pg.ddx.io  pgsql-performance@postgresql.org mailing list archive  
help / color / mirror / Atom feed
From: Ron St-Pierre <rstpierre@syscor.com>
To: revoohc@sermonaudio.com
To: pgsql-performance@postgresql.org
Subject: Re: Help with performance problems
Date: Fri, 23 Apr 2004 11:57:23 -0700
Message-ID: <40896713.2010703@syscor.com> (raw)
In-Reply-To: <200404231115.23198.josh@agliodbs.com>
References: <200404230931.17850.revoohc@sermonaudio.com>
	<4089561B.8010001@speedfc.com>
	<200404231353.17234.revoohc@sermonaudio.com>
	<200404231115.23198.josh@agliodbs.com>

Josh Berkus wrote:

>Chris,
>
>  
>
>>Sorry for the confusion here.  I can't run any sort of vacuum durin the day
>>due to performance hits.  However, I have run vacuums at night.  Several
>>nights a week I run a vacuumdb -f -z  on all of the clusters.  I can take
>>serveral hours to complete, but it does complete.
>>    
>>
>
>Well, here's your first problem:  since your FSM pages is low, and you're only 
>vacuuming once a day, you've got to have some serious table and index bloat.   
>SO you're going to need to do VACUUM FULL on all of your databases, and then 
>REINDEX on all of your indexes.
>
>After that, raise your max_fsm_pages to something useful, like 1,000,000.   Of 
>course, data on your real rate of updates would help more.
>
>If you're getting severe disk choke when you vacuum, you probably are I/O 
>bound.   You may want to try something which allows you to vacuum one table 
>at a time, either pg_autovacuum or a custom script.
>
>  
>
Tom and Josh recently gave me some help about setting the fsm settings 
which was quite useful. The full message is at
http://archives.postgresql.org/pgsql-performance/2004-04/msg00229.php
and the 'most interesting' posrtion was:

 Actually, since he's running 7.4, there's an even better way.  Do a
 "VACUUM VERBOSE" (full-database vacuum --- doesn't matter whether you
 ANALYZE or not).  At the end of the very voluminous output, you'll see
 something like


 INFO:  free space map: 240 relations, 490 pages stored; 4080 total pages needed
 DETAIL:  Allocated FSM size: 1000 relations + 20000 pages = 178 kB shared memory.


 Here, I would need max_fsm_relations = 240 and max_fsm_pages = 4080 to
 exactly cover the present freespace needs of my system.  I concur with
 the suggestion to bump that up a good deal, of course, but that gives
 you a real number to start from.


 The DETAIL part of the message shows my current settings (which are the
 defaults) and what the FSM is costing me in shared memory space.

Good luck
Ron






view thread (14+ messages)  latest in thread

Message-ID: <40896713.2010703@syscor.com>
Permalink:  ../40896713.2010703@syscor.com/
Also on:    postgresql.org/message-id/40896713.2010703@syscor.com

 · 

reply

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: rstpierre@syscor.com, revoohc@sermonaudio.com
  Subject: Re: Help with performance problems
  In-Reply-To: <40896713.2010703@syscor.com>

* 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