Date: Tue, 12 Jun 2007 12:25:55 -0400
From: Bill Moran <wmoran@collaborativefusion.com>
To: "Sabin Coanda" <sabin.coanda@deuromedia.ro>
Cc: pgsql-performance@postgresql.org
Subject: Re: VACUUM vs auto-vacuum daemon
Message-Id: <20070612122555.2b0f4fa0.wmoran@collaborativefusion.com>
In-Reply-To: <f4mfin$1vjf$1@news.hub.org>
References: <f4mfin$1vjf$1@news.hub.org>
Organization: Collaborative Fusion
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

In response to "Sabin Coanda" <sabin.coanda@deuromedia.ro>:

> Hi there,
> 
> Using explicitly VACUUM command give me the opportunity to fine tune my 
> VACUUM scheduling parameters, after I analyze the log generated by VACUUM 
> VERBOSE.
> 
> On the other hand I'd like to use the auto-vacuum mechanism because of its 
> facilities. Unfortunately, after I made some initial estimations for 
> autovacuum_naptime, and I set the specific data into pg_autovacuum table, I 
> have not a feedback from the auto-vacuum mechanism to check that it works 
> well or not.  It would be nice to have some kind of log similar with the one 
> generated by VACUUM VERBOSE. Is the auto-vacuum mechanism able to provide 
> such a useful log ?

Ditto what Alvaro said.

However, you can get some measure of tracking my running VACUUM VERBOSE
on a regular basis to see how well autovacuum is keeping up.  There's
no problem with running manual vacuum and autovacuum together, and you'll
be able to gather _some_ information about how well autovacuum is
keeping up.

-- 
Bill Moran
Collaborative Fusion Inc.
http://people.collaborativefusion.com/~wmoran/

wmoran@collaborativefusion.com
Phone: 412-422-3463x4023