Thread: pg_ctrl: odd messages logged at startup -
[TEST REPORT]
[Release]: 9.0 RC1
[Test Type]: Install, initdb, first startup
[Test]: pg_ctrl startup of a newly-intdb'd cluster.
[Platform]: Solaris 10 SUN E450 Quad
[Parameters]: pg_ctrl startup of a newly-intdb'd cluster.
[Failure]: NO. Server starts perfectly. However,
[Results]:
2010-09-12 16:17:43 GMT - LOG: could not open directory "DIR": No such file or directory (Many repetitions)
2010-09-12 12:17:43 EDT - LOG: could not bind IPv6 socket: Cannot assign requested address
2010-09-12 12:17:43 EDT - HINT: Is another postmaster already running on port 5434? If not, wait a few seconds and retry..[Comments]: None of this a showstopper.
However, we have no 'DIR' in configuration file. What is this?
Is there a way to set 'localhost' listen to use only the IPv4 interface?
There is no postmaster running on 5434 - Could this message either be more informative or - what does it actually mean? We don't see the postmaster.pid file persisting after a shutdown...