Re: Cause of moving-target FSM space-needed reports - Mailing list pgsql-hackers

From Kevin Brown
Subject Re: Cause of moving-target FSM space-needed reports
Date
Msg-id 20060921202043.GG11514@filer
Whole thread Raw
In response to Cause of moving-target FSM space-needed reports  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Tom Lane wrote:
> In fact, maybe vacuum should just throw a WARNING when it
> finds a single rel with more than MaxFSMPages pages with useful free
> space?

I fully agree with this.  This (in particular, how many useful free
pages a rel has) is information a good DBA can make very good use of,
and is needed in the case where it exceeds MaxFSMPages.

I would also suggest having vacuum verbose print an INFO message with
the rel's number of free pages for rels that don't exceed MaxFSMPages
(that number should be printed in the WARNING when the number exceeds
MaxFSMPages).

Are there other ways of getting this information from the system?  If
not, then I'd consider this proposed change important.


-- 
Kevin Brown                          kevin@sysexperts.com


pgsql-hackers by date:

Previous
From: "Jim C. Nasby"
Date:
Subject: Re: Release Notes: Major Changes in 8.2
Next
From: "Jonah H. Harris"
Date:
Subject: Re: pg_upgrade: downgradebility