.76
105.72
105.93
94.90
39.43
50.13
9.85
11.29
32.94
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
9.1%
11.2%
9.1%
9.1%
9.1%
9.2%
8.6%
9.1%
9.2%
1.0%
203

ResMemMgmtByNode Macros
Function
Macro...
Reports memory management activity for...
ResMemMgmtByNode
each individual node.
ResMemMgmtOneNode
a specific node.
ResMemByGroup
a node grouping.

EXEC ResMemMgmtByNode
EXEC ResMemMgmtOneNode
EXEC ResMemByGroup
Usage Notes
Output Examples
The reports in the following sections are sample output reports from the
ResMemMgmtByNode, the ResMemMgmtOneNode macros, and the ResMemByGroup,
respectively, where:
Column...
Reports the...
% Mem Free
current snapshot of the percent of memory that is unused.
Text Alocs/ Sec
average number of text page allocations per second.

text pages are allocations of memory for code that is not associated with
system-level overhead tasks.
204
VPR Alocs/ Sec
average number of vproc-specific page and segment allocations per

second on a node.
KB/ VPR Aloc
average KB per vproc-specific page and segment allocation on a node.
Aloc Fail %
percent of memory allocation attempts that failed.

Column...
Reports the...
Ages/ Sec
average number of times memory pages were aged out per second.
# Proc Swp
current number of processes that are swapped out.
Page Drops/ Sec
average number of text pages dropped from memory per second.

Page drops are text pages that are dropped from memory to increase the
amount of available memory.
Page Reads KBs/ Sec
average number of memory paging KB, read from disk per second.
Page reads include both memory text pages and task context pages, such
as scratch, stack, etc.
Page Writes/ Sec
average number of memory pages written to disk per second.

Page writes include only task context pages.
Swap Drops/ Sec
average number of disk segments dropped from memory per second.

Swap drops include all disk segments dropped from memory because
their ancestor processes were swapped out.
Swap Reads/ Sec
average number of disk segments reread back into memory, after being
swapped, out per second.
Swap reads include all reread disk segments that had been previously
dropped from memory because their ancestor processes were swapped
out.
KB/Swp Drp
average size, in KB, of disk segments dropped from memory because their
ancestor processes were swapped out.
KB/Swp Rd
average size, in KB, of reread disk segments that had been previously
dropped from memory because their ancestor processes were swapped
out.
P+S Drops/ Sec
average number of paged, swapped page, or segment drops per second.

This statistic includes both the memory text pages (Pg Drps/ Sec), and the
disk segments (Swp Drps/ Sec), that were dropped.
P+S Read KBs/ Sec
average total number of paged, swapped page, or segment read KB, per
second. Includes both the memory text pages and task context pages (Pg
Rds/ Sec), and the disk segments (Swp Rds/ Sec), reread back into
memory after being swapped out.
P+S Writes/ Sec
average total number of paged, swapped page, or segment writes per

second.
P+S IO KB %
percent of logical device input and output KB done for paging or

swapping inputs and outputs.
205

ResMemMgmtByNode Sample Output

Note: The NodeID column only appears in the ResMemMgmtByNode output report.
11/02/01
MEMORY MANAGEMENT ACTIVITY BY NODE
Date
-------11/02/01
Time
-------09:15:00
09:16:00
09:17:00
09:18:00
09:19:00
09:20:00
09:21:00
09:22:00
09:23:00
09:24:00
09:25:00
09:26:00
09:27:00
09:28:00
09:29:00
09:30:00
09:31:00
09:32:00
09:33:00
09:34:00
09:35:00
09:36:00
09:37:00
Node
Id
-----001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
%
Mem
Free
---26%
27%
26%
27%
26%
26%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
Page
Read KBs
/Sec
--------0
15
786
407
1,628
648
788
113
145
36
203
139
79
40
6
22
41
14
43
0
0
7
0
Page+Swap
Read KBs
/Sec
--------0
19
813
446
1,698
837
957
143
184
42
234
164
85
40
8
31
47
15
43
3
0
7
0
Page
Page
P+S
IO KB %
------0%
26%
85%
73%
14%
8%
11%
1%
1%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
ResMemMgmtOneNode Sample Output

11/02/01
Memory Management Activity for Node 001-01
Date
-------11/02/01
206
Time
-------09:15:00
09:16:00
09:17:00
09:18:00
09:19:00
09:20:00
09:21:00
09:22:00
09:23:00
09:24:00
09:25:00
09:26:00
09:27:00
09:28:00
09:29:00
09:30:00
09:31:00
09:32:00
09:33:00
09:34:00
09:35:00
09:36:00
09:37:00
09:38:00
09:39:00
09:40:00
09:41:00
09:42:00
09:43:00
09:44:00
09:45:00
09:46:00
09:47:00
09:48:00
09:49:00
%
Mem
Free
---26%
27%
26%
27%
26%
26%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
Page
Read KBs
/Sec
-------0
15
786
407
1628
648
788
113
145
36
203
139
79
40
6
22
41
14
43
0
0
7
0
0
16
2
30
23
0
0
0
1
0
0
6
Page+Swap
Read KBs
/Sec
--------0
19
813
446
1698
837
957
143
184
42
234
164
85
40
8
31
47
15
43
3
0
7
0
1
19
4
31
24
0
0
0
1
1
4
7
P+S
IO KB %
------0%
26%
85%
73%
14%
8%
11%
1%
1%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
1%
1%
0%
0%
0%
10%
1%
2%
5%

09:50:00
27%
3%
ResMemByGroup Sample Output

Note: The GroupID column only appears in the ResMemByGroup output report.
11/02/01
Date
-------11/02/01
MEMORY MGMT ACTIVITY BY GROUP
Time
-------09:15:00
09:16:00
09:17:00
09:18:00
09:19:00
09:20:00
09:21:00
09:22:00
09:23:00
09:24:00
09:25:00
09:26:00
09:27:00
09:28:00
09:29:00
09:30:00
09:31:00
09:32:00
09:33:00
09:34:00
09:35:00
09:36:00
09:37:00
Group
Id
-------AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
AMPNodes
%
Mem
Free
---26%
27%
26%
27%
26%
26%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
27%
Page
Read KBs
/Sec
--------0
15
786
407
1,628
648
788
113
145
36
203
139
79
40
6
22
41
14
43
0
0
7
0
Page
Page+Swap
Read KBs
/Sec
--------0
19
813
446
1,698
837
957
143
184
42
234
164
85
40
8
31
47
15
43
3
0
7
0
P+S
IO KB %
------0%
26%
85%
73%
14%
8%
11%
1%
1%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
207

ResNetByNode Macros
ResNetByNode Macros
Function
Macro...
Reports net traffic for...
ResNetByNode
each node.
ResNetOneNode
a specific node.
ResNetByGroup
nodes summarized by node groups.

EXEC ResNetByNode
EXEC ResNetOneNode
EXEC ResNetByGroup
Usage Notes
Output Examples
The reports in the following sections are sample output reports from the ResNetByNode, the
ResNetOneNode, and the ResNetByGroup macros, respectively, where:
Column...
Reports the...
% Retries
percent of total net circuit attempts that caused software backoffs (BNS
service-blocked occurrences).
Note: This value reflects how many times the hardware backed off a
connection because the switch nodes could not route to the end point. That
implies that the end point was busy or, in switch node terms, the routing
path was busy. A value over 100% does not imply a problem, but shows that
there were multiple attempts to send new messages while the Bynet path was
busy. On a busy system, this can be a normal level of activity.
208
Total Reads/ Sec
average number of net reads per second.
Total Writes/ Sec
average number of net writes per second.

ResNetByNode Macros
Column...
Reports the...
Total IOs/ Sec
average number of net reads and writes per second.
KB/ IO
average KB per net read or write.
% PtP
percent of total net reads and writes that are point-to-point reads and writes.
% Brd
percent of total net reads and writes that are broadcast reads and writes.
Note: In the following examples, the NodeID column appears only in the ResNetByNode
output report. The GroupID column only appears in the ResNetByGroup output report. For
all the examples, the values in the Total Reads/ Sec and Total Writes/ Sec are expected to be
equal on SMP (single-node, vnet) systems.
ResNetByNode Sample Output

00/10/16
NET ACTIVITY BY NODE
Date
Time
---------- -------2000/10/16 11:20:00
Page 2
Total
Total
Total
Node
% ReReads
Writes
IOs
KB
Id
tries
/Sec
/Sec
/Sec
/IO
------ ------ ------- ------- ------- -----001-03
0.0%
0.46
0.39
0.85
0.4
001-04
0.0%
0.55
0.47
1.02
0.4
%
PtP
--92
93
%
Brd
--8
7
11:25:00
001-03
001-04
0.0%
0.0%
0.39
0.39
0.33
0.32
0.72
0.71
0.4
0.4
90
90
10
10
11:30:00
001-03
001-04
0.0%
0.0%
0.44
0.55
0.37
0.47
0.81
1.02
0.4
0.4
91
92
9
8
11:35:00
001-03
001-04
2.5%
2.5%
20.84
23.07
12.53
15.51
33.37
38.58
1.8
1.8
73
74
27
26
11:40:00
001-03
001-04
24.7%
20.6%
35.44
37.16
35.56
38.87
71.00
76.03
17.4
13.8
93
93
7
7
11:45:00
001-03
001-04
15.9%
28.1%
13.47
11.79
10.71
12.63
24.18
24.42
8.3
12.8
76
83
24
17
11:50:00
001-03
001-04
3.3%
4.1%
18.92
22.77
14.18
20.97
33.11
43.74
1.3
1.9
77
75
23
25
ResNetOneNode Sample Output

00/10/16
Net Activity for Node 001-03
Date
-------00/10/16
Time
-------10:19:00
10:20:00
10:21:00
10:22:00
10:23:00
10:30:00
10:35:00
10:40:00
10:45:00
10:50:00
10:55:00
11:00:00
11:05:00
11:10:00
11:15:00
11:20:00
% Retries
-----0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Total
Reads
/Sec
------0.78
2.87
3.08
2.13
2.23
0.25
0.53
0.51
0.48
0.52
0.46
0.58
0.57
0.54
0.46
0.46
Total
Writes
/Sec
------0.07
2.65
2.07
2.07
2.17
0.18
0.47
0.44
0.42
0.45
0.39
0.51
0.38
0.47
0.40
0.39
Total
IOs
/Sec
------0.85
5.52
5.15
4.20
4.40
0.43
1.00
0.95
0.90
0.97
0.85
1.09
0.95
1.01
0.86
0.85
Page 1
KB
/IO
-----1.1
0.5
0.6
0.5
0.5
0.4
0.4
0.5
0.4
0.5
0.4
0.4
0.5
0.4
0.5
0.4
%
PtP
--8
96
80
98
98
84
93
93
92
93
92
94
79
93
92
92
%
Brd
--92
4
20
2
2
16
7
7
8
7
8
6
21
7
8
8
209

ResNetByNode Macros
ResNetByGroup Sample Output

00/10/16
NET ACTIVITY BY Group
Date
---------2000/10/16
210
Group
% ReTime
Id
tries
-------- ------ -----11:20:00
B
0.0%
Page 2
Reads
Writes
/Sec
/Sec
------- ------0.55
0.47
IOs
/Sec
------1.02
KB
%
%
/IO PtP Brd
------ --- --0.4
93
7
11:25:00
A
B
0.0%
0.0%
0.39
0.39
0.33
0.32
0.72
0.71
0.4
0.4
90
90
10
10
11:30:00
A
B
0.0%
0.0%
0.44
0.55
0.37
0.47
0.81
1.02
0.4
0.4
91
92
9
8
11:35:00
A
B
2.5%
2.5%
20.84
23.07
12.53
15.51
33.37
38.58
1.8
1.8
73
74
27
26
11:40:00
A
B
24.7%
20.6%
35.44
37.16
35.56
38.87
71.00
76.03
17.4
13.8
93
93
7
7
11:45:00
A
B
15.9%
28.1%
13.47
11.79
10.71
12.63
24.18
24.42
8.3
12.8
76
83
24
17
11:50:00
A
B
3.3%
4.1%
33.11
43.74
1.3
1.9
77
75
23
25
11:55:00
A
B
55.8%
41.1%
73.46
88.72
22.3
17.7
95
96
5
4
12:00:00
A
B
5.1%
5.8%
32.27
33.16
2.0
1.7
73
70
27
30
12:05:00
A
B
24.2%
10.6%
33.09
26.97
28.57
25.97
61.67
52.94
14.6
5.5
90
91
10
9
12:10:00
A
B
73.8%
57.4%
17.33
28.12
14.01
26.65
31.34
54.77
23.2
23.0
91
93
9
7
12:15:00
A
B
3.9%
6.3%
37.12
36.83
2.0
1.8
73
73
27
27
12:20:00
A
B
48.7%
34.9%
70.83
71.86
18.0
13.2
95
93
5
7
18.92
22.77
40.01
44.35
19.11
22.13
21.02
22.13
36.18
38.16
14.18
20.97
33.45
44.37
13.16
11.03
16.10
14.70
34.65
33.70

ResNode Macros
ResNode Macros
Function
Macro...
Provides a summary of resource usage ...
ResNode
averaged across all nodes, excluding any PE-only nodes.
ResOneNode
by returning the node requested.
ResNodeByNode
by returning the nodes requested.
ResNodeByGroup
by defined groups of nodes.

The input forms of these four macros are described below.
EXEC ResNode
Note: The ResNode macro syntax does not include the FromNode and ToNode parameters to
specify a range of nodes.
EXEC ResOneNode
EXEC ResNodeByNode
EXEC ResNodeByGroup
ToTime, FromNode, ToNode, and Node parameters.
Usage Notes
Output Examples
The reports in the following sections are sample output reports from the ResNode, the
ResOneNode, the ResNodebyNode, and the ResNodeByGroup macros, respectively.
ResNode output report.
Statistics columns
Description
1 through 3
CPU usage.
4 through 8
Logical device interface.
211

ResNode Macros
Statistics columns
Description
9 through 14
Memory interface.
15 through 17
Net interface.
18 and 19
General node process scheduling.
ResOneNode output report.
Statistics columns
Description
1 and 2
CPU usage.
3 through 6
7 through 11
Memory interface.
12 through 14
Net interface.
15 and 16
ResNodebyNode output report.
Statistics columns
Description
1 and 2
CPU usage.
3 through 6
7 through 12
Memory interface.
13 through 15
Net interface.
16 and 17
ResNodeByGroup output report.
212
Statistics columns
Description
GroupId as defined in the associated view as a grouping of one or

more nodes.
2 and 3
CPU usage.
4 through 7
8 through 12
Memory interface.

