A number of extract jobs are failing with incomplete collections and IEC161I virtual storage errors. How can we resolve this problem?

Products

• RC Compare for DB2 for z/OS
• Database Analyzer for DB2 for z/OS
• Fast Unload for DB2 for z/OS
• Fast Check for DB2 for z/OS
• Fast Index for DB2 for z/OS
• Rapid Reorg for DB2 for z/OS
• CA Distributed Database Management
• CA Database Analyzer for DB2 for Oracle
• Database Analyzer (IMS Tools)
• Mainframe Configuration Manager for IMS for z/OS
• IMS TOOLS - MISC
• Compress Data Compression (IMS Tools)
• Database Analyzer for IMS for z/OS
• Database Copier for IMS for z/OS
• Database Organizer for IMS for z/OS
• Mainframe Extended Terminal Manager (IMS Tools)
• High Performance Recovery for IMS for z/OS
• Database Organizer (IMS Tools)
• Mainframe Program Restart Manager for IMS for z/OS
• Secondary Index Builder for IMS for z/OS
• Secondary Index for IMS for z/OS

Issue/Introduction

Description:

We are running extract jobs, some of which are failing with messages indicating that some Tablespace collections are incomplete.
The Database Analyzer report in the job output shows the following messages for some of the Collections of data statistics. The majority of the collections have ended successfully.

OPEN PROCESS FAILED FOR DB2 DATASET
10:28:43
Collection of TS - dbname.tsname          - Incomplete
10:28:43

The job JESMSGLG messages in the job output also show the following z/OS system messages:

IEC161I 020-053,jobname,stepname,ddname,dev,ser,xxx,dsname,cat
IEC161I 020-054,jobname,stepname,ddname,dev,ser,xxx,dsname,cat

How do we resolve this collection problem?

Solution:

The return code 020 in the IEC161I 020-053 and 020-054 messages suggest that there is a Getmain problem and not enough virtual storage is available to open the VSAM datasets.

Check the REGION size for your job. Increasing this may resolve the Getmain failure and allow the collections to complete successfully.

If this does not work then we have to look at reducing the amount of virtual storage required for the extract.

We have two options, reduce the number of scans, or reduce the buffer size.

Check the final settings for the collection options in the job output, to confirm what settings were in operation for the failing job:

Collection of data statistics Started - Subsystem - ssid
Access: D/EXCP Buffers: 0360 Scans: 0008 Scantype: FASTSCAN

The number of buffers requested affects the amount of memory needed for the extract. The more buffers you assign, the less physical I/O Database Analyzer performs. If you set the Access field to DIV, the VSAM Buffers field is not used.

Entering a high number of scans does not reduce CPU time, but could reduce elapsed time. A value of 3 to 5 is recommended. Scan values above 5 can severely impact other address spaces depending on hardware and software limitations, current workload, and so forth.

In this example with Buffers set to 360, and Scans set to 8, (8*360) buffers are allocated.

Try one of the following to reduce the amount of virtual storage required for the extract and to resolve the problem:

• Reduce the buffer amount from 360 to 180
• Reduce the number of scans by from 8 to 4

Environment

Release:
Component: DBA