Re: port conflicts when running tests concurrently on windows.

From: Andres Freund <andres(at)anarazel(dot)de>
To: pgsql-hackers(at)postgresql(dot)org
Subject: Re: port conflicts when running tests concurrently on windows.
Date: 2021-12-09 01:22:27
Message-ID: 20211209012227.cksdoli22exjjnxl@alap3.anarazel.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi,

On 2021-12-08 17:03:07 -0800, Andres Freund wrote:
> On 2021-12-08 16:36:14 -0800, Andres Freund wrote:
> > On 2021-12-08 14:45:50 -0800, Andres Freund wrote:
> > > Is it perhaps time to to use unix sockets on windows by default
> > > (i.e. PG_TEST_USE_UNIX_SOCKETS), at least when on a new enough windows?
> >
> > On its own PG_TEST_USE_UNIX_SOCKETS doesn't work at all on windows - it fails
> > trying to use /tmp/ as a socket directory. Using PG_REGRESS_SOCK_DIR fixes
> > that for PG_REGRESS. But the tap tests don't look at that :(.
>
> The tap failures in turn are caused by the Cluster.pm choosing a socket
> directory with backslashes. Those backslashes are then treated as an escape
> character both by guc.c and libpq.
>
> I think this can be addressed by something like
> [...]

That indeed seems to pass [1] where it previously failed [2]. There's another
failure that I haven't diagnosed yet, but it's independent of tcp vs unix sockets.

[1] https://cirrus-ci.com/task/5055530235330560?logs=ssl_test#L5
[2] https://cirrus-ci.com/task/5524596901281792?logs=ssl_test#L5

Greetings,

Andres Freund

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Peter Geoghegan 2021-12-09 01:40:37 Re: Why doesn't GiST VACUUM require a super-exclusive lock, like nbtree VACUUM?
Previous Message Zhenghua Lyu 2021-12-09 01:20:32 Re: Question on not-in and array-eq