ResNode Macros
Statistics columns
Description
13 through 15
Net interface.
16 and 17
The following table describes the statistics columns in all output reports (with the exception
of ResNodeByNode, which has a NodeID column, and ResNodeByGroup, which has a
GroupId column).
Column
Reports the
CPU Bsy %
percent of time the CPUs are busy, based on average CPU usage
per node.
CPU Eff % (ResNode report)
parallel efficiency of node CPU usage.

Parallel efficiency is the total percent of time nodes are busy. It is
the average for all nodes of total busy divided by the total busy
time of the busiest node.
WIO %
percent of time the CPUs are idle and waiting for completion of an
I/O operation.
Ldv IO KBs /Sec
average number of logical device read and write KB per second.
Ldv Eff % (ResNode report)
parallel efficiency of the logical device (disk) I/Os. It is the average

number of I/Os per node divided by the number of I/Os
performed by the node with the most I/Os.
P+S % of IO KBs
percent of logical device read and write KB that are for paging
purposes.
Read % of IO KBs
percent of logical device read and write KB that are reads.
Ldv KB / IO
average size of a non-swap/page logical device read or write.
Fre Mem %
percent of memory that is unused.
TPtP IOs /Sec
total point-to-point net reads and writes per second, per node.
TMlt IOs /Sec
total multicast (broadcast or merge) net reads and writes per

second, per node.
Net Rtry %
percent of transmission attempts that resulted in retries.
Prc Blks / Sec
number of times per second, per node, that processes other than
message and timer waits are blocked.
ms /Blk
average time, in milliseconds, spent waiting for a blocked process

other than message and timer waits.
Net Rx % Bsy
percent of time the network was busy either receiving.
Net Tx % Bsy %
percent of time the network was busy transmitting.
213

ResNode Macros
ResNode Sample Output

11/02/01
Date
Time
-------- -------11/02/01 09:15:00
09:16:00
09:17:00
09:18:00
09:19:00
09:20:00
09:21:00
09:22:00
09:23:00
09:24:00
09:25:00
09:26:00
09:27:00
09:28:00
09:29:00
09:30:00
09:31:00
09:32:00
09:33:00
09:34:00
09:35:00
09:36:00
09:37:00
09:38:00
09:39:00
09:40:00
09:41:00
09:42:00
09:43:00
09:44:00
09:45:00
09:46:00
09:47:00
09:48:00
09:49:00
09:50:00
214
GENERAL RESUSAGE SUMMARY

Average across all nodes
CPU
Bsy
%
--0
0
1
2
32
76
45
42
37
16
16
13
4
4
5
10
9
8
5
5
4
4
4
4
3
2
40
2
0
0
0
0
0
0
1
1
Page
CPU
Ldv Ldv
P+S
Read Ldv Fre TPtP TMlt Net Proc
ms Net Net
Eff WIO
IO KBs Eff
%of
% of
KB Mem
IOs
IOs Rtry Blks
/ Rx % Tx %
%
%
/Sec
% IO KBs IO KBs /IO
% /Sec /Sec
% /Sec
Blk Busy Busy
--- --- ---------- --- ------ ------ ---- --- ----- ----- ---- ----- ------ ---- ---100
0
70 100
0
16
21 26
0
0
?
49
344
?
?
100
1
73 100
26
37
17 27
0
0
?
49
1735
?
?
100
6
954 100
85
88
25 26
0
0
?
56
1804
?
?
100 14
609 100
73
76
18 27
0
0
?
63
685
?
?
100 57
11,719 100
14
46
74 26
0
0
? 2385
16
?
?
100 23
10,706 100
8
40
72 26
0
0
?
187
260
?
?
100 47
11,607 100
11
11
87 27
0
0
? 2215
19
?
?
100 54
26,460 100
1
14
92 27
0
0
?
747
659
?
?
100 61
24,873 100
1
8
64 27
0
0
?
607
69
?
?
100 81
41,507 100
0
1
64 27
0
0
?
710
44
?
?
100 84
50,374 100
0
62 157 27
0
0
?
383
152
?
?
100 87
54,921 100
0
100 1546 27
0
0
?
140
330
?
?
100 96
58,023 100
0
100 1589 27
0
0
?
164
244
?
?
100 96
58,644 100
0
100 1683 27
0
0
?
156
248
?
?
100 95
59,642 100
0
100 1558 27
0
0
?
182
452
?
?
100 87
36,044 100
0
63 117 27
0
0
?
431
104
?
?
100 91
27,847 100
0
57
95 27
0
0
?
441
81
?
?
100 90
32,465 100
0
53 100 27
0
0
?
483
68
?
?
100 95
32,193 100
0
50 123 27
0
0
?
365
75
?
?
100 95
33,125 100
0
50 124 27
0
0
?
300
94
?
?
100 96
27,514 100
0
50 124 27
0
0
?
287
219
?
?
100 96
27,725 100
0
50 123 27
0
0
?
244
109
?
?
100 96
27,764 100
0
50 125 27
0
0
?
200
141
?
?
100 96
25,126 100
0
59 122 27
0
0
?
201
370
?
?
100 97
18,595 100
0
99 120 27
0
0
?
257
503
?
?
100 79
13,516 100
0
97 114 27
0
0
?
195
627
?
?
100 47
4,406 100
1
54
57 27
0
0
?
122
474
?
?
100 27
2,039 100
1
49
35 27
0
0
?
113
309
?
?
100
1
170 100
0
53
32 27
0
0
?
51
503
?
?
100
2
255 100
0
53
32 27
0
0
?
54
593
?
?
100
1
118 100
0
48
25 27
0
0
?
50
335
?
?
100
1
158 100
10
38
25 27
0
0
?
51
329
?
?
100
1
156 100
1
45
28 27
0
0
?
52
1450
?
?
100
1
198 100
2
46
29 27
0
0
?
53
1596
?
?
100
1
129 100
5
47
25 27
0
0
?
52
1284
?
?
100
1
97 100
3
41
23 27
0
0
?
50
955
?
?

ResNode Macros
ResOneNode Sample Output

11/02/01
Date
Time
-------- -------11/02/01 09:15:00
09:16:00
09:17:00
09:18:00
09:19:00
09:20:00
09:21:00
09:22:00
09:23:00
09:24:00
09:25:00
09:26:00
09:27:00
09:28:00
09:29:00
09:30:00
09:31:00
09:32:00
09:33:00
09:34:00
09:35:00
09:36:00
09:37:00
09:38:00
09:39:00
09:40:00
09:41:00
09:42:00
09:43:00
09:44:00
09:45:00
09:46:00
09:47:00
09:48:00
09:49:00
09:50:00
General Resource Usage Summary for Node 001-01
Page
CPU
Ldv
P+S
Read
Ldv Fre TPtP TMlt Net Proc
ms Net Net
Bsy WIO
IO KBs
%of
% of
KB Mem
IOs
IOs Rtry Blks
/ Rx % Tx %
%
%
/Sec IO KBs IO KBs
/IO
% /Sec /Sec
% /Sec
Blk Busy Busy
--- --- ---------- ------ ------ ------ --- ----- ----- ---- ----- ------ ---- ---0
0
70
0
16
21 26
0
0
?
49
344
?
?
0
1
73
26
37
17 27
0
0
?
49
1735
?
?
1
6
954
85
88
25 26
0
0
?
56
1804
?
?
2 14
609
73
76
18 27
0
0
?
63
685
?
?
32 57
11,719
14
46
74 26
0
0
? 2385
16
?
?
76 23
10,706
8
40
72 26
0
0
?
187
260
?
?
45 47
11,607
11
11
87 27
0
0
? 2215
19
?
?
42 54
26,460
1
14
92 27
0
0
?
747
659
?
?
37 61
24,873
1
8
64 27
0
0
?
607
69
?
?
16 81
41,507
0
1
64 27
0
0
?
710
44
?
?
16 84
50,374
0
62
157 27
0
0
?
383
152
?
?
13 87
54,921
0
100 1,546 27
0
0
?
140
330
?
?
4 96
58,023
0
100 1,589 27
0
0
?
164
244
?
?
4 96
58,644
0
100 1,683 27
0
0
?
156
248
?
?
5 95
59,642
0
100 1,558 27
0
0
?
182
452
?
?
10 87
36,044
0
63
117 27
0
0
?
431
104
?
?
9 91
27,847
0
57
95 27
0
0
?
441
81
?
?
8 90
32,465
0
53
100 27
0
0
?
483
68
?
?
5 95
32,193
0
50
123 27
0
0
?
365
75
?
?
5 95
33,125
0
50
124 27
0
0
?
300
94
?
?
4 96
27,514
0
50
124 27
0
0
?
287
219
?
?
4 96
27,725
0
50
123 27
0
0
?
244
109
?
?
4 96
27,764
0
50
125 27
0
0
?
200
141
?
?
4 96
25,126
0
59
122 27
0
0
?
201
370
?
?
3 97
18,595
0
99
120 27
0
0
?
257
503
?
?
2 79
13,516
0
97
114 27
0
0
?
195
627
?
?
40 47
4,406
1
54
57 27
0
0
?
122
474
?
?
2 27
2,039
1
49
35 27
0
0
?
113
309
?
?
0
1
170
0
53
32 27
0
0
?
51
503
?
?
0
2
255
0
53
32 27
0
0
?
54
593
?
?
0
1
118
0
48
25 27
0
0
?
50
335
?
?
0
1
158
10
38
25 27
0
0
?
51
329
?
?
0
1
156
1
45
28 27
0
0
?
52
1450
?
?
0
1
198
2
46
29 27
0
0
?
53
1596
?
?
1
1
129
5
47
25 27
0
0
?
52
1284
?
?
1
1
97
3
41
23 27
0
0
?
50
955
?
?
ResNodeByNode Sample Output

11/02/01
Date
Time NodeID
-------- -------- -----11/02/01 09:15:00 001-01
09:16:00 001-01
09:17:00 001-01
09:18:00 001-01
09:19:00 001-01
09:20:00 001-01
09:21:00 001-01
09:22:00 001-01
09:23:00 001-01
09:24:00 001-01
09:25:00 001-01
09:26:00 001-01
09:27:00 001-01
09:28:00 001-01
09:29:00 001-01
09:30:00 001-01
09:31:00 001-01
09:32:00 001-01
09:33:00 001-01
09:34:00 001-01
09:35:00 001-01
09:36:00 001-01
09:37:00 001-01
Node by Node General Resource Usage Summary
Page
CPU
Ldv
P+S
Read
Ldv Fre TPtP TMlt Net Proc
ms Net Net
Bsy WIO
IO KBs
%of
% of
KB Mem
IOs
IOs Rtry Blks
/ Rx % Tx %
%
%
/Sec IO KBs IO KBs
/IO
% /Sec /Sec
% /Sec
Blk Busy Busy
--- --- ---------- ------ ------ ------ --- ----- ----- ---- ----- ------ ---- ---0
0
70
0
16
21 26
0
0
?
49
344
?
?
0
1
73
26
37
17 27
0
0
?
49
1735
?
?
1
6
954
85
88
25 26
0
0
?
56
1804
?
?
2 14
609
73
76
18 27
0
0
?
63
685
?
?
32 57
11,719
14
46
74 26
0
0
? 2385
16
?
?
76 23
10,706
8
40
72 26
0
0
?
187
260
?
?
45 47
11,607
11
11
87 27
0
0
? 2215
19
?
?
42 54
26,460
1
14
92 27
0
0
?
747
659
?
?
37 61
24,873
1
8
64 27
0
0
?
607
69
?
?
16 81
41,507
0
1
64 27
0
0
?
710
44
?
?
16 84
50,374
0
62
157 27
0
0
?
383
152
?
?
13 87
54,921
0
100 1,546 27
0
0
?
140
330
?
?
4 96
58,023
0
100 1,589 27
0
0
?
164
244
?
?
4 96
58,644
0
100 1,683 27
0
0
?
156
248
?
?
5 95
59,642
0
100 1,558 27
0
0
?
182
452
?
?
10 87
36,044
0
63
117 27
0
0
?
431
104
?
?
9 91
27,847
0
57
95 27
0
0
?
441
81
?
?
8 90
32,465
0
53
100 27
0
0
?
483
68
?
?
5 95
32,193
0
50
123 27
0
0
?
365
75
?
?
5 95
33,125
0
50
124 27
0
0
?
300
94
?
?
4 96
27,514
0
50
124 27
0
0
?
287
219
?
?
4 96
27,725
0
50
123 27
0
0
?
244
109
?
?
4 96
27,764
0
50
125 27
0
0
?
200
141
?
?
215

ResNode Macros
ResNodeByGroup Sample Output

11/02/01
GENERAL RESOURCE USAGE SUMMARY BY GROUP
Page
CPU
Ldv
P+S
Read
Ldv Fre TPtP TMlt Net
Proc
ms
Net Net
Group
Bsy WIO
IO KBs
%of
% of
KB Mem
IOs IOs
Rtry Blks
/
Rx % Tx %
Date
Time
Id
%
%
/Sec IO KBs IO KBs
/IO
% /Sec /Sec
% /Sec
Blk Busy Busy
-------- -------- -------- --- --- ---------- ------ ------ ------ --- ----- ----- ---- ----- ------ ---- ---11/02/01 09:15:00 AMPNodes
0
0
70
0
16
21 26
0
0
?
49
344
?
?
09:16:00 AMPNodes
0
1
73
26
37
17 27
0
0
?
49
1735
?
?
09:17:00 AMPNodes
1
6
954
85
88
25 26
0
0
?
56
1804
?
?
09:18:00 AMPNodes
2 14
609
73
76
18 27
0
0
?
63
685
?
?
09:19:00 AMPNodes 32 57
11,719
14
46
74 26
0
0
? 2385
16
?
?
09:20:00 AMPNodes 76 23
10,706
8
40
72 26
0
0
?
187
260
?
?
09:21:00 AMPNodes 45 47
11,607
11
11
87 27
0
0
? 2215
19
?
?
09:22:00 AMPNodes 42 54
26,460
1
14
92 27
0
0
?
747
659
?
?
09:23:00 AMPNodes 37 61
24,873
1
8
64 27
0
0
?
607
69
?
?
09:24:00 AMPNodes 16 81
41,507
0
1
64 27
0
0
?
710
44
?
?
09:25:00 AMPNodes 16 84
50,374
0
62
157 27
0
0
?
383
152
?
?
09:26:00 AMPNodes 13 87
54,921
0
100 1,546 27
0
0
?
140
330
?
?
09:27:00 AMPNodes
4 96
58,023
0
100 1,589 27
0
0
?
164
244
?
?
09:28:00 AMPNodes
4 96
58,644
0
100 1,683 27
0
0
?
156
248
?
?
09:29:00 AMPNodes
5 95
59,642
0
100 1,558 27
0
0
?
182
452
?
?
09:30:00 AMPNodes 10 87
36,044
0
63
117 27
0
0
?
431
104
?
?
09:31:00 AMPNodes
9 91
27,847
0
57
95 27
0
0
?
441
81
?
?
09:32:00 AMPNodes
8 90
32,465
0
53
100 27
0
0
?
483
68
?
?
09:33:00 AMPNodes
5 95
32,193
0
50
123 27
0
0
?
365
75
?
?
09:34:00 AMPNodes
5 95
33,125
0
50
124 27
0
0
?
300
94
?
?
09:35:00 AMPNodes
4 96
27,514
0
50
124 27
0
0
?
287
219
?
?
09:36:00 AMPNodes
4 96
27,725
0
50
123 27
0
0
?
244
109
?
?
09:37:00 AMPNodes
4 96
27,764
0
50
125 27
0
0
?
200
141
?
?
216

ResPdskByNode Macros
Function
Macro...
Reports the device traffic...
ResPdskByNode
by a physical node.
ResPdskOneNode
for a specified node.
ResPdskByGroup
node grouping.

