Long runnning IDOC Archiving jobs

[Former Member]

Hi Techies,

Long runnning IDOC Archiving jobs: ARV_IDOC_WRI* are being observed in PRD system with Avg runtime > 150,000 sec.

Job Details:

Main Program: RSEXARCA

Parameter: IDOC ALL MONTH

These jobs are also resource consuming & extensively using CPU & Memory.

For info: These jobs are getting finished but only after long runtime.

Some SQL stmts observed while this job-execution:

SELECT

"OBJKEY" , "OBJTYPE" , "LOGSYS" , "ROLETYPE" , "ROLEID" , "UTCTIME"

FROM

"SRRELROLES"

WHERE

"CLIENT" = :A0 AND "OBJKEY" = :A1 AND "OBJTYPE" = :A2#

SELECT

"MANDT" , "DOCNUM" , "COUNTER" , "PAGENO" , "TIMESTMP" , "PAGELG" , "VARDATA"

FROM

"EDI40"

WHERE

"MANDT" = :A0 AND "DOCNUM" = :A1

ORDER BY

"MANDT" , "DOCNUM" , "COUNTER" , "PAGENO"#

SELECT

"DOCNUM" , "DOCREL" , "STATUS"

FROM

"EDIDC"

WHERE

"MANDT" = :A0 AND "UPDDAT" >= :A1 AND "UPDDAT" BETWEEN :A2 AND :A3 AND "DIRECT" BETWEEN :A4 AND :A5 AND

"STATUS" = :A6 AND NOT "MESTYP" IN ( :A7 , :A8 , :A9 , :A10 , :A11 , :A12 , :A13 , :A14 , :A15 , :A16 ,

:A17 , :A18 , :A19 , :A20 , :A21 , :A22 , :A23 , :A24 , :A25 , :A26 , :A27 , :A28 , :A29 , :A30 , :A31 )

#

Please suggest as to how the job runtime can be reduced.

Awaiting for your reply.

[Former Member]

Hi,

Since the variant used is fetching data for the entire month, you could try splitting the jobs and have multiple runs instead.

You can also try to schedule the jobs when there is minimum system load, incase its not already done.

As far as the program is concerned, if this is standard program, there is little scope for tuning.

You can always check for related index quality and statements with high CPU costs to better understand why the job is running for such a long duration.

[JPReyes]

Most likely the tables are big and is taking long while doing sequencial reads... you may want to look into creating an index to speed up the process

Regards

Juan