Skip site navigation (1) Skip section navigation (2)

Re: PG writes a lot to the disk

From: Bill Moran <wmoran(at)collaborativefusion(dot)com>
To: "Laurent Raufaste" <analogue(at)glop(dot)org>
Cc: pgsql-performance(at)postgresql(dot)org
Subject: Re: PG writes a lot to the disk
Date: 2008-03-21 13:54:37
Message-ID: 20080321095437.2a4a9ce6.wmoran@collaborativefusion.com (view raw or flat)
Thread:
Lists: pgsql-performance
In response to "Laurent Raufaste" <analogue(at)glop(dot)org>:

> 2008/3/21, Gregory Stark <stark(at)enterprisedb(dot)com>:
> >
> > Well that's true it does. But only once per row. So analyze would have set the
> >  bit on every row. You could do the same thing with something ligter like
> >  "select count(*) from <table>".
> 
> Well, the table has been analyzed, I did SELECT, PG write on the
> table. That's a fact.
> 
> But it's also true (I juste tested it) that every file of a table is
> modified by a SELECT COUNT.

The real question (to verify Tom's point) is does a _second_ SELECT count()
modify the table again?  If so, then something else is going on than
what Tom suggested.

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

wmoran(at)collaborativefusion(dot)com
Phone: 412-422-3463x4023

In response to

pgsql-performance by date

Next:From: Bruce MomjianDate: 2008-03-21 14:23:45
Subject: Re: PostgreSQL NetApp and NFS
Previous:From: Laurent RaufasteDate: 2008-03-21 13:45:28
Subject: Re: PG writes a lot to the disk

Privacy Policy | About PostgreSQL
Copyright © 1996-2014 The PostgreSQL Global Development Group