From: | Alvaro Herrera <alvherre(at)commandprompt(dot)com> |
---|---|
To: | Arctic Toucan <arctic_toucan(at)hotmail(dot)com> |
Cc: | pgsql-admin(at)postgresql(dot)org |
Subject: | Re: Is it safe to reset relfrozenxid without using vacuum? |
Date: | 2008-11-18 16:24:06 |
Message-ID: | 20081118162406.GD4141@alvh.no-ip.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Arctic Toucan wrote:
>
> I have a heavily partitioned DW type database that has over 5,000
> tables in it. Data is only ever inserted, read and then truncated
> after some period of time. Once the table is truncated, the
> constraints are changed and the table is reused. This works well until
> Postgres hits the autovacuum_freeze_max_age, which I currently have
> set @ 1billion). Since these tables are only ever inserted to and
> truncated, they are not normally vacuumed(which is what I want, since
> data is typically going to be truncated before needing to be
> vacuumed).
AFAICS this should be safe. In fact, in 8.3 TRUNCATE advances
relfrozenxid. (Perhaps you should consider upgrading if possible.)
> UPDATE pg_class SET relfrozenxid = ( select relfrozenxid from pg_class
> where age(relfrozenxid) in (select min(age(relfrozenxid)) from
> pg_class where relkind = 'r') limit 1) WHERE relname = '<table being
> truncated>';
Tou could just obtain the xid of the transaction that's going to do the
import (for example by creating a temp table and getting it's xmin from
pg_class)
--
Alvaro Herrera http://www.CommandPrompt.com/
PostgreSQL Replication, Consulting, Custom Development, 24x7 support
From | Date | Subject | |
---|---|---|---|
Next Message | Mischa Sandberg | 2008-11-18 16:47:24 | Re: Reliably determining whether the server came up |
Previous Message | Tom Lane | 2008-11-18 15:59:42 | Re: Reliably determining whether the server came up |