Re: Autovacuum ideas
От | Brendan Duddridge |
---|---|
Тема | Re: Autovacuum ideas |
Дата | |
Msg-id | BFE1A0A4-146E-453A-AFFC-19DB1C237D81@clickspace.com обсуждение исходный текст |
Ответ на | Re: Autovacuum ideas (Robin Iddon <robin@edesix.com>) |
Список | pgsql-admin |
I guess I was thinking that if you do a vacuum analyze verbose from within psql, it does vacuum the big 'insert only' tables. Of course it never finds any dead rows, but it does take a long time to get past those tables. I didn't know that autovacuum would be any different. ____________________________________________________________________ Brendan Duddridge | CTO | 403-277-5591 x24 | brendan@clickspace.com ClickSpace Interactive Inc. Suite L100, 239 - 10th Ave. SE Calgary, AB T2G 0V9 http://www.clickspace.com On Apr 13, 2006, at 1:36 AM, Robin Iddon wrote: > Brendan Duddridge wrote: >> What I'd like to see is a table exclusion list. I have a few very >> large history tables that are never updated or deleted, only >> inserts and selects. > > Such a table will never trigger the vacuum rules as I understand > them (vacuum only happens on table that have obsolete tuples, which > means update or delete). > > It will correctly trigger the analyze rules now and then, but > analyze is cheap compared to vacuum and is desirable because it > will help the planner do it's job (assuming that querying the table > is important to you). Remember analyze only requires a read-lock > on the table so it can run in parallel with other queries quite > happily. > > If you really need to disable autovac on a table you can disable > pg_autovacuum from running on a specific table by creating a row > for your table in pg_autovacuum table and setting the > pg_autovacuum.enabled to false. > > > > Robin > > > ---------------------------(end of > broadcast)--------------------------- > TIP 6: explain analyze is your friend >
В списке pgsql-admin по дате отправления: