Re: Improve checks for GUC recovery_target_xid

On Thu, Mar 5, 2026 at 10:32 PM David Steele <david(at)pgbackrest(dot)org> wrote:
>
> On 3/5/26 19:42, Fujii Masao wrote:
> > On Thu, Mar 5, 2026 at 5:15 PM David Steele <david(at)pgbackrest(dot)org> wrote:
> >>
> >> On 3/5/26 12:03, Fujii Masao wrote:
> >>> On Thu, Mar 5, 2026 at 1:21 PM David Steele <david(at)pgbackrest(dot)org> wrote:
> >>>> The prior standby is not running because of the invalid config. I
> >>>> figured it was better to start clean but when I update the
> >>>> recovery_target_timeline tests I was planning to use the same standby
> >>>> for all the new tests.
> >>>
> >>> Alternatively, we can use $node_primary, since ALTER SYSTEM SET with
> >>> invalid recovery_target_timeline or recovery_target_xid does not
> >>> affect the primary.
> >>
> >> Well, as it turns out I was using the primary after all because I copied
> >> your example and forgot to update the host. Seems weird to set these
> >> GUCs on the primary but as long as we get the expected errors I don't
> >> suppose it matters.
> >
> > Thanks for updating the patch! I've pushed the patch.
>
> Excellent, thank you!
>
> Attached are the test changes for recovery_target_timeline. I can start
> a new thread and add it to the next CF if you like, but since it is just
> test changes maybe we can fast track it.

Yes, let's discuss and review the patch in this thread.

Thanks for the patch! It looks good to me. Barring any objections, I
will commit it.

Regards,

--
Fujii Masao