Re: postgres_fdw - make cached connection functions tests meaningful

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com>
Cc: PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: postgres_fdw - make cached connection functions tests meaningful
Date: 2021-07-14 22:18:05
Message-ID: 3571624.1626301085@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com> writes:
> While working on [1], I got to know that there is a new GUC
> debug_invalidate_system_caches_always that has been introduced in v14.
> It can be used to switch off cache invalidation in
> CLOBBER_CACHE_ALWAYS builds which makes cache sensitive tests stable.
> Using this GUC, it is quite possible to make cached connection
> management function tests more meaningful by returning original
> values(true/false, all the output columns) instead of SELECT 1.

Note that this needs an update in the wake of d68a00391.

More generally, though, I am not sure that I believe the premise of
this patch. AFAICS it's assuming that forcing debug_discard_caches
off guarantees zero cache flushes, which it does not. (If it could,
we wouldn't need the whole thing; the point of that variable is to
deterministically force flushes which would otherwise be
nondeterministic, not nonexistent.) Even in a contrib test that
seemingly has nothing else running, background activity such as
autovacuum could result in surprises. So I fear that what you have
got here is a patch that will work 99% of the time; which is not
good enough for the buildfarm.

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2021-07-14 23:29:04 Re: Add missing function abs (interval)
Previous Message Tom Lane 2021-07-14 21:42:10 Re: automatically generating node support functions