Re: two memory-consuming postgres processes

From: "Kevin Grittner" <Kevin(dot)Grittner(at)wicourts(dot)gov>
To: "Alexy Khrabrov" <deliverable(at)gmail(dot)com>
Cc: "Greg Smith" <gsmith(at)gregsmith(dot)com>, <pgsql-performance(at)postgresql(dot)org>
Subject: Re: two memory-consuming postgres processes
Date: 2008-05-02 22:29:00
Message-ID: 481B4F3E.EE98.0025.0@wicourts.gov
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

>>> Alexy Khrabrov wrote:

> OK. I've cancelled all previous attempts at UPDATE and will now
> create some derived tables. See no changes in the previous huge
table
> -- the added column was completely empty. Dropped it. Should I
> vacuum just in case, or am I guaranteed not to have any extra rows
> since no UPDATE actually went through and none are showing?

The canceled attempts would have left dead space. If you have
autovacuum running, it probably made the space available for reuse,
but depending on exactly how you got to where you are, you may have
bloat. Personally, I would do a VACUUM ANALYZE VERBOSE and capture
the output. If bloat is too bad, you may want to CLUSTER the table
(if you have the free disk space for a temporary extra copy of the
table) or VACUUM FULL followed by REINDEX (if you don't have that much
free disk space).

Let us know if you need help interpreting the VERBOSE output.

-Kevin

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Matthew Wakeling 2008-05-03 09:25:28 Re: two memory-consuming postgres processes
Previous Message Tom Lane 2008-05-02 22:20:58 Re: Very slow INFORMATION_SCHEMA