| From: | Thomas Munro <thomas(dot)munro(at)gmail(dot)com> |
|---|---|
| To: | Alexander Korotkov <aekorotkov(at)gmail(dot)com> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Kirill Reshke <reshkekirill(at)gmail(dot)com>, Andrey Borodin <x4mmm(at)yandex-team(dot)ru>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, Melanie Plageman <melanieplageman(at)gmail(dot)com>, Michael Paquier <michael(at)paquier(dot)xyz> |
| Subject: | Re: VM corruption on standby |
| Date: | 2025-08-22 01:41:44 |
| Message-ID: | CA+hUKGL87V9Tr3KHu2VsGEWZfoXY7kzCBfyDy1wpQhh7=UAn_Q@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Fri, Aug 22, 2025 at 10:27 AM Alexander Korotkov
<aekorotkov(at)gmail(dot)com> wrote:
> And let's retry it for v19.
+1
I'm hoping we can fix PM death handling soon, and then I assume this
can go straight back in without modification. CVs are an essential
low level synchronisation component that really should work in lots of
environments, but we need to straighten out some historical mistakes
and rough edges. Commit cfdf4dc4 removed open-coded exits to fix a
lot of bugs of omission, but it failed to fully consider all the
consequences of "composition", ie hiding that behaviour out of sight.
We really need a sort of postmasterless PANIC here, and I am happy to
work on that (see new thread), not least because it aligns better with
the behaviour of a multithreaded server. There, the answer is "what
other backends?" so I'd already been looking sideways at system states
including the lingering logger, lingering query execution and the
special but inherently flaky error reporting in a few spots.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Euler Taveira | 2025-08-22 01:59:14 | Re: event trigger support for PL/Python |
| Previous Message | Richard Guo | 2025-08-22 01:41:15 | Re: Remove unneeded cast in heap_xlog_lock. |