Re: Impact of checkpointer during pg_upgrade

On Fri, Sep 08, 2023 at 09:14:59AM +0530, Amit Kapila wrote:
> On Fri, Sep 8, 2023 at 9:00 AM Zhijie Hou (Fujitsu)
> <houzj.fnst@fujitsu.com> wrote:
>>>  I
>>> mean that doing the latter is benefitial for the sake of any patch committed and
>>> as a long-term method to rely on.
>
> What is your worry here? Are you worried that unknowingly in the
> future we could add some other way to invalidate slots during upgrades
> that we won't be able to detect?

Exactly.  A safety belt would not hurt, especially if the belt added
is simple.  The idea of a backend side elog(ERROR) with
isBinaryUpgrade is tempting in the invalidation slot path.
--
Michael