EXEC ResPdskByNode
EXEC ResPdskOneNode
EXEC ResPdskByGroup
Usage Notes
You must have logging enabled on the ResUsageSpdsk table.
Output Examples
of ResPdiskByNode, which reports by NodeID columns, and ResPdiskByGroup, which reports
by NodeType column).
Column...
Reports the...
Reads/Sec
average number of device reads per second.
Writes/Sec
average number of device writes per second.
Rd KB/ I/O
average number of KB per device read.
Wrt KB/ I/O
average number of KB per device write.
Avg I/O Resp
average response time for a device read or write in seconds.
Max Concur Rqsts
Out Rqst Time %
217

ResPdskByNode Sample Output

Note: The NodeID column only appears in the ResPdskByNode output report.
07/11/28
Date
-------07/11/28
218
PDISK TRAFFIC BY NODE
Pdisk
Type
-----DISK
Page
Time
-------13:20:00
Node
Id
-----001-01
Reads
/ Sec
-------0.10
Writes
/ Sec
-------0.23
Rd KB
/ I/O
------*******
Wrt KB
/ I/O
------*******
Avg
I/O
Resp
------0.004
13:21:00
001-01
0.12
0.41
*******
4202.15
0.002
0.0%
13:22:00
001-01
0.13
0.49
*******
3623.05
0.001
0.0%
13:24:00
001-01
0.05
0.20
4717.71
5262.77
0.002
0.0%
13:25:00
001-01
0.07
0.35
2560.00
4637.26
0.001
0.0%
13:26:00
001-01
0.11
0.42
6537.85
4485.12
0.001
0.0%
13:28:00
001-01
0.06
0.19
3396.27
4785.63
0.001
0.0%
13:29:00
001-01
0.14
0.45
3602.29
4943.64
0.003
0.0%
13:30:00
001-01
0.12
0.40
5961.14
5274.67
0.002
0.0%
13:32:00
001-01
0.07
0.18
3990.59
3856.34
0.000
0.0%
13:33:00
001-01
0.17
0.53
4532.71
5745.23
0.001
0.0%
13:34:00
001-01
0.11
0.38
5730.46
4975.30
0.002
0.0%
13:36:00
001-01
0.05
0.17
5218.46
5677.51
0.002
0.0%
13:37:00
001-01
0.18
0.51
3990.59
5125.33
0.001
0.0%
13:38:00
001-01
0.11
0.37
5218.46
4846.55
0.001
0.0%
13:39:00
001-01
0.14
0.42
3990.59
4758.59
0.001
0.0%
13:41:00
001-01
0.05
0.17
5139.69
4660.36
0.002
0.0%
13:42:00
001-01
0.11
0.28
5017.60
5546.67
0.001
0.0%
13:43:00
001-01
0.14
0.41
6731.29
5616.33
0.002
0.0%
13:45:00
001-01
0.06
0.19
4176.00
4874.04
0.001
0.0%
13:46:00
001-01
0.07
0.34
2523.43
6260.36
0.002
0.0%
13:47:00
001-01
0.13
0.38
6192.00
5990.40
0.002
0.0%
Out
Rqst
Time %
-----0.0%

ResPdskOneNode Sample Output

Note: The NodeID column only appears in the ResPdskOneNode output report.
07/11/28
PDISK Traffic for Node 001-01
Date
-------07/11/28
Pdisk
Type
-----DISK
Time
-------13:20:00
13:21:00
13:22:00
13:24:00
13:25:00
13:26:00
13:28:00
13:29:00
13:30:00
13:32:00
13:33:00
13:34:00
13:36:00
13:37:00
13:38:00
13:39:00
13:41:00
13:42:00
13:43:00
13:45:00
13:46:00
13:47:00
13:49:00
Reads
/ Sec
-------0.10
0.12
0.13
0.05
0.07
0.11
0.06
0.14
0.12
0.07
0.17
0.11
0.05
0.18
0.11
0.14
0.05
0.11
0.14
0.06
0.07
0.13
0.05
Writes
/ Sec
-------0.23
0.41
0.49
0.20
0.35
0.42
0.19
0.45
0.40
0.18
0.53
0.38
0.17
0.51
0.37
0.42
0.17
0.28
0.41
0.19
0.34
0.38
0.21
Rd KB
/ I/O
------*******
*******
*******
4717.71
2560.00
6537.85
3396.27
3602.29
5961.14
3990.59
4532.71
5730.46
5218.46
3990.59
5218.46
3990.59
5139.69
5017.60
6731.29
4176.00
2523.43
6192.00
3693.71
Page
Wrt KB
/ I/O
------*******
4202.15
3623.05
5262.77
4637.26
4485.12
4785.63
4943.64
5274.67
3856.34
5745.23
4975.30
5677.51
5125.33
4846.55
4758.59
4660.36
5546.67
5616.33
4874.04
6260.36
5990.40
5878.52
Avg
I/O
Resp
------0.004
0.002
0.001
0.002
0.001
0.001
0.001
0.003
0.002
0.000
0.001
0.002
0.002
0.001
0.001
0.001
0.002
0.001
0.002
0.001
0.002
0.002
0.002
Out
Rqst
Time %
-----0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
ResPdskByGroup Sample Output

Note: The GroupID column only appears in the ResPdskByGroup output report.
06/09/26
PDISK TRAFFIC BY GROUP
Node
Pdisk
Date Type Type
-------- ---- -----06/09/26 4400 DISK
Time
-------10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
10:12:15
10:12:30
10:12:45
Avg
ReadCnt
WriteCnt Rd KB
Wrt KB
I/O
/ Sec
/ Sec
/ I/O
/ I/O
Resp
-------- -------- ------- ------- ------0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
0.00
0.00
?
?
?
Page
Max
Concur
Rqsts
-----0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Out
Rqst
Time %
-----0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
219

ResPs Macros
ResPs Macros
Function
Macro...
Provides a summary of the Priority Scheduler resource usage...
ResPsByNode
By TheDate, TheTime, NodeId, PGId columns on SLES 10 or

earlier systems.
By the pWDId column on SLES 11 or later systems.
The ResPsByNode macro produces one row of data for every set as
defined in the GROUP BY clause.
ResPsByGroup
by coexistence group, produces one set of rows of data for each for
each node type in the system for each log interval.

The input forms of the macros are described below.
EXEC ResPsByNode
EXEC ResPsByGroup
ToTime, FromNode, ToNode, and Node parameters.
Note: Coexistence support can be accomplished using the NodeType column to do a group by
in SQL directly. Therefore, the GroupId column is not needed and the ResUsageSps table view
is not provided.
Usage Notes
To use ResPs macros, existing rows must exist in the table.
These macros can be used to report historical data.
Output Examples
The reports in the following sections are sample output reports from the ResPsByNode and
ResPsByGroup macros.
ResPsByNode output report.
220

ResPs Macros
Statistics columns
Description
Node ID.
On SLES 10 or later systems, this column contains the PG ID.

On SLES 11 or later systems, this column contains the pWDId.
Note: In the ResPs macro outputs, the RowIndex1 column
displays as PGid, instead of pWDid, when running SLES 11 or
later systems. For a description of the RowIndex1 column, see
Chapter 11: ResUsageSps.
3 through 12
Summary of the Priority Scheduler statistics.
ResPsByGroup output report.
Statistics columns
Description
Node type.
2 through 11
Summary of the Priority Scheduler statistics.
of ResSpsByNode, which reports by NodeID columns, and ResSpsByGroup, which reports by
NodeType column).
Column...
Reports the...
CPU Bsy %
percent of time the CPUs are busy, based on average CPU usage per node.
IO Blks Sec
number of logical data blocks read or written.
Num Procs
average number of tasks of online nodes.
Num Requests
number of AMP Worker Task messages or requests that got assigned AMP
Worker Tasks to them.
Avg QWait Time
average time in milliseconds that work requests waited on an input queue

Max QWait Time
maximum time in milliseconds that work requests waited on an input queue

Q Length
average number of messages queued for output to the host.
Q Len Max
maximum number of messages queued for output to the host.
Avg Svc Time
average time in milliseconds that work requests required for service
Max Svc Time
maximum time in milliseconds that work requests required for service.
For a complete description of the above columns, see Chapter 11: ResUsageSps.
221

ResPs Macros
ResPsByNode Macro Sample Output

Note: The column label, RowIndex1, displays as PGid, instead of pWDid, when running SLES
11 or later systems. For a description of the RowIndex1 column, see Chapter 11:
ResUsageSps.
ResPsByNode:
10/12/01
PS by Node Usage Summary
----0
1
250
0
1
250
CPU
Bsy
%
--0
29
0
0
26
8
IO
Blks
/Sec
----0
0
2
0
0
2
Num
Procs
------19
66
722
18
87
697
Num
Requests
-------0
0
3
0
162033
3
Avg
QWait
Time
-----?
?
0
?
1
0
Max
QWait
Time
------0
0
0
0
31
0
Q
Length
------0
0
0
0
1
0
Q
Len
Max
------0
0
0
0
2
0
Avg
Svc
Time
------?
?
36
?
5
26
Max
Svc
Time
------0
0
78
0
265
46
2-09
2-09
2-09
2-10
2-10
2-10
0
1
250
0
1
250
0
31
0
0
26
9
0
0
2
0
0
2
19
66
722
18
121
705
0
0
3
0
164728
2
?
?
0
?
1
0
0
0
0
0
31
0
0
0
0
0
1
0
0
0
0
0
2
0
?
?
25
?
5
78
0
0
46
0
218
93
14:37:00
2-09
2-09
2-09
2-10
2-10
2-10
0
1
250
0
1
250
0
30
0
0
26
8
0
0
2
0
0
2
19
66
722
18
96
698
0
0
47
0
164847
50
?
?
0
?
1
0
0
0
15
0
31
15
0
0
0
0
1
0
0
0
0
0
2
0
?
?
6
?
4
9
0
0
78
0
265
46
14:38:00
2-09
2-09
2-09
2-09
2-10
2-10
2-10
2-10
0
1
3
250
0
1
3
250
0
27
0
0
0
31
0
9
0
0
0
2
0
0
0
2
19
66
0
722
18
100
0
697
0
27
9
3
0
158281
9
12
?
3
0
0
?
1
10
0
0
15
0
0
0
31
15
0
0
0
0
0
0
1
0
0
0
0
0
0
0
2
0
0
?
3
0
78
?
5
3
14
0
31
0
156
0
234
15
78
14:39:00
2-09
2-09
2-09
2-10
2-10
2-10
0
1
250
0
1
250
0
30
0
0
23
8
0
0
2
0
0
2
19
66
722
18
91
698
0
0
4
0
165816
2
?
?
0
?
1
8
0
0
0
0
31
15
0
0
0
0
1
0
0
0
0
0
2
0
?
?
12
?
4
78
0
0
31
0
203
109
14:40:00
2-09
2-09
2-09
2-09
2-10
2-10
2-10
2-10
0
1
3
250
0
1
3
250
0
26
0
0
0
33
0
11
0
0
0
2
0
0
0
2
19
66
0
722
18
95
0
698
0
36
9
5
0
161946
9
8
?
0
0
0
?
1
2
0
0
0
0
0
0
31
15
0
0
0
0
0
0
1
0
0
0
0
0
0
0
2
0
0
?
1
0
18
?
4
0
18
0
46
0
46
0
203
0
109
PGid
Date
-------10/12/01
Page
Time
-------14:35:00
NodeID
-----2-09
2-09
2-09
2-10
2-10
2-10
14:36:00
ResPsByGroup Macro Sample Output

Note: The column label, RowIndex1, displays as PGid, instead of pWDid, when running SLES
11 or later systems. For a description of the RowIndex1 column, see Chapter 11:
ResUsageSps.
ResPsByGroup:
10/12/01
PS by Group Usage Summary
Page
CPU
IO
Avg
Max
Q
Avg
Max
Node
Bsy Blks
Num
Num
QWait QWait
Q
Len
Svc
Svc
Date
Time Type
NodeID PGid
% /Sec
Procs Requests
Time
Time
Length
Max
Time
Time
-------- -------- -------- ------ ----- --- ----- ------- -------- ------ ------- ------- ------- ------- ------10/12/01 14:35:00 5400
2-09
0
0
0
19
0
?
0
0
0
?
0
5400
2-09
1 29
0
66
0
?
0
0
0
?
0
5400
2-09
250
0
2
722
3
0
0
0
0
36
78
5400
2-10
0
0
0
18
0
?
0
0
0
?
0
5400
2-10
1 26
0
87
162033
1
31
1
2
5
265
5400
2-10
250
8
2
697
3
0
0
0
0
26
46
14:36:00 5400
5400
5400
5400
222
2-09
2-09
2-09
2-10
0
1
250
0
0
31
0
0
0
0
2
0
19
66
722
18
0
0
3
0
?
?
0
?
0
0
0
0
0
0
0
0
0
0
0
0
?
?
25
?
0
0
46
0

