Yes, you gather a separate set of stats for each significantly distinct
workload and swap them in for that workload. The key word is significantly.
Don't try to micromanage statistics. You are dealing with statistics, don't
try to interpret more accuracy into them as there is.
At 10:23 AM 3/5/2004, you wrote:
So you define the timeframe for gathering the stats. And in a mixed case workload, say OLTP during the day, big batch processing during the night, do you gather two sets of stats with an appropriate timeframe, and swap them out, just as described in chapter 3 of the 9i tuning guide? Do you find yourself needing to manually tweak the system stats? And I've seen systems where there is a predictable change in workload during the day, simply based on how people do their work, so in a case like that, maybe even swap out system stats during the day (assuming you have a good handle on how the system is used)?
Regards,
Larry G. Elkins elkinsl@xxxxxxxxx 214.954.1781
---------------------------------------------------------------- Please see the official ORACLE-L FAQ: http://www.orafaq.com ---------------------------------------------------------------- To unsubscribe send email to: oracle-l-request@xxxxxxxxxxxxx put 'unsubscribe' in the subject line. -- Archives are at //www.freelists.org/archives/oracle-l/ FAQ is at //www.freelists.org/help/fom-serve/cache/1.html -----------------------------------------------------------------
