Re: Performance of autovacuum and full vacuum of database - Mailing list pgsql-general

From Carlos Oliva
Subject Re: Performance of autovacuum and full vacuum of database
Date
Msg-id 200511111813.NAA21395@pbsi.pbsinet.com
Whole thread Raw
In response to Re: Performance of autovacuum and full vacuum of database  ("Matthew T. O'Connor" <matthew@zeut.net>)
Responses Re: Performance of autovacuum and full vacuum of database  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-general
Hi Matthew,
Would rebooting the server interfere with the work of pg_autovacuum?  I
imagine that pg_autovacuum would loose the information that it gathered
prior to the reboot.

Would not pg_autovacuum need to gather data about the database again before
it can judge if it should analyze/vacuum a table of the database?

-----Original Message-----
From: pgsql-general-owner@postgresql.org
[mailto:pgsql-general-owner@postgresql.org] On Behalf Of Matthew T. O'Connor
Sent: Thursday, November 10, 2005 3:44 PM
To: Carlos Oliva
Cc: pgsql-general@postgresql.org
Subject: Re: [GENERAL] Performance of autovacuum and full vacuum of database

Carlos Oliva wrote:
> Thank you for your response Matthew.  Currently I run pg_autovacuum with
the
> following scripts.
> su -l postgres -c "pg_autovacuum -D -U postgres > /dev/null 2>&1"&
>
> Do you suggest that I could change it to something like the following:
> su -l postgres -c "pg_autovacuum -d2 -D -U postgres > /tmp/vacuum.log
2>&1"&

Yes exactly, and if you find that pg_autovacuum is never or not often
enough firing off vacuum comands, then you will need to play with the
threshold settings.  The default thresholds for pg_autovacuum are too
conservative for most people, so you may very well have to do this.


---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
       choose an index scan if your joining column's datatypes do not
       match



pgsql-general by date:

Previous
From: Tom Lane
Date:
Subject: Re: Best way to use indexes for partial match at
Next
From: Bob Pawley
Date:
Subject: Re: Where