| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Zhijie Hou <houzhijie22(at)gmail(dot)com> |
| Cc: | Melanie Plageman <melanieplageman(at)gmail(dot)com>, Robert Haas <robertmhaas(at)gmail(dot)com>, Rogers Wang <rogers(dot)ww(at)qq(dot)com>, pgsql-hackers(at)lists(dot)postgresql(dot)org, Matthias van de Meent <boekewurm+postgres(at)gmail(dot)com>, Álvaro Herrera <alvherre(at)kurilemu(dot)de>, Andres Freund <andres(at)anarazel(dot)de> |
| Subject: | Re: WAL_LOG CREATE DATABASE strategy broken for non-standard page layouts |
| Date: | 2026-11-20 00:01:27 |
| Message-ID: | av-N17qyMg30Fx7v@paquier.xyz |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Tue, Nov 17, 2026 at 11:20:54PM +0800, Zhijie Hou wrote:
> I think we also need to increase shared_buffers (e.g., to 16MB), because
> $node->init(allows_streaming => 1) sets shared_buffers to 1MB internally,
> causing the pg_class pages to be evicted from shared_buffers before CREATE
> TABLE runs, so the test could not reach the old bug path.
If we care about this level of tuning, I'd suspect that just bumping
shared_buffers if the wal_consistency_checking switch is enabled
would be fine to do instead of having a buildfarm member that uses a
custom configuration file. wal_consistency_checking consumes a lot of
resources, so this would not hurt more. Your ideas around lifting
some of the masking of the VM pages would make that more valuable in
the long run, as least it seems so to me.
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Fujii Masao | 2026-11-20 00:14:18 | Re: A bug in recovery prefetcher, triggering asserts / FATAL: checkpoint request failed |
| Previous Message | Michael Paquier | 2026-11-19 23:52:03 | Re: Compile warnings under -DMEMORY_CONTEXT_CHECKING and no assertions |