Too-short timeouts in test code - Mailing list pgsql-hackers

We've had more buildfarm failures due to hard-coded, short timeouts:

http://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=chipmunk&dt=2018-10-13%2021%3A06%3A58
  10s timeout while running pg_recvlogical
http://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=hornet&dt=2018-12-03%2005%3A52%3A12
  30s timeout while running pg_recvlogical
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=mandrill&dt=2018-11-30%2014%3A31%3A18
  60s timeout in isolationtester try_complete_step()

The 180s timeout in poll_query_until has been trouble-free since 2a0f89c
introduced it two years ago.  I plan to raise the timeouts in question to
180s, as attached.