Re: PATCH: standby crashed when replay block which truncated in standby but failed to truncate in master node

On Mon, Dec 16, 2019 at 12:22:18PM +0900, Fujii Masao wrote:
> > + Detection of WAL records having references to invalid pages during
> > + recovery causes <productname>PostgreSQL</productname> to report
> > + an error, aborting the recovery. Setting
> > Well, that's not really an error. This triggers a PANIC, aka crashes
> > the server. And in this case the actual problem is that you may not
> > be able to move on with recovery when restarting the server again,
> > except if luck is on your side because you would continuously face
> > it..
>
> So you're thinking that "report an error" should be changed to
> "trigger a PANIC"? Personally "report an error" sounds ok because
> PANIC is one of "error", I think. But if that misleads people,
> I will change the sentence.

In the context of a recovery, an ERROR is promoted to a FATAL, but
here are talking about something that bypasses the crash of the
server. So this could bring confusion. I think that the
documentation should be crystal clear about that, with two aspects
outlined when the parameter is disabled, somewhat like data_sync_retry
actually:
- A PANIC-level error is triggered.
- It crashes the cluster.
--
Michael