ResVdskByNode Macros
5400
5400
2-10
2-10
1
250
26
9
0
2
121
705
164728
2
1
0
31
0
1
0
2
0
5
78
218
93
14:37:00 5400
5400
5400
5400
5400
5400
2-09
2-09
2-09
2-10
2-10
2-10
0
1
250
0
1
250
0
30
0
0
26
8
0
0
2
0
0
2
19
66
722
18
96
698
0
0
47
0
164847
50
?
?
0
?
1
0
0
0
15
0
31
15
0
0
0
0
1
0
0
0
0
0
2
0
?
?
6
?
4
9
0
0
78
0
265
46
14:38:00 5400
5400
5400
5400
5400
5400
5400
5400
2-09
2-09
2-09
2-09
2-10
2-10
2-10
2-10
0
1
3
250
0
1
3
250
0
27
0
0
0
31
0
9
0
0
0
2
0
0
0
2
19
66
0
722
18
100
0
697
0
27
9
3
0
158281
9
12
?
3
0
0
?
1
10
0
0
15
0
0
0
31
15
0
0
0
0
0
0
1
0
0
0
0
0
0
0
2
0
0
?
3
0
78
?
5
3
14
0
31
0
156
0
234
15
78
14:39:00 5400
5400
5400
5400
5400
5400
2-09
2-09
2-09
2-10
2-10
2-10
0
1
250
0
1
250
0
30
0
0
23
8
0
0
2
0
0
2
19
66
722
18
91
698
0
0
4
0
165816
2
?
?
0
?
1
8
0
0
0
0
31
15
0
0
0
0
1
0
0
0
0
0
2
0
?
?
12
?
4
78
0
0
31
0
203
109
14:40:00 5400
5400
5400
5400
5400
5400
5400
5400
2-09
2-09
2-09
2-09
2-10
2-10
2-10
2-10
0
1
3
250
0
1
3
250
0
26
0
0
0
33
0
11
0
0
0
2
0
0
0
2
19
66
0
722
18
95
0
698
0
36
9
5
0
161946
9
8
?
0
0
0
?
1
2
0
0
0
0
0
0
31
15
0
0
0
0
0
0
1
0
0
0
0
0
0
0
2
0
0
?
1
0
18
?
4
0
18
0
46
0
46
0
203
0
109
Function
Macro...
Reports the logical device traffic by...
ResVdskByNode
a physical node.
ResVdskOneNode
for a specified node.
ResVdskByGroup
a node grouping.

EXEC ResVdskByNode
EXEC ResVdskOneNode
EXEC ResVdskByGroup
Usage Notes
To use the macros, existing rows must exist in the ResUsageSvdsk table.
223

Output Examples
of ResVdiskByNode, which has the NodeID column, and ResVdskByGroup, which has the
NodeType column).
Column...
Reports the...
Read Cnt / Sec
average number of logical device reads per second.
Write Cnt / Sec
average number of logical device writes per second.
Rd KB / I/O
average number of KB per logical device read.
Wrt KB / I/O
average number of KB per logical device write.
Avg I/O Resp
average response time for a logical device read or write in seconds.
Out Rqst Time %
Max Concur Rqsts
ResVdskByNode Sample Output

Note: The NodeID column only appears in the ResVdskByNode output report.
06/09/26
VDISK TRAFFIC BY NODE
Date
-------06/09/26
Time
-------10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
10:12:15
10:12:30
10:12:45
Node
Id
-----001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
001-01
Page
Read Cnt Write Cnt Rd KB

/ Sec
/ Sec
I/O /
-------- -------- ------0.73
5.20
121.27
0.40
41.17
127.50
0.00
47.40
?
2.43
45.17
111.79
2.77
45.90
12.00
2.67
43.77
12.00
2.83
46.10
12.00
4.87
14.13 1374.25
5.37
1.77 1785.00
5.43
0.30 1785.00
3.20
9.70 1759.00
0.50
12.60
275.47
0.00
2.17
?
Avg
Wrt KB
I/O
I/O
Resp
------- ------46.69
0.023
116.34
0.152
118.98
0.143
113.54
0.126
109.93
0.118
112.07
0.119
112.32
0.107
100.01
0.081
38.15
0.065
71.56
0.064
16.76
0.039
18.84
0.038
32.94
0.017
Out
Rqst
Time %
-----9.7%
85.3%
98.8%
98.4%
97.7%
98.0%
97.3%
54.9%
41.0%
36.2%
45.8%
38.6%
3.1%
ResVdskOneNode Sample Output

06/09/26
VDISK Traffic for Node 001-01
Date
-------06/09/26
224
Time
-------10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
ReadCnt
/ Sec
--------0.73
0.40
0.00
2.43
2.77
2.67
2.83
4.87
5.37
5.43
Page
WriteCnt
Rd KB
Wrt KB
/ Sec
/ I/O
/ I/O
-------- ------- ------5.20
121.27
46.69
41.17
127.50
116.34
47.40
?
118.98
45.17
111.79
113.54
45.90
12.00
109.93
43.77
12.00
112.07
46.10
12.00
112.32
14.13 1374.25
100.01
1.77 1785.00
38.15
0.30 1785.00
71.56
Avg
I/O
Resp
------0.023
0.152
0.143
0.126
0.118
0.119
0.107
0.081
0.065
0.064
Out
Rqst
Time %
-----9.7%
85.3%
98.8%
98.4%
97.7%
98.0%
97.3%
54.9%
41.0%
36.2%

10:12:15
10:12:30
10:12:45
3.20
0.50
0.00
9.70
12.60
2.17
1759.00
275.47
?
16.76
18.84
32.94
0.039
0.038
0.017
45.8%
38.6%
3.1%
ResVdskByGroup Sample Output

Note: The NodeType column only appears in the ResVdskByGroup output report.
06/09/26
VDISK TRAFFIC BY GROUP
Date
-------06/09/26
Node
Type Time
---- -------4400 10:09:45
10:10:00
10:10:15
10:10:30
10:10:45
10:11:00
10:11:15
10:11:30
10:11:45
10:12:00
10:12:15
10:12:30
10:12:45
ReadCnt
/ Sec
-------0.73
0.40
0.00
2.43
2.77
2.67
2.83
4.87
5.37
5.43
3.20
0.50
0.00
WriteCnt Rd KB
/ Sec
/ I/O
-------- ------5.20
121.27
41.17
127.50
47.40
?
45.17
111.79
45.90
12.00
43.77
12.00
46.10
12.00
14.13 1374.25
1.77 1785.00
0.30 1785.00
9.70 1759.00
12.60
275.47
2.17
?
Wrt KB
/ I/O
------46.69
116.34
118.98
113.54
109.93
112.07
112.32
100.01
38.15
71.56
16.76
18.84
32.94
Page
Avg
I/O
Resp
------0.023
0.152
0.143
0.126
0.118
0.119
0.107
0.081
0.065
0.064
0.039
0.038
0.017
Max
Concur
Rqsts
-----4.5
20.5
20.0
17.5
13.5
14.5
14.0
12.5
3.0
2.0
2.0
3.5
2.0
Out
Rqst
Time %
-----9.7%
85.3%
98.8%
98.4%
97.7%
98.0%
97.3%
54.9%
41.0%
36.2%
45.8%
38.6%
3.1%
225

226
APPENDIX A
How to Read Syntax Diagrams
This appendix describes the conventions that apply to reading the syntax diagrams used in
this book.
Syntax Diagram Conventions

Notation Conventions
Item
Definition / Comments
Letter
An uppercase or lowercase alphabetic character ranging from A through Z.
Number
A digit ranging from 0 through 9.

Do not use commas when typing a number with more than 3 digits.
Word
Keywords and variables.

UPPERCASE LETTERS represent a keyword.
Syntax diagrams show all keywords in uppercase, unless operating system
restrictions require them to be in lowercase.
lowercase letters represent a keyword that you must type in lowercase, such as a
Linux command.
Mixed Case letters represent exceptions to uppercase and lowercase rules. The
exceptions are noted in the syntax explanation.
lowercase italic letters represent a variable such as a column or table name.
Substitute the variable with a proper value.
lowercase bold letters represent an excerpt from the diagram. The excerpt is
defined immediately following the diagram that contains it.
UNDERLINED LETTERS represent the default value.
This applies to both uppercase and lowercase words.
Spaces
Use one space between items such as keywords or variables.
Punctuation
Type all punctuation exactly as it appears in the diagram.
Paths
The main path along the syntax diagram begins at the left with a keyword, and proceeds, left
to right, to the vertical bar, which marks the end of the diagram. Paths that do not have an
arrow or a vertical bar only show portions of the syntax.
The only part of a path that reads from right to left is a loop.
227
Appendix A: How to Read Syntax Diagrams

Continuation Links
Paths that are too long for one line use continuation links. Continuation links are circled
letters indicating the beginning and end of a link:
A
FE0CA002
When you see a circled letter in a syntax diagram, go to the corresponding circled letter and
continue reading.
Required Entries
Required entries appear on the main path:
SHOW
FE0CA003
If you can choose from more than one entry, the choices appear vertically, in a stack. The first
entry appears on the main path:
SHOW
CONTROLS
VERSIONS
FE0CA005
Optional Entries
You may choose to include or disregard optional entries. Optional entries appear below the
main path:
SHOW
CONTROLS
228
FE0CA004

If you can optionally choose from more than one entry, all the choices appear below the main
path:
READ
SHARE
JC01A010
ACCESS
Some commands and statements treat one of the optional choices as a default value. This
value is UNDERLINED. It is presumed to be selected if you type the command or statement
without specifying one of the options.
Strings
String literals appear in apostrophes:
'msgtext '
JC01A004
Abbreviations
If a keyword or a reserved word has a valid abbreviation, the unabbreviated form always
appears on the main path. The shortest valid abbreviation appears beneath.
SHOW
CONTROLS
CONTROL
FE0CA042
In the above syntax, the following formats are valid:
SHOW CONTROLS
SHOW CONTROL
Loops
A loop is an entry or a group of entries that you can repeat one or more times. Syntax
diagrams show loops as a return path above the main path, over the item or items that you can
repeat:
,
,
(
cname
3
4
)
JC01B012
229

Read loops from right to left.

The following conventions apply to loops:
IF...
THEN...
there is a maximum number of

entries allowed
the number appears in a circle on the return path.
there is a minimum number of

entries required
the number appears in a square on the return path.
a separator character is required

between entries
the character appears on the return path.
In the example, you may type cname a maximum of 4 times.
In the example, you must type at least three groups of column

names.
If the diagram does not show a separator character, use one

blank space.
In the example, the separator character is a comma.
a delimiter character is required

around entries
the beginning and end characters appear outside the return

path.
Generally, a space is not needed between delimiter characters
and entries.
In the example, the delimiter characters are the left and right
parentheses.
Excerpts
Sometimes a piece of a syntax phrase is too large to fit into the diagram. Such a phrase is
indicated by a break in the path, marked by (|) terminators on each side of the break. The
name for the excerpted piece appears between the terminators in boldface type.
The boldface excerpt name and the excerpted phrase appears immediately after the main
diagram. The excerpted phrase starts and ends with a plain horizontal line:
LOCKING
excerpt
A
HAVING
con
excerpt
where_cond
,
cname
,
col_pos
JC01A014
230

Multiple Legitimate Phrases

In a syntax diagram, it is possible for any number of phrases to be legitimate:
dbname
DATABASE
tname
TABLE
vname
VIEW
JC01A016
In this example, any of the following phrases are legitimate:
dbname
DATABASE dbname
tname
TABLE tname
vname
VIEW vname
Sample Syntax Diagram

,
CREATE VIEW
viewname
AS
A
LOCKING
cname
CV
LOCK
ACCESS
dbname
A
DATABASE
tname
FOR
SHARE
IN
READ
TABLE
WRITE
EXCLUSIVE
vname
VIEW
EXCL
,
SEL
B
MODE
expr
,
FROM
qual_cond
tname
.aname
C
HAVING cond
;
qual_cond
,
WHERE cond
GROUP BY
cname
,
col_pos
JC01A018
231

Diagram Identifier
The alphanumeric string that appears in the lower right corner of every diagram is an internal
identifier used to catalog the diagram. The text never refers to this string.
232
APPENDIX B
ResUsageIpma Table
The ResUsageIpma table includes resource usage data for system-wide, node information.
Note: The ResUsageIpma table is generally not used at customer sites and is only used by
Teradata engineers.
This table is created as a MULTISET table. For more information, see Relational Primary
Index on page 37.
Note: Summary Mode is not applicable to this table.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT

columns.
NodeID
n/a

denotes the three-digit cabinet number and MM denotes the twodigit chassis number of the node. For example, a node in chassis 9
of cabinet 3 has a node ID of '003-09'.
INTEGER


These columns provide a generalized picture of the vprocs running on this node, shown as
Type n virtual processors where n = 1 to 7. Under the current implementation, only Type 1
(AMP), Type 2 (PE), Type 3 (GTW), Type 4 (RSG), and Type 5 (TVS) vprocs exist; vproc
types 6 through 7 are not currently used.
233
Appendix B: ResUsageIpma Table

Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
NodeType
n/a
CHAR(8)
NCPUs
n/a
SMALLINT

Vproc1
n/a
Current count of type 1 (AMP) virtual processors running under

the node.
SMALLINT
Vproc2
n/a
Current count of type 2 (PE) virtual processors running under the

node.
SMALLINT
Vproc3
n/a
Current count of type 3 (GTW) virtual processors running under

the node.
SMALLINT
Vproc4
n/a
Current count of type 4 (RSG) virtual processors running under

the node.
SMALLINT
Vproc5
n/a
Current count of type 5 (TVS) virtual processors running under the

node.
SMALLINT
Vproc6
n/a
SMALLINT

Vproc7
n/a
SMALLINT

VprocType1
n/a
Type of virtual processor for Vproc1. Value is always AMP.
CHAR(4)
VprocType2
n/a
Type of virtual processor for Vproc2. Value is always PE.
CHAR(4)
VprocType3
n/a
Type of virtual processor for Vproc3. Value is always GTW.
CHAR(4)
VprocType4
n/a
Type of virtual processor for Vproc4. Value is always RSG.
CHAR(4)
VprocType5
n/a
Type of virtual processor for Vproc5. The value is always TVS (see
Teradata Virtual Storage).
CHAR(4)
VprocType6
n/a
The type of virtual processor for Vproc6.
CHAR(4)

