From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Justin Pasher <justinp(at)newmediagateway(dot)com> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: Continual increase of age(datfrozenxid) for template0 |
Date: | 2009-02-11 00:46:46 |
Message-ID: | 27557.1234313206@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Justin Pasher <justinp(at)newmediagateway(dot)com> writes:
> As you can see, it is starting to approach the 2 billion mark. If I
> change the datallowconn setting to TRUE, connect to template0, then
> perform a VACUUM FREEZE, the age(datfrozenxid) goes way down (probably
> around the 3 million mark). However, over time (typically less than two
> weeks), the number slowly creeps higher and higher, until our custom
> monitoring software starts spitting out a warning/error that the
> database should be vacuumed soon.
Recommendation: fix that monitoring software to ignore databases with
datallowconn = false. They're not supposed to need vacuuming.
(autovacuum will process them anyway for safety's sake, but they're
not supposed to need it.)
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2009-02-11 01:04:17 | Re: encoding of PostgreSQL messages |
Previous Message | Lee Hughes | 2009-02-11 00:16:47 | Re: Referencing Cursor/Row/Record Fields in PL/PgSQL |