1 of 1 people found this helpful
Hi Lana. I'm the Director for our DB2 Customer Success team. My support team have filled me in on what you're experiencing. Hopefully, some of their suggestions have helped, but your Idea is also something to be considered for additional improvements. I'm going to have one our support members reach back out to you on this and help get this added to our DB2 product Ideas. Our support team will also update this discussion with a related Knowledge Article link that discusses some of the suggestions to help minimize the CPU consumption when populating the PADB.
We'd be happy to take a look at your process and see if configuration changes might help you. You might also check to see if you have BQU2831 applied. This was part of our May 2020 SQL Performance enhancement set, and will be included in the RSL2010 quarterly maintenance release. With BQU2831, the generation of records to load into object-related tables is more efficient, with fewer sorts.
This KA on performance tips for the PADB also might be of interest:
Technical Support Analyst
DB2 System and SQL Performance
Beth, the KA is full of great and necessary information. Will this also be added to the manuals that discuss how to set up Output Groups (if it’s not there already) and the use of the DMDBMERG utility? It would be extremely helpful to have this information in those places.
Thanks so much for your help.
DOMPLEX Output Group recommendations are provided in our online documentation here:
That said, I see that there is no real verbiage here related to output group configuration in terms of how it can align effectively with the PADB extracts. Although
there are comments sprinkled throughout the DMDBMERG utility section related to efficiency, your comment about perhaps adding a performance tips or
recommendations section is noted. I'll be glad to take your suggestion forward.