Re: Cygwin cleanup - Mailing list pgsql-hackers

From Thomas Munro
Subject Re: Cygwin cleanup
Date
Msg-id CA+hUKGL1kpPwt8DzU7cm_VMOi5mpSKSwCrdJAPDQfr_tg0E=NA@mail.gmail.com
Whole thread Raw
In response to Re: Cygwin cleanup  (Thomas Munro <thomas.munro@gmail.com>)
Responses Re: Cygwin cleanup
List pgsql-hackers
On Wed, Nov 9, 2022 at 2:04 PM Justin Pryzby <pryzby@telsasoft.com> wrote:
> +data_sync_retry = on

Sharing with the list some clues that Justin and I figured out about
what that part is doing.  Without it, you get failures like:

  PANIC:  could not open file "pg_logical/snapshots/0-14FE6B0.snap":
No such file or directory

That's been seen before:

  https://www.postgresql.org/message-id/flat/17827.1549866683%40sss.pgh.pa.us

That thread concluded that the operating system must have a non-atomic
rename(), ie a kernel bug.  I don't know why Cygwin would display that
behaviour and our native Windows build not; maybe timing, or maybe our
own open() and rename() wrappers for Windows do something important
differently than Cygwin's open() and rename().

On reflection, that seems a bit too flimsy to have in-tree without
more investigation, which I won't have time for myself, so I'm going
to withdraw this entry.



pgsql-hackers by date:

Previous
From: Dean Rasheed
Date:
Subject: Re: Improve tab completion for ALTER FUNCTION/PROCEDURE/ROUTINE
Next
From: Michail Nikolaev
Date:
Subject: Re: BUG: Postgres 14 + vacuum_defer_cleanup_age + FOR UPDATE + UPDATE