Thread: Random make check failures

I'm seeing random failures in the parallel tests run by
"make check".  Sometimes (40% or so) all tests will succeed.
The rest of the time, I will get one or more failures.

The specific failures are connection failures:

! psql: connectDBStart() -- connect() failed: Connection refused
!       Is the postmaster running locally
!       and accepting connections on Unix socket '/tmp/.s.PGSQL.65432'?

I tried running pg_regress with --debug to see if it would
shed some light on the matter.  I didn't see anything in the
output indicating what was failing on the backend.

I'm running on Solaris 8, with the 2nd update installed.

I'm new to postgresql, could someone with some experience give
me a few pointers on tracking this one down?

Steve Nicolai

Steve Nicolai <snicolai@mac.com> writes:
> I'm seeing random failures in the parallel tests run by
> "make check".  Sometimes (40% or so) all tests will succeed.
> The rest of the time, I will get one or more failures.

> The specific failures are connection failures:

> ! psql: connectDBStart() -- connect() failed: Connection refused
> !       Is the postmaster running locally
> !       and accepting connections on Unix socket '/tmp/.s.PGSQL.65432'?

> I'm running on Solaris 8, with the 2nd update installed.

We've heard reports that this goes away if you tweak the pg_regress
script to use TCP rather than Unix-socket connections (look for the
bit that forces unix_sockets=no).  Nobody seems to have any idea why
Solaris has a problem with Unix-socket connections ...
        regards, tom lane