Re: determining max_fsm_pages

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Patrick Hatcher <pathat(at)comcast(dot)net>
Cc: pgsql-performance(at)postgresql(dot)org
Subject: Re: determining max_fsm_pages
Date: 2004-10-29 14:37:23
Message-ID: 15929.1099060643@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Patrick Hatcher <pathat(at)comcast(dot)net> writes:
> My question is this: I have about 8 databases running on this server.
> When I do a vacuum full on each of these databases, there is a INFO
> section that I assume is the total pages used for that database. Should
> add ALL these individual pages together and pad the total and use this
> as my new max_fsm_pages? Should I do the same thing with max_fsm_relations?

No, the numbers shown at the end of a vacuum verbose printout reflect
the current cluster-wide FSM demand. BTW you do *not* want to use FULL
because that's not going to reflect the FSM requirements when you are
just running normal vacuums.

I would vacuum all your databases (to make sure each one's FSM contents
are pretty up-to-date) and then take the numbers shown by the last one
as your targets.

If you find yourself having to raise max_fsm_relations, it may be
necessary to repeat the vacuuming cycle before you can get a decent
total for max_fsm_pages. IIRC, the vacuum printout does include in
"needed" a count of pages that it would have stored if it'd had room;
but this is only tracked for relations that have an FSM relation entry.

regards, tom lane

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Jan Wieck 2004-10-30 13:40:39 Re: [PATCHES] ARC Memory Usage analysis
Previous Message Tatsuo Ishii 2004-10-29 14:31:51 Re: determining max_fsm_pages