DB2 Enterprise Edition Satge- Max Log issue

ysrini9 · Post by **ysrini9** » Tue Jul 03, 2012 3:58 am

Hi,

The job cleans the old data in staging table for unique month,year(i.e.,delete the records from target Db2 table- records count: 7.5 lakhs)
write Method:Delete rows
Delete Rows Mode:Auto-generated Delete.
fatal errors:
1.tblCCMIPreStagdel,0: ERROR during execution of DB2-CLI requests: SQL_ERROR; SQLSTATE = 40003; NativeErrorCode = -1224; Message = [IBM][CLI Driver] SQL1224N The database manager is not able to accept new requests, has terminated all requests in progress, or has terminated the specified request because of an error or a forced interrupt. SQLSTATE=55032.
2.tblCCMIPreStagdel,0: Failed to a execute a DB2statement; examine the printed SQL error.
3.tblCCMIPreStagdel,0: ERROR during execution of DB2-CLI requests: SQL_ERROR; SQLSTATE = 08003; NativeErrorCode = -99999; Message = [IBM][CLI Driver] CLI0106E Connection is closed. SQLSTATE=08003 4.tblCCMIPreStagdel,0: Rollback failed.
---
Note: we Tested with diffrent array size and row commit interval.
pls do the needful...
Thansk in advance.
Srini Y

ArndW · Post by **ArndW** » Tue Jul 03, 2012 4:04 am

The Database has run out of space to (temporarily) store the records in your delete transaction. Either define a smaller transaction size or have your DBA increase the rollback space.

What commit sizes did you experiment with? Which stage are you using on which release?

ysrini9 · Post by **ysrini9** » Tue Jul 03, 2012 4:19 am

Thanks for your response...
we tried with
1.array size : 1000 and commit size: 2000
2.array size : 10000 and commit size: 20000
3.array size : 5000 and commit size: 10000
4.array size : 10000 and commit size: 10000
5.array size : 100 and commit size: 200

DB2_UDB_Enterprise stage(DB2 Version:9.5)
Datastage 8.1.

ysrini9 · Post by **ysrini9** » Tue Jul 03, 2012 4:40 am

For this issue.. We have one script .. that script( for unique month,year) is working fine from command line.. we are getting unique data based on the month,year columns from source file.
How can I call the same script in the datastage job level?

ArndW · Post by **ArndW** » Tue Jul 03, 2012 5:00 am

What does your DELETE statement look like. Basically, if the delete is only called one time, i.e. one input row and the delete is "DELETE * from <table> where <condition>;" then it might burst the rollback space regardless of the stage settings and that is what it looks like is happening. This is a typical problem with DB2 when trying truncate a table (there used to be no "truncate" command, one needs to execute a tricky "LOAD FROM /dev/null of del REPLACE INTO <table>").
If performance is not the highest priority, you could do a job where you SELECT all the keys of records to be deleted in one stage, then call up the DB2 stage with DELETE mode - in that case your settings for commit frequency and array size will be used.

MT · Post by MT » Tue Jul 03, 2012 6:35 am

Hi ysrini9

SQL1224N The database manager is not able to accept new requests, has terminated all requests in progress, or has terminated the specified
request because of an error or a forced interrupt. SQLSTATE=55032.

If you read the message carefully it says that your connection has been terminated - i.e. forced of by an admin.
So I would suggest you talk to your DBA first.

It can be log space problem - which is a common one for deletes - but it can also be a WLM rule which has been violated.

For these kind of deletes you could also talk to your DBA and suggest using MDC tables which have a very good delete performance and reduced logging. I like them for getting a good delete performance when deleting a big number of rows from a huge table.

DSXchange