From: | "Ian Harding" <ianh(at)tpchd(dot)org> |
---|---|
To: | <pgsql-general(at)postgresql(dot)org> |
Subject: | VACUUM ANALYZE makes things worse! |
Date: | 2002-04-12 20:33:34 |
Message-ID: | scb6e27e.075@mail.tpchd.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
My troublesome view is called vtb_timecard.
This view runs acceptably quickly after re-creating and reloading the database, but is intolerable slow after vacuum analyze. I assume I have inadvertently tricked the optimizer into doing The Wrong Thing somehow, and I hope what it is will be apparent to someone on the list.
I have indexes that help out on all the joins in the "s" query, and have tried creating the additional ones such as on timedetail.shiftdiff and timedetail.adjperiod but with no affect. I thought the issue might not only be a missing index on timedetail.adjperiod, but the fact that that field is usually NULL, so I tried populating the nulls with zeroes, creating and index, and analyzing again, still no luck.
I have put the schema and explain output on a website where I hope you can find them...
http://www.pakrat.com/pgtrouble
PS If you are squeamish about an occasional obscenity in remarks, please don't even look, I am sure there are one or two.
Ian A. Harding
Programmer/Analyst II
Tacoma-Pierce County Health Department
(253) 798-3549
mailto: iharding(at)tpchd(dot)org
From | Date | Subject | |
---|---|---|---|
Next Message | Ian Harding | 2002-04-12 20:43:18 | Nested Loop WAS: VACUUM ANALYZE makes things worse! |
Previous Message | Justin Clift | 2002-04-12 19:37:10 | Binaries for Solaris |