Re: Speedup of relation deletes during recovery

From: Andres Freund <andres(at)anarazel(dot)de>
To: Michael Paquier <michael(at)paquier(dot)xyz>
Cc: Fujii Masao <masao(dot)fujii(at)gmail(dot)com>, Kyotaro HORIGUCHI <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>, Robert Haas <robertmhaas(at)gmail(dot)com>
Subject: Re: Speedup of relation deletes during recovery
Date: 2018-06-21 17:41:29
Message-ID: 20180621174129.hogefyopje4xaznu@alap3.anarazel.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 2018-06-21 14:40:58 +0900, Michael Paquier wrote:
> On Wed, Jun 20, 2018 at 08:43:11PM -0700, Andres Freund wrote:
> > On 2018-06-18 11:13:47 -0700, Andres Freund wrote:
> >> We could do that - but add_to_unowned_list() is actually a bottleneck in
> >> other places during recovery too. We pretty much never (outside of
> >> dropping relations / databases) close opened relations during recovery -
> >> which is obviously problematic since nearly all of the entries are
> >> unowned. I've come to the conclusion that we should have a global
> >> variable that just disables adding anything to the global lists.
> >
> > On second thought: I think we should your approach in the back branches,
> > and something like I'm suggesting in master once open.
>
> +1. Let's also make sure that the removal of smgrdounlink and
> smgrdounlinkfork happens only on master.

FWIW, I'd just skip removing them. They seem useful functionality and
don't hurt. I don't see the point in mucking around with them.

Greetings,

Andres Freund

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2018-06-21 17:44:19 Re: Fast default stuff versus pg_upgrade
Previous Message Andres Freund 2018-06-21 17:40:08 Re: Fast default stuff versus pg_upgrade