SB37 ABEND on IXX601 in Datacom/AD for CA 11
search cancel

SB37 ABEND on IXX601 in Datacom/AD for CA 11

book

Article ID: 54792

calendar_today

Updated On:

Products

Workload Automation Restart Option for z/OS Schedulers (CA 11) Datacom/AD

Issue/Introduction

A rapid and abnormal increase in IXX601 utilization occurs, leading to an Abend SB37 on the index file.

This typically happens in environments supporting CA 11 where the index area fills despite organic growth predictions.

Additional indicators found in the MUF joblog may include:

  • DB01705I - DYNAMIC EXTEND START IXX601

  • DB01709I - DYNAMIC EXTEND END IXX601

  • DB02827W - DYNAMIC EXTEND NEAR LIMIT

Rapid jumps in extent allocation (e.g., reaching the 16-extent limit within minutes.

Environment

  • z/OS

  • Datacom/AD (All versions)

  • CA Workload Automation Restart Option for z/OS Schedulers (CA 11)

Cause

This issue is caused by one or more of the following:

  1. Index Fragmentation: Batch activities allocating new blocks rather than reusing fragmented space when the index is not regularly maintained.

  2. Sizing: The IXX601 is not sized large enough to accommodate peak batch activity.

  3. CA11 MANT task: Not being run on a regular basis. 

  4. Product Defect: A known bug (LU00112) can cause a loop in the IXX resulting in rapid, indefinite allocation attempts during MUF restart or active processing.

Resolution

Recovering the IXX601

Suspend or shut down the CA 11 application and make DB 601 unavailable to CA 11 using DBUTLTY:

ACCESS DBID=601,STATUS=OFF,USERS=FAIL
COMM DBID=601,OPTION=CLOSE

If the MUF is down (e.g., during a RESTART error), add the following to the beginning of your JCL SYSIN and omit the ACCESS and COMM commands:

SET OPTION1=MUF_NOT_ENABLED

Allocation and Rebuild

1. Delete or rename the existing IXX601 dataset.

2. Allocate a larger IXX601 dataset. Note: As a rule of thumb, the IXX601 should be at least as large as the larger of your JHF601 or DSN601 datasets.

3. Rebuild the index entries using the sample JCL below:

//INITDB   EXEC PGM=DBUTLTY,REGION=0M
//STEPLIB DD DISP=SHR,DSN=DATACOM.CUSLIB
//         DD DISP=SHR,DSN=DATACOM.CAAXLOAD
//IXX601   DD DISP=(,CATLG),DSN=<HLQ>..IXX601,
//             SPACE=(CYL,(nnn,nn)),UNIT=3390
//SYSIN    DD *
 INIT   DBID=601,AREA=IXX
 RETIX  DBID=601,OPTIMIZE=YES,SORT=999,SORTDFLT=YES,OPTION1='(I)'
 COMM   DBID=601,OPTION=STATS
 REPORT DBID=601,AREA=CXX,TYPE=A
 ACCESS DBID=601,STATUS=WRITE
/*
 


Preventative Maintenance

Apply Required PTFs

Apply the following PTFs to resolve the index loop defect:

LU00265 and LU00589 (Fixes for LU00112)

• Latest level set recommended.

Implement Regular DEFRAG

Run the DBUTLTY DEFRAG function monthly or more frequently following the CA 11 MANT process. This reclaims space while the database remains active:

DEFRAG DBID=601

 

Additional Information

To speak with a customer representative or a Support Engineer, see Contact Support  or refer to KB 206567.