| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Mailing Lists -> Oracle-L -> RE: CBO irregularity
I believe the comment is intended to reflect the fact that old statistics may well be valid statistics. If you gathered statistics a year ago and the data hasn't changed substantially, the statistics would still be valid. Many DBA's don't gather statistics on a regular basis so as not to disrupt a production database with acceptable plans.
Justin Cave
Distributed Database Consulting, Inc.
http://www.ddbcinc.com/askDDBC
-----Original Message-----
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org]
On Behalf Of Jacques Kilchoer
Sent: Monday, June 07, 2004 5:46 PM
To: oracle-l_at_freelists.org
Subject: RE: CBO irregularity
When would up-to-date statistics not be valid? Do you mean, for example, a different number of histograms or different sample sizes were collected with the "up-to-date" statistics?
-----Original Message-----
Daniel Fink
...
If you don't have valid (not the
same as up-to-date) statistics on all the objects (and corollary
objects) referenced in the query, you might get some defaults that are
unrealistic or it may perform dynamic sampling.
 |
 |