|
|
Archives of the TeradataForum
Message Posted: Fri, 09 Mar 2007 @ 22:10:22 GMT
Subj: | | Re: CPU sec variations |
|
From: | | Anomy Anom |
<-- Anonymously Posted: Friday, March 09, 2007 16:39 -->
| How are you measuring the CPU? AMPUsage, DBQL, or what? Could it be the effect of accessing cached data and/or SyncScan? | |
I verified that we do have current statistics on all tables involved. We are getting the CPU sec info from the AMPUsage tables. I was
thinking the cached query might be an issue but the odd thing is that the earlier executed queries are the ones with the shorter runtime and I
would think that it's the later queries that would benefit from cached SQL. The SyncScan sounds like it could be a possibility. is there anyway
to verify this?
Thanks again!
| |