From: Csaba Nagy <nagy@ecircle-ag.com>
To: postgres performance list <pgsql-performance@postgresql.org>
Subject: How long it takes to vacuum a big table
Date: Fri, 28 Oct 2005 17:14:21 +0200
Message-ID: <1130512461.25950.20.camel@coppola.muc.ecircle.de> (raw)
Hi all,
I wonder what is the main driving factor for vacuum's duration: the size
of the table, or the number of dead tuples it has to clean ?
We have a few big tables which are also heavily updated, and I couldn't
figure out a way to properly vacuum them. Vacuuming any of those took
very long amounts of time (I started one this morning and after ~5h30min
it's still running - and it's not even the biggest or most updated
table), which I can't really afford because it prevents other vacuum
processes on smaller tables to do their job due to the transaction open
for the long-running vacuum.
BTW, is it in any way feasible to implement to make one vacuum not
blocking other vacuums from cleaning dead tuples after the first one
started ? I know it's the transaction not the vacuum which blocks, but
then wouldn't be a way to run vacuum somehow in "out of transaction
context" mode ?
Another issue: vacuum is not responding to cancel requests, at least not
in a reasonable amount of time...
Thanks in advance,
Csaba.
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: pgsql-performance@postgresql.org
Cc: nagy@ecircle-ag.com
Subject: Re: How long it takes to vacuum a big table
In-Reply-To: <1130512461.25950.20.camel@coppola.muc.ecircle.de>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox