Make new crash restart test a bit more robust.
authorAndres Freund <andres@anarazel.de>
Tue, 19 Sep 2017 17:32:51 +0000 (10:32 -0700)
committerAndres Freund <andres@anarazel.de>
Tue, 19 Sep 2017 17:39:52 +0000 (10:39 -0700)
commit1910353675bd149e1020b29c0fae02538fc358cd
tree201e1bc7b45f6137e41c79671e8070fb3d8b6649
parentd61f5bb7c444255b064a60df782907f7dddad61a
Make new crash restart test a bit more robust.

Add timeouts in case psql doesn't deliver the expected output, and try
to cause the monitoring psql to be fully connected to a backend.  This
isn't necessarily everything needed, but at least the timeouts should
reduce the pain for buildfarm owners.

Author: Andres Freund
Reported-By: Tom Lane, BF animals prairiedog and calliphoridae
Discussion: https://postgr.es/m/E1du6ZT-00043I-91@gemulon.postgresql.org
src/test/recovery/t/013_crash_restart.pl