Re: Improve behavior of concurrent ANALYZE/VACUUM

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Michael Paquier <michael(at)paquier(dot)xyz>
Cc: Postgres hackers <pgsql-hackers(at)postgresql(dot)org>, "Bossart, Nathan" <bossartn(at)amazon(dot)com>, Kyotaro HORIGUCHI <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp>
Subject: Re: Improve behavior of concurrent ANALYZE/VACUUM
Date: 2018-08-14 15:26:29
Message-ID: CA+TgmobF86+TFYT5aj47TgtGF_x=ShRh4NtjLY64zgabSo14ig@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Sun, Aug 12, 2018 at 10:21 PM, Michael Paquier <michael(at)paquier(dot)xyz> wrote:
> In the previous thread, we discussed a couple of approaches, but I was
> not happy with any of those, hence I have been spending more time in
> getting to a solution which has no user-facing changes, and still solves
> the problems folks have been complaining about, and the result is the
> patch attached. The patch changes a couple of things regarding ACL
> checks, by simply centralizing the ownership checks into a single
> routine used by both ANALYZE and VACUUM. This routine is then used in
> two more places for manual ANALYZE and VACUUM:
> - When specifying directly one or more relations in the command, in
> expand_vacuum_rel().
> - When building the complete list of relations to work on in the case of
> a database-wide operation, in get_all_vacuum_rels().

I feel like you're not being very clear about exactly what this new
approach is. Sorry if I'm being dense.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Alexander Korotkov 2018-08-14 15:38:56 Re: [HACKERS] Bug in to_timestamp().
Previous Message Robert Haas 2018-08-14 15:10:27 Re: Tid scan improvements