 |
 |
Archives of the TeradataForum
Message Posted: Wed, 03 Aug 2011 @ 12:51:09 GMT
| Subj: | | Re: Difference between collect stats on a column vs. a table |
| |
| From: | | De pons, Paul |
One note on table level stats is that it is all or nothing. Stats are not committed on any column until all stats are collected. Since stats
are resource intensive, if you have a large table with lots of column stats defined it can take a long time to complete. Killing it in the middle
to temporally free up resources could waist allot of CPU. Where as if you are collecting column by column, you can kill the script, get the
benefits of the refreshed stats already collected and then pick up from where you left off once resources are available.
Paul Martin De Pons
| |