VprocType7
n/a
The type of virtual processor for Vproc7.
CHAR(4)

MemSize
234
n/a
Amount of memory on this node in megabytes. Useful for

performing memory usage calculations.
INTEGER

Statistics Columns
Column Name
Mode
Description
Data Type
NodeNormFactor
n/a
A per CPU normalization factor that is used to normalize the

reported CPU values of the ResUsageSpma table to a single 5100
CPU.
INTEGER
This value is scaled by a factor of 100. For example, if the actual

factor is 5.25, the value of the NodeNormFactor will be 525.
Note: The normalization factor is related to the NodeType value
reported in the ResUsageSpma table.
For information on this value, see Chapter 6: ResUsageSpma
Table.
Secs
n/a

cases:
SMALLINT

CentiSecs
n/a

INTEGER
NominalSecs
n/a
SMALLINT
Active
max

FLOAT

usage tables.
TheTimestamp
n/a
BIGINT
CODFactor
n/a

SMALLINT
Statistics Columns
Process Scheduling CPU Switching Columns
These columns identify the number of times CPUs were switched by the scheduler from one
type of work to another type of work.
235

Statistics Columns
Column Name
Mode
Description
Data Type
CPUProcSwitches
count
Number of times the scheduler switched a currently active process

of a CPU to a new process.
FLOAT
Net Columns
Message Type Columns
These columns subdivide all messages sent and received into the type of message, where:
Hash messages (Hash) are data sent to a destination through its primary or fallback hash
value
Processor messages (Proc) are data sent to a destination through a vproc ID
Group messages (Group) are broadcasted messages to be received by members of a group
Local messages (Local) are messages communicated locally within the node
Channel messages (Chan) are data sent between vprocs through channel IDs for purposes
of a private conversation to perform functions such as row redistribution, and so on
Mailbox messages (Mbox) are data sent between vprocs through mailbox IDs for similar
purposes as channel messages.
A duplicated accounting is done with two different perspectives, since Hash + Proc + Group +
Local messages = Chan + MBox messages.
Column Name
Mode
Description
Data Type
MsgChanReads
count
Number of channel messages read by this node.
FLOAT
MsgChanWrites
count
Number of channel messages written by this node.
FLOAT
MsgHashReads
count
Number of hash messages read by this node.
FLOAT
MsgHashWrites
count
Number of hash messages written by this node.
FLOAT
MsgGroupReads
count
Number of group messages read by this node.
FLOAT
MsgGroupWrites
count
Number of group messages written by this node.
FLOAT
MsgLocalReads
count
Number of local messages read by this node.
FLOAT
MsgLocalWrites
count
Number of local messages written by this node.
FLOAT
MsgMboxReads
count
Number of mailbox messages read by this node.
FLOAT
MsgMboxWrites
count
Number of mailbox messages written by this node.
FLOAT
MsgProcReads
count
Number of processor messages read by this node.
FLOAT
MsgProcWrites
count
Number of processor messages written by the node.
FLOAT
236

Statistics Columns
Message Delivery Times Columns

These columns identify the time it took for hash, processor, group and local messages to reach
their destination. Two times are provided:
Message transmission to mailbox delivery (MDelivery)
Mailbox delivery to process delivery (PDelivery)
Column Name
Mode
Description
Data Type
MsgHashMDelivery
count
Total amount of time read hash messages took for mailbox delivery.
FLOAT

MsgHashPDelivery
count
Total amount of time read hash messages took for process delivery.
FLOAT

MsgGroupMDelivery
count
Total amount of time read group messages took for mailbox

delivery.
FLOAT

MsgGroupPDelivery
count
Total amount of time read group messages took for process

delivery.
FLOAT

MsgLocalMDelivery
count
Total amount of time read local messages took for mailbox delivery.
FLOAT

MsgLocalPDelivery
count
Total amount of time read local messages took for process delivery.
FLOAT

MsgProcMDelivery
count
Total amount of time read processor messages took for mailbox

delivery.
FLOAT

MsgProcPDelivery
count
Total amount of time read processor messages took for process

delivery.
FLOAT
Net Circuit Management Columns

For names and descriptions of these columns, see Net Circuit Management Columns in
Chapter 6 ResUsageSpma Table on page 49.
Per Bynet Network Transport Data Columns
For the names and descriptions of these columns, see Per-Bynet Network Transport Data
Columns of Chapter 6 ResUsageSpma Table on page 49.
Net Miscellaneous Contention Management Columns
These columns identify some additional contention management not addressed in the other
contention management areas.
237

Statistics Columns
Column Name
Mode
Description
Data Type
NetActiveMrg
count
The number of concurrent active merges on all Bynets.
FLOAT
NetBrdWindowOverrun
count
Broadcast window overruns on all Bynets.
FLOAT
Note: This column is net-specific, that is, it relates to each specific

Bynet. On a single-node system, net-specific statistics are not
meaningful and are always zero
NetCBRErrror
count
Note: This column is not valid and returns a zero value.
FLOAT
NetDanglingAborts
count
FLOAT
NetDanglingCommits
count
FLOAT
NetMrgBlock
count
Number of times a merge message was blocked until delivery of

outstanding outgoing messages.
FLOAT
NetMrgBufWaits
count
Number of times an IOP task encountered an empty row-block

buffer on all Bynets.
FLOAT
NetMsgChannelBlock
count
Number of times the net software was blocked because the channel
was not in RxReady state on the receiver.
FLOAT
NetMrgHeapFails
count
Number of times a merge operation heap space request failed.
FLOAT
NetMrgHeapRequests
count
Number of times a merge operation requested heap space.
FLOAT
NetMsgFCBlock
count
Number of times the net software was blocked because the receiver
was flow controlled.
FLOAT
NetMsgFCSleep
count
Number of times a transmitter process was put to sleep because it

was flow controlled.
FLOAT
NetMsgGroupBlock
count
could not implicitly enter the group.
FLOAT
NetMsgResourceBlock
count
could not get the necessary resources.
FLOAT
NetMsgRxBlock
count
Number of times the net software could not accept a message and
caused a transmitter to block.
FLOAT
Net Queues Columns

These columns identify lengths of the various internal queues used by the network controllers.
238
NetSamples can be used to normalize all aggregated sampled statistics to an average

queue-length basis.
Example: Dividing (NetPtPQueue/NetSamples) yields the average point-to-point queue

length over all samples on all networks taken during this log interval.
All of the aggregated sampled statistics columns in the following table are net-specific, that
is, they relate to each specific Bynet. On a single-node system, net-specific statistics are not
meaningful and are always zero.

Statistics Columns
Column Name
Mode
Description
Data Type
NetPtPQueue
count
Aggregated sample point-to-point normal priority queue length on

all Bynets.
FLOAT
NetBlockQueue
track
Total number of services on the BlockableService queue at the

current time.
FLOAT
Services can be blocked for a variety of reasons including receiver

flow control, receiver resource usage, and daemon services.
NetBlockQueueMax
max
Maximum number of services on the BlockableServices queue in

this log interval.
FLOAT
NetBlockQueueTotal
count
Total number of services added to the BlockableServices queue in

this log interval.
FLOAT
NetBrdQueue
count
Aggregated sample broadcast normal priority queue length on all

Bynets.
FLOAT
NetPtPQueueMax
max
Maximum reported value of NetPtPQueue in this reporting

interval.
FLOAT
NetBrdQueueMax
max
Maximum reported value of NetBrdQueue in this reporting

interval.
FLOAT
NetHPBrdQueue
count
Aggregated sample broadcast high priority queue length on all

networks.
FLOAT
NetHPBrdQueueMax
max
Maximum value of NetHPBrdQueue in this reporting interval.
FLOAT
NetPendMrgQueue
count
Current count of pending merges, regardless of which net. A merge

may be queued for reasons such as:
FLOAT
The local IOP memory is saturated

System memory is trashing.
NetHPPtPQueue
count
Aggregated sample point-to point high priority queue length on all

Bynets.
FLOAT
NetHPPtPQueueMax
max
Maximum value of NetHPPtPQueue in this reporting interval.
FLOAT
Reserved Columns
These reserved columns expand to values 00 - 19, so that column names would be
RSSInternal00, RSSInternal01, and so on.
Column Name
Mode
Description
Data Type
RSSInternal[00-06]
track
Reserved for use by the RSS application.
FLOAT
RSSInternal[07-08, 10-11, 13-14, 16-19]
count
FLOAT
RSSInternal[09, 12, 15]
max
FLOAT
239

Spare Columns
Spare Columns
The ResUsageIpma table spare fields are named Spare00 through Spare09, and SpareInt.
Column Name
Description
Spare00

Teradata Database 15.0.
1000 on SLES 10.
Related Topics
240
APPENDIX C
ResUsageIvpr Table
The ResUsageIvpr table includes resource usage data for system-wide information.
Note: The ResUsageIvpr table is generally not used at customer sites and is only used by
Teradata engineers.
These columns taken together form the primary index.
Column Name
Mode
Description
Data Type
TheDate
n/a
DATE
TheTime
n/a
FLOAT

columns.
NodeID
n/a

INTEGER


Column Name
Mode
Description
Data Type
GmtTime
n/a

FLOAT
NodeType
n/a
CHAR(8)
VprId
n/a
Identifies the vproc number (non-Summary Mode) or the vproc

type (Summary Mode; 0 = NODE, 1 = AMP, 2 = PE, 3=GTW,
4=RSG, 5=TVS).
INTEGER
241
Appendix C: ResUsageIvpr Table

Column Name
Mode
Description
Data Type
VprType
n/a
The values can be NODE, AMP, PE, GTW, RSG, or TVS (see
Teradata Virtual Storage).
CHAR(4)
Secs
n/a

cases:
SMALLINT

CentiSecs
n/a

INTEGER
NominalSecs
n/a
SMALLINT
SummaryFlag
n/a
is a non-summary row, and 'S' if the row is a summary row.
CHAR (1)
NCPUs
n/a
SMALLINT

Active
max

FLOAT

usage tables.
TheTimestamp
n/a
BIGINT
CODFactor
n/a

SMALLINT
242

Statistics Columns
Statistics Columns
File System Columns
Cylinder Defragmentation Overhead Columns
These columns identify background file system overhead associated with fragmented free
space to achieve one large free space within that cylinder (CylDefrag).
Event counts are found in ResUsageSvpr.
Each cylinder defragment event implies one logical cylinder index read and one logical
cylinder index write.
Only logical I/Os and the amount moved (KB) are identified. Cylinder defragments are
done on cylinders containing permanent tables (including append and transient journal
tables) only.
Column Name
Mode
Description
Data Type
FileDbCylDefragKB
count
KB moved by the FileDbCylDefragIO.
FLOAT
FileDbCylDefragIO
count
Number of permanent data block logical I/Os due to cylinder

defragmentation.
FLOAT
Cylinder MiniCylPack Overhead Columns

These columns identify file system overhead associated with MiniCylPack operations that get
performed to make available a free cylinder when one is needed but not available.
Event counts are found in ResUsageSvpr.
Only logical I/Os and the amount moved (KB) are identified, except the amount moved
for cylinder indexes because they can be calculated by multiplying the current cylinder
index fixed size and the I/Os.
MiniCylPack operations are done on cylinders containing permanent tables (including

append and transient journal tables) only.
Column Name
Mode
Description
Data Type
FilePCiMCylPackIO
count
Number of permanent cylinder index logical I/Os due to

performing MiniCylPack operations.
FLOAT
FilePDbMCylPackKB
count
KB moved by FilePDbCylPackIOs.
FLOAT
FilePDbMCylPackIO
count
Number of permanent data block logical I/Os due to performing

MiniCylPack operations.
FLOAT
Cylinder Split and Migrate Overhead Columns

These columns further identify file system cylinder split/migrate (CylMigr) overhead
performed when cylinders can not accommodate new data.
243

Statistics Columns
Note: Event counts are found in Chapter 13: ResUsageSvpr Table..

