Horton

Re: [HACKERS] VACUUM and ANALYZE disagreeing on what reltuples means

Tomas Vondra <tomas.vondra@2ndquadrant.com>

View thread View on postgr.es
From: Tomas Vondra <tomas.vondra@2ndquadrant.com>
To: Stephen Frost <sfrost@snowman.net>, Michael Paquier <michael.paquier@gmail.com>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Haribabu Kommi <kommi.haribabu@gmail.com>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2018-01-08T18:34:59Z
Lists: pgsql-hackers 
On 01/05/2018 05:25 AM, Stephen Frost wrote:
> Tomas,
> 
> * Michael Paquier (michael.paquier@gmail.com) wrote:
>> On Sun, Nov 19, 2017 at 6:40 AM, Tomas Vondra
>> <tomas.vondra@2ndquadrant.com> wrote:
>>> Thanks for looking into this. I agree your patch is more consistent and
>>> generally cleaner.
>>
>> This is classified as a bug fix, and is marked as waiting on author. I
>> am moving it to next CF as work continues.
> 
> This is still in waiting-on-author state; it'd be great to get your 
> thoughts on where this patch is and what the next steps are to move
> it forward. If you need additional feedback or there needs to be
> more discussion (perhaps with Tom) then maybe this should be in
> needs-review instead, but it'd be great if you could review the
> discussion and patch again and at least provide an update on it (last
> update from you was in November).
> 

Hmmm, I'm not sure, TBH.

As I already mentioned, Tom's updated patch is better than what I posted
initially, and I agree with his approach to the remaining issues he
pointed out. But I somehow assumed that he's already looking into that.

Tom, do you plan to look into this patch soon, or should I?

regards

-- 
Tomas Vondra                  http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

Commits

  1. Sync up our various ways of estimating pg_class.reltuples.

  2. Fix tuple counting in SP-GiST index build.

  3. Fix errors in contrib/bloom index build.

  4. When updating reltuples after ANALYZE, just extrapolate from our sample.

  5. Avoid holding AutovacuumScheduleLock while rechecking table statistics.