On Tue, Feb 14, 2023 at 5:42 AM Andres Freund <andres@anarazel.de> wrote:
>
> Hi,
>
> I'm working on rebasing [1], my patch to make relation extension scale
> better.
>
> As part of that I'd like to add tests for relation extension. To be able to
> test the bulk write strategy path, we need to have a few backends concurrently
> load > 16MB files.
>
> It seems pretty clear that doing that on all buildfarm machines wouldn't be
> nice / welcome. And it also seems likely that this won't be the last case
> where that'd be useful.
>
> So I'd like to add a 'large' class to PG_TEST_EXTRA, that we can use in tests
> that we only want to execute on machines with sufficient resources.
>
Oh, I was been thinking about a similar topic recently, but I was
unaware of PG_TEST_EXTRA [1]
I've observed suggested test cases get rejected as being overkill, or
because they would add precious seconds to the test execution. OTOH, I
felt such tests would still help gain some additional percentages from
the "code coverage" stats. The kind of tests I am thinking of don't
necessarily need a huge disk/CPU - but they just take longer to run
than anyone has wanted to burden the build-farm with.
~
Sorry for the thread interruption -- but I thought this might be the
right place to ask: What is the recommended way to deal with such
tests intended primarily for better code coverage?
I didn't see anything that looked pre-built for 'coverage'. Did I miss
something, or is it a case of just invent-your-own extra tests/values
for your own ad-hoc requirements?
e.g.
make check EXTRA_TESTS=extra_regress_for_coverage
make check-world PG_TEST_EXTRA='extra_tap_tests_for_coverage'
Thanks!
------
[1] https://www.postgresql.org/docs/devel/regress-run.html
Kind Regards,
Peter Smith.
Fujitsu Australia