Re: [HACKERS] More detail on settings for pgavd?

Matthew T. O'Connor <matthew@zeut.net>

From: "Matthew T. O'Connor" <matthew@zeut.net>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Chester Kustarz <chester@arbor.net>, Josh Berkus <josh@agliodbs.com>, pgsql-performance@postgresql.org, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2003-11-20T23:18:27Z
Lists: pgsql-hackers, pgsql-performance
Tom Lane wrote:

>Chester Kustarz <chester@arbor.net> writes:
>  
>
>>vacuum is to reclaim dead tuples. this means it depends on update and
>>delete. analyze depends on data values/distribution. this means it depends on
>>insert, update, and delete. thus the dependencies are slightly different
>>between the 2 operations, an so you can come up with use-cases that
>>justify running either more frequently.
>>    
>>
>Agreed.
>  
>

And that is why pg_autovacuum looks at insert, update and delete when 
deciding to do an analyze, but only looks at update and delete when 
deciding to do a vacuum.  In addition, this is why pg_autovacuum was 
given knobs so that the vacuum and analyze thresholds can be set 
independently.

Matthew