On Wed, 5 Apr 2023 at 11:15, Andres Freund <andres@anarazel.de> wrote:
>
> The freebsd test that failed is running tests in parallel, against an existing
> cluster. In that case it's expected that there's some concurrency.
>
> Why does this cause your tests to fail? They're in separate databases, so the
> visibility effects of the concurrent tests should be somewhat limited.
Because I'm checking that relfrozenxid was updated but any concurrent
transactions even in other databases hold back the xmin.
Honestly I'm glad I wrote the test because it was hard to know whether
my code was doing anything at all without it (and it wasn't in the
first cut...) But I don't think there's much value in having it be in
the regression suite. We don't generally write tests to ensure that a
specific internal implementation behaves in the specific way it was
written to.
--
greg