Re: handling TOAST tables in autovacuum

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Alvaro Herrera <alvherre(at)commandprompt(dot)com>
Cc: Pg Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: handling TOAST tables in autovacuum
Date: 2008-06-09 00:23:25
Message-ID: 25296.1212971005@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Alvaro Herrera <alvherre(at)commandprompt(dot)com> writes:
> The point here is that if the user disables autovac for the main table,
> then it's expected that it is automagically disabled for the toast table
> as well, for the usual case where they are disabling it because the
> table is too big.

Hmm, good point. OK, two passes it is. (I thought about remembering
the toast table rows in memory so as not to scan the catalog twice,
but I'm not sure you really save much that way.)

Another thing to think about here is locking: I believe you need to get
a vacuum-type lock on the parent table not only the toast table, so
vacuuming a toast table without any knowledge of which table is its
parent ain't gonna fly anyway.

regards, tom lane

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Andrew Dunstan 2008-06-09 00:33:52 Re: libpq support for arrays and composites
Previous Message Alvaro Herrera 2008-06-09 00:15:15 Re: handling TOAST tables in autovacuum