Only logical I/Os and the amount moved (KB) for data blocks are identified. Each cylinder
migration event implies one logical read and three logical writes of the cylinder index. Only
permanent tables (including append and transient journal tables) are migrated.
Column Name
Mode
Description
Data Type
FileDbCylMigrKB
count
KB moved by FileDbCylMigrIOs.
FLOAT
FileDbCylMigrIO
count
Number of data block logical I/Os due to cylinder migration.
FLOAT
Data Block Creation Columns

These columns identify the file system operations required when a data block is being created
(BlkCreate).
Note: These columns do not include data blocks created due to any of the new data blocks
created when a data block was updated as described in the Data Block Update Operations
Columns description.
Column Name
Mode
Description
Data Type
FilePDbCreateKB
count
KB created by FilePDbCreates.
FLOAT
FilePDbCreates
count
Number of permanent table (including append and transient

journal tables) data blocks created.
FLOAT
FileSDbCreateKB
count
KB created by FileSDbCreates.
FLOAT
FileSDbCreates
count
Number of spool data blocks created.
FLOAT
Data Block Update Operations Columns

These columns identify the file system operations required when a data block is being updated
(BlkUpd). When a block is updated, it can be in place and requires no new data blocks, or it
could spill over the current data block and require one, two, three or more new data blocks in
addition to the current data block. Only logical I/Os and the amount moved (KB) are
identified, except for the amount moved for cylinder indexes because they can be calculated by
multiplying the current fixed cylinder index size by the I/Os. Data block updates should only
be performed on permanent tables (including append and transient journal tables), so no
attempt is made to separate permanent and spool data segments.
Column Name
Mode
Description
Data Type
FileCiUpd0IO
count
Number of cylinder index logical I/Os performed for a block

update operation requiring no new data blocks.
FLOAT
FileCiUpd1IO
count

update operation requiring 1 new data blocks.
FLOAT
244

Statistics Columns
Column Name
Mode
Description
Data Type
FileCiUpd2IO
count

update operation requiring two new data blocks.
FLOAT
FileCiUpd3IO
count

update operation requiring three new data blocks.
FLOAT
FileCiUpdNIO
count

update operation requiring over three new data blocks.
FLOAT
FileDbUpd0IO
count
Number of data block logical I/Os performed for a block update

operation requiring no new data blocks.
FLOAT
FileDbUpd1IO
count

operation requiring one new data blocks.
FLOAT
FileDbUpd2IO
count

operation requiring two new data blocks.
FLOAT
FileDbUpd3IO
count

operation requiring three new data blocks.
FLOAT
FileDbUpd0KB
count
KB moved by FileDbUpd0IO.
FLOAT
FileDbUpd1KB
count
FLOAT
FileDbUpd2KB
count
FLOAT
FileDbUpd3KB
count
FLOAT
FileDbUpdNKB
count
KB moved by FileDbUpdNIO.
FLOAT
FileDbUpdNIO
count

operation requiring over three new data blocks.
FLOAT
Multi-Row Requests Columns

These columns identify the significant multi-row requests made by application software on
the file system. Rows are distinguished as permanent data (P), spool (S) or user append table /
permanent journal table (APt).
Column Name
Mode
Description
Data Type
FileAPtBlkRead
count
Number of requests for an append data block read.
FLOAT
FileAPtBlkReplace
count
Number of requests for an append data block replace.
FLOAT
FileAPtRownins
count
Number of requests for an append data multi-row insert.
FLOAT
FileAPtRowNDel
count
Number of requests for an append data multi-row delete.
FLOAT
FileAPtRowNUpd
count
Number of requests for an append data multi-row update.
FLOAT
FileAPtSortable
count
Number of requests for an append table sort.
FLOAT
FileAPtTabdelete
count
Number of requests for an append table delete.
FLOAT
245

Statistics Columns
Column Name
Mode
Description
Data Type
FileAPtTabdelra
count
Number of requests for an append multi-row delete.
FLOAT
FileAPtTabmrows
count
Number of requests for an append table modification.
FLOAT
FileAPtTabrblocks
count
Number of requests for an append table multi-block read.
FLOAT
FilePBlkRead
count
Number of requests for a permanent data block read.
FLOAT
FilePBlkReplace
count
Number of requests for a permanent data block replace.
FLOAT
FilePRowNDel
count
Number of requests for a permanent data multi-row delete.
FLOAT
FilePRownins
count
Number of requests for a permanent data multi-row insert.
FLOAT
FilePRowNUpd
count
Number of requests for a permanent data multi-row update.
FLOAT
FilePSortable
count
Number of requests for permanent table sort.
FLOAT
FilePTabdelete
count
Number of requests for a permanent table delete.
FLOAT
FilePTabdelra
count
Number of requests for a multi-row delete.
FLOAT
FilePTabmrows
count
Number of requests for a permanent table modification.
FLOAT
FilePTabrblocks
count
Number of requests for a permanent table multi-block read.
FLOAT
FileSBlkRead
count
Number of requests for a spool data block read.
FLOAT
FileSBlkReplace
count
Number of requests for a spool data block replace.
FLOAT
FileSRowNDel
count
Number of requests for a spool data multi-row delete.
FLOAT
FileSRownins
count
Number of requests for a spool data multi-row insert.
FLOAT
FileSRowNUpd
count
Number of requests for a spool data multi-row update.
FLOAT
FileSSortable
count
Number of requests for spool table sort.
FLOAT
FileSTabdelete
count
Number of requests for a spool table delete.
FLOAT
FileSTabdelra
count
Number of requests for a spool multi-row delete.
FLOAT
FileSTabmrows
count
Number of requests for a spool table modification.
FLOAT
FileSTabrblocks
count
Number of requests for a spool table multi-block read.
FLOAT
Single-Row Requests Columns

These columns identify the significant single-row requests made by application software on
the file system. Rows are distinguished as permanent data (P), spool (S) or user append table /
permanent journal table (APt).
Column Name
Mode
Description
Data Type
FileAPtRowAppend
count
Number of requests for an append row append.
FLOAT
FileAPtRowDelete
count
Number of requests for an append row delete.
FLOAT
246

Statistics Columns
Column Name
Mode
Description
Data Type
FileAPtRowInsert
count
Number of requests for an append row insert.
FLOAT
FileAPtRowReadInit
count
Number of requests for an initial append row read.
FLOAT
FileAPtRowReadCont
count
Number of requests for an continued append row read.
FLOAT
FileAPtRowReplace
count
Number of requests for an append row replace/update.
FLOAT
FilePRowAppend
count
Number of requests for a row append.
FLOAT
FilePRowDelete
count
Number of requests for a permanent row delete.
FLOAT
FilePRowInsert
count
Number of requests for a permanent row insert.
FLOAT
FilePRowReadCont
count
Number of requests for a continued permanent row read.
FLOAT
FilePRowReadInit
count
Number of requests for an initial permanent row read.
FLOAT
FilePRowReplace
count
Number of requests for a permanent row replace.
FLOAT
FileSRowAppend
count
Number of requests for a row append.
FLOAT
FileSRowDelete
count
Number of requests for a spool row delete.
FLOAT
FileSRowInsert
count
Number of requests for a spool row insert.
FLOAT
FileSRowReadCont
count
Number of requests for a continued spool row read.
FLOAT
FileSRowReadInit
count
Number of requests for an initial spool row read.
FLOAT
FileSRowReplace
count
Number of requests for a spool row replace/update.
FLOAT
Transient Journal Overhead Columns

These columns identify file system overhead associated with maintaining a transient journal
(TJ).
Column Name
Mode
Description
Data Type
FileTJBufUpdates
count
Number of transient journal buffer updates.
FLOAT
Transient Journal Requests Column

These columns identify the significant transient journal requests made by application software
on the file system.
Column Name
Mode
Description
Data Type
FileTJCalls
count
Number of transient journal calls.
FLOAT
FileTJDbUpdates
count
Number of WAL data blocks modified. The modification can either

be an update or a delete of an existing WAL or TJ record.
FLOAT
247

Statistics Columns
Write Ahead Logging

These columns identify the log-based file system recovery scheme in which modifications to
permanent data are written to a log file, the WAL log.
Column Name
Mode
Description
Data Type
FileTJAppends
count
Number of times transient journal records were appended to the

WAL log. A single append call can append multiple transient
journal rows. A transient journal append by itself does not imply a
write of a WAL block, nor a WAL Cylinder Index (WCI)
modification.
FLOAT
FileTJFlush
count
Number of times a request to force transient journal records within

the WAL log to be written to disk is issued. An increment of this
counter may or may not result in an I/O depending on whether the
request was to flush records that were already on disk.
FLOAT
FileWAppends
count
Number of times a record was appended to the WAL log. A single

append call can append multiple rows. Subtracting FileTJAppends
from this counter results in the number of times non-transient
journal rows were appended to the WAL log.
FLOAT
FileWDBCreates
count
Number of WAL data blocks created. The block can contain either
TJ records, WAL records or both
FLOAT
FileWFlush
count
Number of times a request to force any record in the WAL log to be

written to disk is issued. An increment of this counter may or may
not result in an I/O depending on whether the request was to flush
records that were already on disk. Subtracting FileTJFlush from this
counter results in the number of times a non-transient journal
WAL flush was issued.
FLOAT
FileWRowDelete
count
Number of times rows were deleted from the WAL log.
FLOAT
FileWTabDelRa
count
Number of requests for a WAL multi-row delete.
FLOAT
General Concurrency Control Monitor Management Columns

These columns identify monitor activities for Teradata Database concurrency control.
Column Name
Mode
Description
Data Type
MonAllocates
count
Number of monitors allocated.
FLOAT
MonBlocks
count
Number of times entry into a monitor was blocked. For example,

the number of requests - the number of blocks = immediate grants.
FLOAT
MonEnters
count
Number of times entry into a monitor was requested.
FLOAT
MonYields
count
Number of times a monitor yield was requested.
FLOAT
248

Statistics Columns
Net Columns
Message Type Columns
These columns subdivide all messages sent and received into the type of message, where:
Hash messages (Hash) are data sent to a destination through its primary or fallback hash
value.
Processor messages (Proc) are data sent to a destination through a vproc ID.
Group messages (Group) are broadcasted messages to be received by members of a group.
local messages (Local) are messages communicated locally within the node.
Channel messages (Chan) are data sent between vprocs through channel IDs for purposes
of a private conversation to perform functions such as row redistribution, and so on.
Mailbox messages (Mbox) are data sent between vprocs through mailbox IDs for similar
purposes as channel messages.
A duplicated accounting is done with two different perspectives, since Hash + Proc + Group +
Local messages = Chan + MBox messages.
Column Name
Mode
Description
Data Type
MsgChanReads
count
Number of channel messages read by this vproc.
FLOAT
MsgChanWrites
count
Number of channel messages written by this vproc.
FLOAT
MsgHashReads
count
Number of hash messages read by this vproc.
FLOAT
MsgHashWrites
count
Number of hash messages written by this vproc.
FLOAT
MsgGroupReads
count
Number of group messages read by this vproc.
FLOAT
MsgGroupWrites
count
Number of group messages written by this vproc.
FLOAT
MsgLocalReads
count
Number of local messages read by this vproc.
FLOAT
MsgLocalWrites
count
Number of local messages written by this vproc.
FLOAT
MsgProcReads
count
Number of processor messages read by this vproc.
FLOAT
MsgProcWrites
count
Number of processor messages written by this vproc.
FLOAT
MsgMboxReads
count
Number of mailbox messages read by this vproc.
FLOAT
MsgMboxWrites
count
Number of mailbox messages written by this vproc.
FLOAT
249

Statistics Columns
Message Delivery Times Columns

These columns identify the time it took for hash, processor, group and local messages to reach
their destination. Two times are provided:
Message transmission to mailbox delivery (MDelivery).
Mailbox delivery to process delivery (PDelivery).
Column Name
Mode
Description
Data Type
MsgHashMDelivery
count
Total amount of time read hash messages took for mailbox delivery.
FLOAT

MsgHashPDelivery
count
Total amount of time read hash messages took for process delivery.
FLOAT

MsgGroupMDelivery
count
Total amount of time read group messages took for mailbox

delivery.
FLOAT

MsgGroupPDelivery
count
Total amount of time read group messages took for process

delivery.
FLOAT

MsgLocalMDelivery
count
Total amount of time read local messages took for mailbox delivery.
FLOAT

MsgLocalPDelivery
count
Total amount of time read local messages took for process delivery.
FLOAT

MsgProcMDelivery
count
Total amount of time read processor messages took for mailbox

delivery.
FLOAT

MsgProcPDelivery
count
Total amount of time read processor messages took for process

delivery.
FLOAT
Transient Journal Purge Overhead Columns

These columns identify the background overhead associated with the occasional transient
journal purge operation.
Column Name
Mode
Description
Data Type
TJPurges
count
The number of purge passes in which a block-by-block scan is

done.
FLOAT
250

Summary Mode
Column Name
Mode
Description
Data Type
TJDbPurgeDeletes
count
The number of blocks mapped in during the scan that were

included in the ranges of blocks that were deleted.
FLOAT
Before Write Ahead Logging (WAL), the ratio of deletes to reads

would have been a useful measure of the effectiveness of the purge
processing. However, with WAL, the ratio cannot be interpreted
quite so simply because:
1 The range of deleted blocks could include blocks that were not
actually mapped in (and therefore not counted). Blocks that

contain only WAL records are not mapped in during the scan, as
they are automatically filtered out. Under typical conditions,
there are probably relatively few such blocks. TJ and WAL
records are typically generated in an interleaved sequence by
regular SQL transactions. But during periods when the system
workload is dominated by MultiLoad/FastLoad work, there will
be relatively few TJ records written, so the proportion of WALonly blocks would probably be significant.
2 Post-WAL, neither TJDbPurgeReads nor TJDbPurgeDeletes gets
incremented during a normal purge pass. Instead of scanning

the active data blocks, a pointer to the oldest active transaction is
maintained which is a quicker method. Therefore, PurgeTJ() can
simply compute the bounds of the range of records that can be
deleted in the part of the WAL/TJ that precedes the start of the
oldest transaction. This does not require any scanning and the
system cannot definitely determine how many blocks actually
get deleted.
3 If the oldest transaction remains open for a long time, the quick
purge method is not effective. Therefore, the system reverts back

to the full scan method. The TJDbPurgeReads and
TJDbPurgeDeletes are only incremented during a full scan.
TJDbPurgeReads
count
The number of blocks actually mapped in during the purge scan.

This is a reasonable approximate measure of the I/O load. The
system uses full-cylinder read mode, but the block count would still
be roughly proportionate to the I/O load.
FLOAT
Summary Mode
When Summary Mode is active for the ResUsageIvpr table, one row is written to the database
for each type of vproc on each node in the system, summarizing the vprocs of that type on
that node, for each log interval.
that row.
251

Spare Columns
'S'
in Summary Mode.
'N'
normally.
Spare Columns
The ResUsageIvpr table spare fields are named Spare00 through Spare09, and SpareInt.
Column Name
Description
Spare00

Teradata Database 15.0.
1000 on SLES 10.
Related Topics
252
APPENDIX D
ResIpmaView and ResIvprView

Views
This chapter provides the definitions of the ResIpmaView and ResIvprView views.
Note: These views are intended primarily for Teradata engineers.
253
Appendix D: ResIpmaView and ResIvprView Views

ResIpmaView
ResIpmaView
ResIpmaView is based on the ResUsageIpma table.
REPLACE VIEW DBC.ResIpmaView
AS SELECT
TheDate,
TheTime,
GmtTime,
NodeType,
TheTimestamp,
CentiSecs,
Secs,
NominalSecs,
CodFactor,
NCPUs,
Reserved,
Vproc1,
VprocType1,
Vproc2,
VprocType2,
Vproc3,
VprocType3,
Vproc4,
VprocType4,
Vproc5,
VprocType5,
Vproc6,
VprocType6,
Vproc7,
VprocType7,
MemSize,
NodeNormFactor,
/* PM/WM CODs */
CodFactor
AS
/
/* RSS time stamps
RssInternal00 AS
RssInternal01 AS
RssInternal02 AS
RssInternal04 AS
RssInternal05 AS
RssInternal06 AS
RssInternal07 AS
RssInternal08 AS
RssInternal09 AS
RssInternal10 AS
RssInternal11 AS
RssInternal12 AS
RssInternal13 AS
RssInternal14 AS
RssInternal15 AS
PM_CPU_COD,
/* PM CPU Capacity On Demand factor *
in seconds since 1970.

