Robert Haas <robertmhaas@gmail.com> writes:
> On Thu, Mar 31, 2022 at 10:52 AM Andrew Dunstan <andrew@dunslane.net> wrote:
>> We should probably fix the test though, so it doesn't require Unix
>> sockets. It should be possible, although I haven't looked yet to see how.
> Our mutual colleague Neha Sharma pointed out this email message to me:
> http://postgr.es/m/106926.1643842376@sss.pgh.pa.us
Ah, right.
> I actually don't understand why using pg_regress --auth-extra would
> fix it, or what that option does, or why we're even running pg_regress
> at all in PostgreSQL::Test::Cluster::init. I think it might be to fix
> this exact issue, but there's no SGML documentation for pg_regress,
I'm not volunteering to fix that, but this comment in pg_regress.c
is probably adequately illuminating:
* Rewrite pg_hba.conf and pg_ident.conf to use SSPI authentication. Permit
* the current OS user to authenticate as the bootstrap superuser and as any
* user named in a --create-role option.
This script is creating users manually rather than letting the TAP
infrastructure do it, which is an antipattern.
regards, tom lane