Re: Corruption during WAL replay

At Thu, 4 Mar 2021 22:37:23 +0500, Ibrar Ahmed <ibrar(dot)ahmad(at)gmail(dot)com> wrote in
> The regression is failing for this patch, do you mind look at that and send
> the updated patch?
>
> https://api.cirrus-ci.com/v1/task/6313174510075904/logs/test.log
>
> ...
> t/006_logical_decoding.pl ............ ok
> t/007_sync_rep.pl .................... ok
> Bailout called. Further testing stopped: system pg_ctl failed
> FAILED--Further testing stopped: system pg_ctl failed
> make[2]: *** [Makefile:19: check] Error 255
> make[1]: *** [Makefile:49: check-recovery-recurse] Error 2
> make: *** [GNUmakefile:71: check-world-src/test-recurse] Error 2
> ...

(I regret that I sent this as .patch file..)

Thaks for pointing that!

The patch assumed that CHKPT_START/COMPLETE barrier are exclusively
used each other, but MarkBufferDirtyHint which delays checkpoint start
is called in RelationTruncate while delaying checkpoint completion.
That is not a strange nor harmful behavior. I changed delayChkpt to a
bitmap integer from an enum so that both barrier are separately
triggered.

I'm not sure this is the way to go here, though. This fixes the issue
of a crash during RelationTruncate, but the issue of smgrtruncate
failure during RelationTruncate still remains (unless we treat that
failure as PANIC?).

regards.

--
Kyotaro Horiguchi
NTT Open Source Software Center