GatherStart,
GatherCpuInfo,
GatherMemInfo,
GatherDiskInfo,
Kcollect,
GatherMmBufs,
OpenMsgCnt,
OpenMsgTotal,
OpenMsgMax,
GatherMsgCnt,
GatherMsgTotal,
GatherMsgMax,
CloseMsgCnt,
CloseMsgTotal,
CloseMsgMax,

/* PM/WM CODs */
Spare00
AS WM_CPU_COD,
*/
/* PM/WM CODs */
/* The default GroupId setting below shows how to
* as well as some PE-only nodes the CASE expression might look like:
* WHEN (VPROCTYPE1='AMP' AND NodeType
* WHEN (VPROCTYPE1='AMP' AND NodeType
* WHEN (VPROCTYPE1='AMP' AND NodeType NOT IN ('5600H', '5650')) THEN
'5550Nodes'
* ELSE 'PEonly'
254

ResIpmaView
*/
CASE
IN ('5600H')) THEN 'AMPNodes'
WHEN (VPROCTYPE1='AMP' AND NodeType NOT IN ('5600H')) THEN 'AMPNodes'
ELSE 'PEonly'
END AS GroupId,
/* table data fields */
FROM DBC.ResUsageIpma WITH CHECK OPTION;
Note: The ResUsageIpma table fields have been removed from this sample output.
255

ResIvprView
ResIvprView
ResIvprView is based on the ResUsageIvpr table.
REPLACE VIEW DBC.ResIvprView
AS SELECT
thedate,
NodeID
thetime,
GmtTime,
NodeType
TheTimestamp,
CentiSecs
Secs
NominalSecs
CodFactor
SummaryFlag
Reserved
NCPUs
vprid,
VprType
(FORMAT '999-99')
AS NodeID,
(FORMAT 'X(8)')
AS NodeType,
(FORMAT
(FORMAT
(FORMAT
(FORMAT
(FORMAT
(FORMAT
(FORMAT
AS
AS
AS
AS
AS
AS
AS
'-------9')
'----9')
'ZZZ9')
'-----9')
'X(1)')
'X(3)')
'ZZ9')
(FORMAT 'X(4)')
CentiSecs,
Secs,
NominalSecs,
CodFactor,
SummaryFlag,
Reserved,
NCPUs,
AS VprType,
/* PM/WM CODs */
CodFactor
AS PM_CPU_COD,

/* PM/WM CODs */
Spare00
AS WM_CPU_COD,
/* PM/WM CODs */
( PM_CPU_COD * WM_CPU_COD / 1000 ) (FORMAT '----9') AS CPU_COD, /* eff
ective CPU COD */
/*
*
*
*

resulting groups. If a coexistence system had 5550, 5600 and 5650 n
odes
* WHEN NodeType
* WHEN NodeType
* ELSE '5550Nodes'
*/
CASE
WHEN NodeType
IN ('5650H') THEN 'A'
ELSE 'A'
END AS GroupId,
FROM DBC.ResUsageIvpr WITH CHECK OPTION;
Note: The ResUsageIvpr table fields have been removed from this sample output.
256
APPENDIX E
Partition Assignments
With regards to Teradata Database, there is more than one definition of partition. The
partitions here refer to the following Parallel Database Extensions (PDE) and vproc definition:
A partition is a collection of tasks and associated resources grouped within a virtual

processor according to the function of the tasks. There are multiple partitions within a
single virtual processor. Partitions are the primary mechanism used by Teradata Database
for managing parallel programs.
Partitions are the subdivision of vproc software processes into 48 semi-isolated domains.
For example, in an AMP vproc, Partition 11 is the AMP Worker Task Partition. In all other
vproc types, Partition 11 is unused.
Another partition description is only meaningful in a dialog between client programs and
Teradata Database. It has nothing to do with PDE vproc partitions, but is a way of enforcing
rules about what a client session is allowed to do and of keeping client sessions isolated from
each other. This concept of partitions is centered in the CLIv2 interface, specifically the
CONNECT parcel.
Partition reservation is as follows:
Partitions 0 through 6 are reserved by PDE
Partitions 7 through 47 are for use by Teradata Database
The table listed under Partition Assignment Listing on page 259 describes the individual
partitions. Teradata Database uses the following vprocs:
Vproc Type
Description
AMP
Access module processors perform database functions, such as executing database

queries. Each AMP owns a portion of the overall database storage.
GTW
Gateway vprocs provide a socket interface to Teradata Database.
Node
The node vproc handles PDE and operating system functions not directly related to
AMP and PE work. Node vprocs cannot be externally manipulated, and do not
appear in the output of the Vproc Manager utility.
PE
Parsing engines perform session control, query parsing, security validation, query
optimization, and query dispatch.
RSG
Relay Services Gateway provides a socket interface for the replication agent, and for
relaying dictionary changes to the Teradata Meta Data Services utility.
TVS
Manages Teradata Database storage. AMPs acquire their portions of database

storage through the TVS vproc.
257
Appendix E: Partition Assignments

Table Conventions
For more information on partition usage, see CPU Utilization Columns on page 151.
Table Conventions
The following table describes the table symbols used in the partition assignments table below.
The symbol used in the Partition
Assignment Listing
Indicates
partition is unused.
activity is observed but not identified.
258
Partition Assignment Listing

The following table lists the Node, AMP, PE, GTW, RSG, and TVS usage of PDE vproc partitions by PDE and Teradata
Database.
Partition:
Usage in Vprocs of Type
259
Number
Name
Node
AMP
PE
GTW
RSG
Kernel
PDE daemons
--------
System Debugger
System Debugger tasks
Console
PDE console
control process
(cnscim)
3-6
Interactive 1 through 4
Console interactive partition programs
Service
Console utilities
scpstart
--------
Host Utility Procedures
--------
utadvtsk
--------
Filesys
--------
File System
processes
--------
10
Gateway
--------
Gateway
processes
--------
11
AWT
--------
12
Session
--------
Session Control
tasks
--------
13
Dispatch
--------
Dispatcher/
Parser Partition
--------
14
[Unused]
--------
15
Startup
--------
16
[Unused]
--------
--------
AMP Worker
Tasks
Startup tasks
--------
--------
TVS
260
Partition:
Usage in Vprocs of Type
Number
Name
Node
AMP
PE
GTW
RSG
TVS
17
RSS Startup
--------
File System RSS

startup
--------
18
Distributed Database
File (DDF) Server
DDF services
--------
19
Metadata
Metadata
Services Gateway
rsgmain
--------
20-29
Interactive Partitions
--------
30
[Unused]
--------
31
Allocator
--------
tvsaallocator
32
Node Agent
--------
tvsa_agent
33
Clique Coordinator
--------
Clique
Coordinator
services
34-41
[Unused]
--------
42
GDO Monitor
gdom
--------
43
Parallel Data Collector
pdcmaster
--------
44
Dump save slaves
csp
--------
45
Dump save or clear
csp
--------
46
Dump list
csp
--------
47
Replication
--------
DBCCONS utilities
Replication
Gateway
rsgdbsmain
Glossary
AG
Allocation Group
AMP
Access Module Processor
API Application Programming Interface

AWS
Administration Workstation
AWT
AMP Worker Task
BLC
Block-level compression
BYNET
Banyan Network (high-speed connection)
COD Capacity on Demand

DBW
Database Window
DDF
Distributed Database File
DDL
Data Definition Language
FSG
File Segment
GTW
Teradata Gateway
I/O Input/Output
MPP Massively Parallel Processing
NOMOD No modification
NUPI
Nonunique Primary Index
PDE Parallel Database Extensions. PDE is a software interface layer between the operating
system and the Teradata Database software. It provides Teradata Database the ability to run in
a parallel environment, execute vprocs, and more.
PE
Parsing Engine
PG
Performance Group
PGWL
PM
Platform Metering
PMPC
PM/API
PP
Performance Group Workload
Performance Monitor and Production Control

Performance Monitor Application Programming Interface
Performance Period
261
Glossary
pWDid
Priority Scheduler Workload Definition ID
RDBMS
Relational Database Management System
ResUsage Resource Usage. The data stored in the database system resource usage tables.
RSG
Relay Services Group
RSS Resource Sampling Subsystem. The RSS provides a method to gather statistics from
across the Teradata Database system, and provides the ability to access the statistics through
an API. Resource usage uses the RSS data from the RSS API to log data to the selected resource
usage tables.
SLG
Service Level Goal
SMP
Symmetric Multi-Processing
TBBLC
Temperature-based Block-level Compression
TCHN
Teradata Channel
Tactical Workload Management Method This workload yields the fastest available response
time and executes at the highest tier, preempting all resource needs of other tiers. This method
is well suited for critical, short-running queries that require fast response times. For more
information, see Teradata Viewpoint User Guide.
TASM
Teradata Active System Management
Timeshare Workload Management Method This workload can be assigned to one of four
stepped access levels, Top, High, Medium, or Low. The higher access levels are given larger
access rates than the lower levels. For example, an SQL request assigned to a Timeshare WD
with a Top access level, which has an access rate of 8, would receive eight times the amount of
resources than an SQL request assigned to a Low access level.
Timeshare workloads are assigned resources remaining after all allocations have been made
for tactical and Workload Share Percent workloads. For more information, see Teradata
Viewpoint User Guide.
VNET
Virtual Network. Virtual BYNET for a single-node.
VP Virtual Partition. A virtual partition divides a system so that a percentage of resources

are allocated to a collection of workloads. A virtual partition can consist of WDs from all
management methods.
TPA
vproc
TVS
Trusted Parallel Application

Virtual Processor
Teradata Virtual Storage
VH Cache Very hot cache. This cache holds the hottest permanent table cylinders. For tables
that are assigned a temperature of very hot, they are kept in the FSG cache as long as they:
262
Stay very hot.
Glossary
Fit into the memory assigned. If the tables cannot fit, the FSG cache considers a sorted list
of the hottest segments and assigns them to the very hot cache in temperature sorted
order. The temperature of the segment is hot enough to qualify.
The temperature takes into account both physical and logical disk accesses and cache hits
(such as, physical and logical I/Os).
WCI
WAL Cylinder Index
Workload Share Percent Management Method This workload is assigned a proportion of

the resources that are available after allocations have been made for tactical workloads. The
percentage of resources is divided equally between all requests running in the WD. For
example, if the Workload Share Percent is 5% and there are five SQL requests, each SQL
request will get 1% of the share resources. For more information, see Teradata Viewpoint User
Guide.
WM
Workload Management
263
Glossary
264
Index
Symbols
?, meaning in macro outputs 182
A
Allocation columns
for ResUsageSpdsk 95
for ResUsageSvdsk 125
for ResUsageSvpr 155
AMP information
macros 183, 188
table 131
view 164
AMP Worker Task columns
for ResUsageSawt 73
for ResUsageSpma 64
API 18
Application programming interfaces. See API
AutoCylPack columns, ResUsageSvpr 143
B
BLC columns, ResUsageSvpr 144
Block-level compression columns. See BLC columns
Broadcast net traffic columns
for ResUsageSpma 57
for ResUsageSps 116
ByGroup macro 30
Bynet network transport data columns, ResUsageIpma 237
C
Channel management columns, ResUsageShst 82
Channel traffic columns
for ResUsageShst 82
for ResUsageSpma 55
Chnsignal status tracking columns, ResUsageSvpr 151
Co-existing node macros. See ByGroup macros
CPU use by AMPs macros
normalized viewing 191
ResAmpCpuByGroup 188
ResCPUByAMP 188
ResCPUByAMPOneNode 188
CPU use by each node macros
ResCPUByGroup 195
ResCPUByNode 195
ResCPUOneNode 195
CPU use by PEs macros

normalized viewing 194
ResCPUByPE 192
ResCPUByPEOneNode 192
ResPeCpuByGroup 192
CPU utilization columns
for ResUsageScpu 46
for ResUsageSpma 63
for ResUsageSps 109
Cylinder defragmentation overhead columns, ResUsageIvpr
243
Cylinder management overhead columns, ResUsageSvpr 141
Cylinder read columns, ResUsageSvpr 151
D
Data block merge columns, ResUsageSvpr 143
Data block prefetches columns
for ResUsageSpma 53
for ResUsageSps 113, 135
Data segment lock requests columns
for ResUsageSpma 54
Database commands
SET LOGTABLE 28
SET RESOURCE 27
SET SUMLOGTABLE 28
Database Window Supervisor
setting logging rates 28
Database Window. See DBW
DBW 27
enabling RSS logging
Depot columns, ResUsageSpma 54
DISABLE LOGONS
effects on logging 34
E
Example
executing a ResCPUByAmp macro 33
ResAmpCpuByGroup macro report 190
ResAWT macro report 186
ResAWTByAMP macro report 186
ResAWTByNode macro report 187
ResCPUByAMP macro report 189
ResCPUByGroup macro report 197
ResCPUByNode macro report 196
265
Index
ResCPUByPE macro report 193

