Archives of the TeradataForum

Message Posted: Wed, 10 Jun 2009 @ 16:08:22 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Joerg,

If you have an application that is using the dictionary tables to look up these types of things, then those queries are going to be subject to whatever workload management scheme you have in place.

Certainly, one way to help improve these would be to use TASM workload classification to direct queries that are expected to use very little CPU, to a workload in a very highly weighted AG. (This workload should have a tight CPU time exception to move over consumers who get in here to a lower priority AG.)

You could do something similar with Priority Scheduler using query milestones to allow short queries a high priority.

One other thing to consider is the collection of statistics on some of the DBC tables themselves. Using Diagnostic Helpstats on for session, you can do an explain on your dictionary look up queries for columns to consider for the collection of statistics.

There have been cases where this really improved the performance of applications like this.

Steve Woodrow
Performance and Workload Management COE
Teradata
Americas Professional Services

	<Prev	Next>		<<First	<Prev	Next>	Last>>