Re: pgsql: Improve runtime and output of tests for replication slots checkp

On Thu, Jun 19, 2025 at 7:31 PM Alexander Korotkov <akorotkov@postgresql.org> wrote:

Improve runtime and output of tests for replication slots checkpointing.

The TAP tests that verify logical and physical replication slot behavior
during checkpoints (046_checkpoint_logical_slot.pl and
047_checkpoint_physical_slot.pl) inserted two batches of 2 million rows each,
generating approximately 520 MB of WAL.  On slow machines, or when compiled
with '-DRELCACHE_FORCE_RELEASE -DCATCACHE_FORCE_RELEASE', this caused the
tests to run for 8-9 minutes and occasionally time out, as seen on the
buildfarm animal prion.

Quite a few animals have started failing since this commit (for example [1]) . I haven't looked into why, but I suspect something is wrong.

stderr:
#   Failed test 'Logical slot still valid'
#   at /home/bf/bf-build/flaviventris/HEAD/pgsql/src/test/recovery/t/046_checkpoint_logical_slot.pl line 134.
#          got: 'death by signal at /home/bf/bf-build/flaviventris/HEAD/pgsql/src/test/perl/PostgreSQL/Test/Cluster.pm line 181.
# '
#     expected: ''

I see this failing on my fork's CI, so it seems like it could have been caught earlier?