ResCPUByPEOneNode macro report 193
ResCPUOneNode macro report 196
ResHostByGroup macro report 200
ResHostByLink macro report 199
ResHostOneNode macro report 200
ResLdvByGroup macro report 202, 219
ResLdvByNode macro report 201, 218, 219
ResLdvOneNode macro report 202
ResMemByGroup macro report 207
ResMemMgmtByNode macro report 206
ResMemMgmtOneNode macro report 206
ResNetByGroup macro report 210
ResNetByNode macro report 209
ResNetOneNode macro report 209
ResNode macro report 214
ResNodeByGroup macro report 216
ResNodeByNode macro report 215
ResOneNode macro report 215
ResPeCpuByGroup macro report 194
ResPsByGroup macro report 222
ResPsByNode macro report 222
ResVdskByGroup macro report 225
ResVdskByNode macro report 224
ResVdskOneNode macro report 224
EXECUTE MACRO statement
parameters 32, 33
syntax of 31
Extent driver I/O column, ResUsageSvpr 156
F
File system columns
AutoCylPack 143
BLC 144
cylinder management overhead events 141
cylinder overhead 243
data block merge 143
data block prefetches 53, 113, 135
data segment lock requests 54, 141
Depot 54
FSG Cache Wait 142
FSG I/O 142
MI 142
segment acquired columns 133
segments acquired 52, 112
segments released 52, 114, 136
synchronized full table scans 133
write ahead logging 142
File system columns, ResUsageSpma 52
Formats
using FromDate 32
using FromNode 33
using Node 33
266
using ToDate 32
using ToNode 33
using ToTime 32
FSG Cache Wait columns, ResUsageSvpr 142
FSG I/O column, ResUsageSvpr 142
G
Gather Buffer 16
General concurrency control database locks columns
for ResUsageSpma 55
General concurrency monitor management columns,
ResUsageIvpr 248
General format of macros 181
GmtTime 37
Group coordination message columns, ResUsageSpma 59
H
Host communications traffic macros
ResHostByGroup 198
ResHostByLink 198
ResHostOneNode 198
Host controller columns
channel management 82
channel traffic 55, 82
Network traffic 55
user command 82
user command arrival and departure 83
Housekeeping columns
for miscellaneous housekeeping 45, 49, 72, 79, 85, 93, 102,
131, 241
for relational primary index 45, 49, 71, 79, 85, 93, 101, 123,
131, 233, 241
How to re-enable logging 34
I
I/O statistics columns
Input and output traffic columns, ResUsageSldv 87
Input format for macros 29
L
Log Buffer 16
Logging period 16
Logging rates 27
Logging resource usage tables 20
Logic device columns
input and output traffic 87
outstanding requests 88
response time 87
Index
Logical device information

macros 200
table 85
M
Macros
All-node 29
ByGroup 29, 30
examples of 33
executing 31
input format 29
Multiple-node 29
One-node 29, 30
output format 181
overview 17
syntax for 31
types of 181
Master Index columns. See MI 142
Memory allocation columns
for ResUsageSpma 56
for ResUsageSps 116
for the ResUsageSvpr table 149
Memory availability management columns, ResUsageSpma
56
Memory management by node macros
ResMemByGroup 204
ResMemMgmtByNode 204
ResMemMgmtOneNode 204
Memory resident columns, ResUsageSvpr 146
Merge services columns, ResUsageSpma 60
Message delivery times columns, ResUsageIvpr 250
Message type columns, ResUsageIvpr 249
MI columns, ResUsageSvpr 142
Migration columns
Miscellaneous housekeeping columns
for ResUsageIvpr 241
for ResUsageSawt 72
for ResUsageScpu 45
for ResUsageShst 79
for ResUsageSldv 85
for ResUsageSpma 49
for ResUsageSps 102
Modes of resource usage data
count 42
max 42
min 42
track 42
Monitor WD columns, ResUsageSps 104
Multinode macro, example of 33

MULTISET tables 37
N
Net circuit management columns
for ResUsageSpma 58
Net columns 150
broadcast net traffic 57, 116, 150
bynet network transport data 237
group coordination message 59
merge services 60
message delivery time 237
message delivery times 250
message type 236, 249
net circuit management 58, 237
net controller status and miscellaneous management 58
net miscellaneous contention management 237
net queues 238
network transport data 57
point-to-point net traffic 57, 117, 150
work mailbox queue 150
Net miscellaneous contention columns
for ResUsageIpma 237
Net queues columns, ResUsageIpma 238
Network traffic columns, ResUsageSpma 55
Network transport data columns, ResUsageSpma 57
Node agent columns, ResUsageSvpr 158
Node information
macros 183, 195
view 173
Node network traffic macros
ResMemMgmtByNode 208
ResNetByGroup 208
ResNetOneNode 208
Nonunique primary index 37
NUPI. See nonunique primary index
O
Occasional event data 38
Outstanding requests columns, ResUsageSldv 88
Overall resource usage information. See summary macros
P
Parameters, EXECUTE MACRO
FromDate 32
FromNode 33
MacroNameAllNode 32
MacroNameByGroup 32
MacroNameMultiNode 32
Node 33
ToDate 32
ToNode 33
267
Index
Partition assignments
listing 259
table convention 258
PE and AMP UDF CPU columns, ResUsageSvpr 154
PE information
macros 192
table 131
view 166
Point-to-point net traffic columns
for ResUsageSpma 57
for ResUsageSps 117
Priority Scheduler columns, ResUsageSpma 64
Priority Scheduler resource usage 220
Process allocation columns, ResUsageSpma 60
Process block count columns
for ResUsageSpma 61
for ResUsageSps 110
Process pending snapshot columns, ResUsageSpma 60
Process pending wait time columns
for ResUsageSpma 62
for ResUsageSps 111
Process scheduling columns 151
chnsignal status tracking 151
CPU utilization 46, 63, 109, 151
cylinder read 151
for CPU utilization 46
PE and AMP UDF CPU 154
pending wait time 111, 155
process allocation 60
process block count 61, 110, 155
process pending snapshot 60
process pending wait time 62
scheduled CPU switching 235
work type summary 243
Purging old resource usage data 35
Q
Question marks, meaning in macro outputs 182
R
Raw disk drive traffic macros 200, 217
ResLdvByGroup 200
ResLdvByNode 200
ResLdvOneNode 200
Relational primary index
and resource usage tables 37
Relational primary index columns 233
for ResUsageIvpr 241
for ResUsageSawt 71
for ResUsageScpu 45
268
for ResUsageShst 79
for ResUsageSldv 85
for ResUsageSps 101
Relational primary index columns, ResUsageSpma 49
ResAmpCpuByGroup macro
what it reports 188
ResAWT macro
what it reports 183
ResAWTByAMP macro
what it reports 183
ResAWTByNode macro
what it reports 183
ResCPUByAMP macro
what it reports 188
ResCPUByAmp macro
example of 33
ResCPUByAMPOneNode macro
what it reports 188
ResCPUByGroup macro
what it reports 195
ResCPUByNode macro
what it reports 195
ResCPUByPE macro
what it reports 192
ResCPUByPEOneNode macro
what it reports 192
ResCPUOneNode macro
what it reports 195
ResCPUUsageByAMPView, definition 164
ResCPUUsageByPEView, definition 166
Reserved columns
for ResUsageSawt 76
for ResUsageScpu 47
for ResUsageShst 83
for ResUsageSpma 66
for ResUsageSps 117
ResHostByGroup macro
what it reports 198
ResHostByLink macro
what it reports 198
ResHostOneNode macro
what it reports 198
ResIpmaView, definition 254
ResIvprView, definition 256
ResLdvByGroup macro
what it reports 200, 217
ResLdvByNode macro
input format example 200
Index
ResLdvOneNode macro
ResMemByGroup macro
what it reports 204
ResMemMgmtByNode macro
what it reports 204
ResMemMgmtOneNode macro
what it reports 204
ResNetByGroup macro
what it reports 208
ResNetByNode macro
what it reports 208
ResNetOneNode macro
what it reports 208
ResNode macro
what it reports 211
ResNodeByGroup macro
what it reports 211
ResOneNode macro
what it reports 211
Resource usage data
and what it covers 15
benefits of 15
deleting old data 35
saving old data 31
Resource usage macros. See macros
Resource usage tables
enabling Summary Mode 28
naming convention 37
primary index 37
reporting Summary Mode 42
types of 20
Resource usage views. See Views
ResPeCpuByGroup macro
what it reports 192
Response time columns, ResUsageSldv 87
ResPs macros 220
ResPsByGroup macro
what it reports 220
ResPsByNode macro
what it reports 220
ResSawtView, definition 168
ResScpuView, definition 169
ResShstView, definition 170
ResSldvView, definition 171
ResSpdskView, definition 172
ResSpmaView, definition 173
ResSpsView, definition 175
ResSvdskView, definition 178
ResSvprView, definition 179
ResUsageIpma table 20
housekeeping columns 233
mode 233, 234, 236, 237, 238, 239
spare columns 240

statistics columns 235
ResUsageIvpr table 20
mode 241, 243, 244, 245, 246, 247, 248, 249, 250
spare columns 252
Summary Mode 251
ResUsageSawt table
spare columns 76
Summary Mode 76
when enabling 20
ResUsageScpu table
reserved columns 47
spare columns 48
Summary Mode 47
when enabling 20
ResUsageShst table
spare columns 84
Summary Mode 84
when enabling 20
ResUsageSldv table
mode 85, 87
spare columns 89
Summary Mode 89
when enabling 20
ResUsageSpdsk table
mode 93, 97
spare columns 98
Summary Mode 98
when enabling 20
ResUsageSpma table
for housekeeping columns 49
relational primary index columns 49
spare columns 66
Teradata VS columns 65
when enabling 21
ResUsageSps table
spare columns 117
when enabling 21
ResUsageSvdsk table
269
Index
mode 123, 124, 125, 126

spare columns 128
Summary Mode 128
ResUsageSvpr table
allocation columns 155
file system columns 133
for housekeeping columns 131
memory columns 146
net columns 150
process scheduling columns 151
reserved column 158
spare columns 159
Summary Mode 159
when enabling 21
ResVdskByGroup macro
what it reports 223
ResVdskByNode macro
what it reports 223
ResVdskByNode macros
ResVdskByGroup 223
ResVdskByNode 223
ResVdskOneNode 223
ResVdskOneNode macro
what it reports 223
RSS logging
enabling from DBW 27
S
Saving old resource usage data 31
Scheduled CPU switching columns, ResUsageIpma 235
Segments acquired columns
for ResUsageSpma 52
for ResUsageSps 112
Segments released columns
for ResUsageSpma 52
for ResUsageSps 114
SET LOGTABLE command 28
SET RESOURCE command 27
SET SUMLOGTABLE command 28
Single-node. See One-Node
SQL statement
EXECUTE MACRO 31
Statistics columns 95, 125
allocation 125
file system 52, 133, 243
general concurrency control database locks 55, 146
general concurrency control monitor management 248
host controller 82
270
host controller channel and Network traffic 55

logical device 87
memory 56
memory allocation 116
Memory columns 146
migration 126
net 57, 116, 236, 249
process scheduling 60, 109, 235, 243
reserved 47, 66, 76, 83, 117, 127
Reserved columns 98
TASM 64, 73
Teradata VS 65, 95, 125, 155
user command 65
Stopped logging 34
Summary macros 211
ResNode 211
ResNodeByGroup 211
ResNodeByNode 211
ResOneNode 211
Summary Mode
benefits of 22
enabling 23
Synchronized full table scans columns, ResUsageSvpr 133
Syntax for
deleting old resource usage data 35
executing macros 31
System information macros 183
T
Table naming conventions 37
Task context segment usage columns, ResUsageSvpr 149
TASM columns
AMP Worker Task 64, 73, 109
file system 112
Monitor WD 104
Priority Scheduler 64
process scheduling 109
work type descriptions 74
Teradata Active System Management. See TASM columns
Teradata virtual storage. See Teradata VS columns 41
Teradata VS columns 41, 155
allocation 95, 155
extent driver I/O 156
I/O statistics 96
Migration 96
node agent 158
Types of resource usage tables 20
Types of statistics reported 39
U
User command columns
arrival 66, 83
Index
departure 66, 83
ResUsageShst 82, 83
V
Views
ResCPUUsageByAMPView 164
ResCPUUsageByPEView 166
ResIpmaView 254
ResIvprView 256
ResSawtView 168
ResScpuView 169
ResShstView 170
ResSldvView 171
ResSpdskView 172
ResSpmaView 173
ResSpsView 175
ResSvdskView 178
ResSvprView 179
W
WAL columns, ResusageSvpr 142
Work mailbox queue columns, ResUsageSvpr 150
Work type descriptions columns, ResUsageSawt 74
Write ahead logging columns. See WAL 142
271
Index
272

Teradata

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Teradata

Uploaded by

Copyright:

Available Formats

Teradata Database

Resource Usage Macros and Tables

Copyright 2000-2014 by Teradata. All Rights Reserved.

Observe the balance of disk and CPU usage

Obtain pdisk usage information

Check for bottlenecks

Provide an overall history of the system operation

Monitor the utilization of AMP Worker Tasks (AWTs)

Collecting RSS data, see Application Programming Reference.

Carrie Ballinger, Workload Management Capacity on Demand: Teradata Database 14.10

Teradata Viewpoint User Guide

Supported Software Releases and Operating

SUSE Linux Enterprise Server (SLES)10

Teradata Database client applications support other operating systems.

Resource Usage Macros and Tables

Changes to This Book

Teradata Database 14.10.03

Added a new feature: Workload Management Capacity on Demand.

Modified the following field descriptions:

Teradata Database 14.10

Updated the following:

Resource Usage Macros and Tables

Teradata Database 14.10

Updated the following views:

May 2013 continued

Use the Teradata Information Products Publishing Library site

Download a documentation CD-ROM:

The Teradata home page provides links to numerous sources of

Resource Usage Macros and Tables

Teradata Customer Education delivers training that builds skills

Use Teradata @ Your Service to access Orange Books, technical

Teradata Developer Exchange provides articles on using

Teradata Database Optional Features

Teradata Row Level Security

Teradata Virtual Storage (VS)

Unity Source Link

Resource Usage Macros and Tables

Chapter 2: Planning Your Resource Usage Data . . . . . . . . . . . . . . . 19

Resource Usage Macros and Tables

Chapter 3: Resource Usage and Procedures . . . . . . . . . . . . . . . . . . . .27

Chapter 4: Resource Usage Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37

Chapter 5: ResUsageScpu Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45

Resource Usage Macros and Tables

Chapter 6: ResUsageSpma Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Chapter 7: ResUsageSawt Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Chapter 8: ResUsageShst Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Chapter 9: ResUsageSldv Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85

Chapter 10: ResUsageSpdsk Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93

Chapter 11: ResUsageSps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101

Resource Usage Macros and Tables

Chapter 12: ResUsageSvdsk Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Chapter 13: ResUsageSvpr Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Chapter 14: Resource Usage Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

Resource Usage Macros and Tables

Chapter 15: Resource Usage Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . .181

Appendix A: How to Read Syntax Diagrams . . . . . . . . . . . . . . . . . . .227

Appendix B: ResUsageIpma Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .233

Resource Usage Macros and Tables

Appendix C: ResUsageIvpr Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241

Summary Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

Appendix D: ResIpmaView and ResIvprView Views . . . . . . . . 253

Appendix E: Partition Assignments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257