Michael Paquier <mich...@paquier.xyz> writes: > sidewinder is still pissed of as of HEAD, pointing visibly to f912d7d > as the root cause: > https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=sidewinder&dt=2019-05-03%2021%3A45%3A00
Right, the deadlocks are expected when some previous session is slow about cleaning out its temp schema. The plan is to leave that in place till tomorrow to see if any *other* failure modes turn up. But it has to come out before we wrap the releases. I don't think we discussed exactly what "come out" means. My thought is to leave the test scripts in place (so they can be invoked manually with EXTRA_TESTS) but remove them from the schedule files. regards, tom lane