A Database Analyzer Extract that previously worked now fails with MMgr-EXCP I/O errors when perfoming a scan.
The following errors can be found in the 'Collection of data statistics' output. These can be issued for any object type, for example:
MMgr-EXCP I/O error. Call Technical Support . RC=0008 RS=0000
Collection of UR - dbname.tsname.0001 - Incomplete RC=8
Additonally the following messages may be issued in the JESMSGLG output
SIX2: MMgr OPEN failure. Missing [email protected] blk!!!
SIX2: MMgr OPEN not done. Bypassing CLOSE!!!
IEA705I ERROR DURING GETMAIN SYS CODE = 878-10 jobname stepname
IEA705I 00F15E80 009AD998 009AD998 00007210 00004A90
SYSTEM COMPLETION CODE=878 REASON CODE=00000010
TIME=10.17.40 SEQ=23348 CPU=4002 ASID=0337
PSW AT TIME OF ERROR 070C1000 8126CC7A ILC 2 INTC 0D
NO ACTIVE MODULE FOUND
DATA AT PSW 0126CC74 - 00181610 0A0D18CE 18FB180C
Release : 20.0
Component : CA Database Analyzer for DB2 for z/OS
Has the number of objects being processed in this Extract increased since it last ran successfully? If so, PDA may have a storage problem that is related to the number of objects.
If the Scan is currently using the EXCP access method, try using the VSAM access method instead for this Extract. EXCP does provide improved performance for elapsed time but it may be having a negative impact on storage.
To switch to using VSAM access for just this step, add the NOEXCP to the PARM options in the EXEC statement, similar to the following:
//STEP10 EXEC PGM=PTLDRIVM,PARM='EP=RALMCTL/NOEXCP'
One other consideration is to reduce the number of objects being processed in this Extract Procedure, by moving some of the objects to another Extract Procedure and to then be processed in a separate step.