Timeout failure in 019_replslot_limit.pl

Hi all,

Running the recovery tests in a parallel run, enough to bloat a
machine in resources, sometimes leads me to the following failure:
ok 19 - walsender termination logged
# poll_query_until timed out executing this query:
# SELECT wal_status FROM pg_replication_slots WHERE slot_name = 'rep3'

This corresponds to the following part of the test, where a WAL sender
is SIGSTOP'd and SIGCONT'd:
$node_primary3->poll_query_until('postgres',
        "SELECT wal_status FROM pg_replication_slots WHERE slot_name = 'rep3'",
        "lost")
  or die "timed out waiting for slot to be lost";

There is already a default timeout of 180s applied as of the default
of PostgresNode::poll_query_until(), so it seems to me that there
could be a different issue hiding here.

Thanks,
--
Michael