Re: [HACKERS] UPDATE performance degradation (6.5.1) - Mailing list pgsql-hackers

From Tom Lane
Subject Re: [HACKERS] UPDATE performance degradation (6.5.1)
Date
Msg-id 7280.933087456@sss.pgh.pa.us
Whole thread Raw
In response to Re: [HACKERS] UPDATE performance degradation (6.5.1)  (Oleg Bartunov <oleg@sai.msu.su>)
List pgsql-hackers
Oleg Bartunov <oleg@sai.msu.su> writes:
> and after vacuum analyze:
> -rw-------   1 postgres users        8192 Jul 27 18:54 hits
> -rw-------   1 postgres users     1703936 Jul 27 18:54 hits_pkey
> Why hits_pkey is so big ? I have only 7 rows in the table.

Looks like vacuum reclaims the extra space in the table itself,
but does not do so with indexes.  Ugh.

I've thought for some time that vacuum ought to drop and rebuild
indexes instead of trying to update them.  This might be another
reason for doing that...
        regards, tom lane


pgsql-hackers by date:

Previous
From: Rich Shepard
Date:
Subject: Re: [ANNOUNCE] i386 RPMs available for v6.5.1
Next
From: Tom Lane
Date:
Subject: Bug tracking system policy