Re: pgsql: Remove secondary checkpoint

On Wed, Nov 8, 2017 at 4:43 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Andres Freund <andres(at)anarazel(dot)de> writes:
>> I think you misunderstand my point - I'm saying that pg_resetxlog should
>> be able to force the use of older checkpoints, basically as a fallback
>> to cases where the previous approach might actually have worked, not
>> that it needs to work across format changes.
>
> That seems like a completely separate feature --- and one of dubious
> value, frankly. The further back you go, the less likely it'd be
> to work.

Because the less guarantees you would have to reach a consistent point
with a non-corrupted instance. I think that Andres' idea here are
worth debating though. Why not just spawning a new thread on the
matter and summarize what you have in mind?
--
Michael