We have a payments application that selects from a db about every 30 seconds. It usually runs in less than 30 seconds. However, when a db backup is running it takes more than 4 minutes to run. The application has logic that bring the exe down on the client and issues an alert when it runs past 4/5 minutes. This always seems to happen at the end of a backup. I have seen where a backup will continue to do some processing after all the pages have been backed up. Can someone tell me what happens a the very end of a backup before it returns control and how long does it usually take? Whatever it is, would it cause read locks, causing an app to take more time to run ? Since this is a SWIFT payments app, the select can't use with ur.

DB21085I Instance "db2inst1" uses "64" bits and DB2 code release "SQL09053"
with level identifier "06040107".
Informational tokens are "DB2 v9.5.0.3", "special_22945", "U818976_22945", and
Fix Pack "3".
Product is installed at "/app/liqdg/sw/IBM/db2/V9.5".

Solaris 10 OS

Try setting the UTIL_IMPACT_PRIORITY to 10 on the backup command. Obviously, this will make your backup run longer, but other DB2 activities will be less affected by the backup.

I have UTIL_IMPACT_LIM = 10 for the db. Doesn't that do the samething ? Remember I don't have any issues while the backup is running. Its after all the pages have been backed up and db2 is doing something at the very end that I have the issue.

Workload impact by throttled utilities(UTIL_IMPACT_LIM) = 10

Thanks

How do you know all pages have been backed up (from list utilities? ) ? It's most likely including the logs (default with online backup in v9.5). You can execute db2pd with -app -tra -loc -uti to see exactly what's happening.

Try using the UTIL_IMPACT_PRIORITY parameter in the backup command itself.