Hi,
On 2021-06-11 18:44:28 -0400, Tom Lane wrote:
> Tomas Vondra <tomas.vondra@enterprisedb.com> writes:
> > There's one caveat, though - for regular builds the slowdown is pretty
> > much eliminated. But with valgrind it's still considerably slower. For
> > postgres_fdw the "make check" used to take ~5 minutes for me, now it
> > takes >1h. And yes, this is entirely due to the new test case which is
> > generating / inserting 70k rows. So maybe the test case is not worth it
> > after all, and we should get rid of it.
>
> I bet the CLOBBER_CACHE animals won't like it much either.
>
> I suggest what we do is leave it in place for long enough to get
> a round of reports from those slow animals, and then (assuming
> those reports are positive) drop the test.
I just encountered this test because it doesn't succeed on a 32bit system with
address sanitizer enabled - it runs out of memory. At that point there are
"just" 29895 parameters parsed...
It's also the slowest step on skink (valgrind animal), taking nearly an hour.
I think two years later is long enough to have some confidence in this being
fixed?
Greetings,
Andres Freund