Horton

Re: Damage control for planner's get_actual_variable_endpoint() runaway

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Robert Haas <robertmhaas@gmail.com>
Cc: Andres Freund <andres@anarazel.de>, Simon Riggs <simon.riggs@enterprisedb.com>, pgsql-hackers@lists.postgresql.org, Jakub Wartak <jakub.wartak@enterprisedb.com>, Alvaro Herrera <alvherre@alvh.no-ip.org>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Date: 2022-11-21T23:17:52Z
Lists: pgsql-hackers 
Robert Haas <robertmhaas@gmail.com> writes:
> On Mon, Nov 21, 2022 at 5:15 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> I think we should content ourselves with improving the demonstrated
>> case, which is where we're forced to do a lot of heap fetches due
>> to lots of not-all-visible tuples.

> All right. I've been bitten by this problem enough that I'm a little
> gun-shy about accepting anything that doesn't feel like a 100%
> solution, but I admit that the scenario I described does seem a little
> bit far-fetched.
> I won't be completely shocked if somebody finds a way to hit it, though.

Well, if we see a case where the time is indeed spent completely
within the index AM, then we'll have to do something more or less
like what Simon sketched.  But I don't want to go there without
evidence that it's a live problem.  API warts are really hard to
get rid of once instituted.

			regards, tom lane

Commits

  1. YA attempt at taming worst-case behavior of get_actual_variable_range.

  2. Improve performance of get_actual_variable_range with recently-dead tuples.

  3. Use SnapshotDirty rather than an active snapshot to probe index endpoints.