Re: Vacuum ERRORs out considering freezing dead tuples from before OldestXmin

Michael Paquier <michael@paquier.xyz>

From: Michael Paquier <michael@paquier.xyz>

To: Andres Freund <andres@anarazel.de>

Cc: Robert Haas <robertmhaas@gmail.com>, Tom Lane <tgl@sss.pgh.pa.us>, Melanie Plageman <melanieplageman@gmail.com>, Peter Geoghegan <pg@bowt.ie>, Heikki Linnakangas <hlinnaka@iki.fi>, Pg Hackers <pgsql-hackers@postgresql.org>, Noah Misch <noah@leadboat.com>

Date: 2024-07-23T00:25:04Z

Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

Test that vacuum removes tuples older than OldestXmin
- 2c0bc4765741 17.6 landed
- 303ba0573ce6 18.0 landed
- 80c34692e8e6 17.0 landed
- aa607980aee0 18.0 landed
Lower minimum maintenance_work_mem to 64kB
- 2eda3df9ad53 17.0 landed
- bbf668d66fbf 18.0 landed
Add accidentally omitted test to meson build file
- 9d198f4d3e3b 16.4 landed
Use DELETE instead of UPDATE to speed up vacuum test
- 924a08b76f5d 14.13 landed
- 9744fe24118b 15.8 landed
- 571e0ee40ebd 16.4 landed
Revert "Test that vacuum removes tuples older than OldestXmin"
- efcbb76efe40 18.0 landed
- 1a3e90948b50 17.0 landed
Ensure vacuum removes all visibly dead tuples older than OldestXmin
- fd4f12df5e46 17.0 landed
- 83c39a1f7f3f 18.0 landed

On Mon, Jul 22, 2024 at 10:37:17AM -0700, Andres Freund wrote:
> On 2024-07-22 13:16:49 -0400, Robert Haas wrote:
>> Laziness would have been not bothering to develop a TAP test for this
>> at all. Going to the trouble of creating one and not being able to
>> make it as fast or as stable as everybody would like is just being
>> human.
> 
> Yea, I think calling weeks of effort by Melanie lazy is, uhm, not kind.

FWIW, I'm really impressed by what she has achieved here by Melanie,
fixing a hard bug while hacking a crazily-complicated test to make it
reproducible.  This has an incredible amount of value in the long-run.

> It's not like somebody else had a great suggestion for how to do this in a
> better way either.

Sawada-san and John are the two ones in the best position to answer
that.  I'm not sure either how to force a second index pass, either.
Hmm.  An idea would be to manipulate the TIDStore stack under the
injection points switch?  This is run by the CI and some buildfarm
members already.
--
Michael