Horton

Re: pgsql: Add 'basebackup_to_shell' contrib module.

Andres Freund <andres@anarazel.de>

View thread View on postgr.es
From: Andres Freund <andres@anarazel.de>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Robert Haas <robertmhaas@gmail.com>, Nathan Bossart <nathandbossart@gmail.com>, Andrew Dunstan <andrew@dunslane.net>, Thomas Munro <thomas.munro@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>, Magnus Hagander <magnus@hagander.net>
Date: 2022-03-31T05:25:00Z
Lists: pgsql-hackers 
Hi,

On 2022-03-31 00:08:00 -0400, Tom Lane wrote:
> Andres Freund <andres@anarazel.de> writes:
> > I assume SSPI doesn't work over unix sockets? Oh. Maybe it's not even that -
> > we only enable it when not using unix sockets:
> 
> Duh.  But can it work over unix sockets?

I wonder if we should go the other way and use unix sockets by default in the
tests. Even if CI windows could be made to use SSPI, it'd still be further
away from the environment most of us write tests in.

Afaics the only reason we use SSPI is to secure the tests, because they run
over tcp by default. But since we have unix socket support for windows now,
that shouldn't really be needed anymore.

The only animal that might not be new enough for it is hamerkop. I don't
really understand when windows features end up in which release.

Looking at 8f3ec75de40 it seems we just assume unix sockets are available, we
don't have a version / feature test (win32.h just defines
HAVE_STRUCT_SOCKADDR_UN).

Greetings,

Andres Freund

Commits

  1. Print information about type of test and subdirectory before running tests.

  2. In basebackup_to_shell tests, properly set up pg_hba.conf.

  3. Add .gitignore for basebackup_to_shell.

  4. basebackup_to_shell: Add TAP test.

  5. Document basebackup_to_shell.required_role.

  6. Add 'basebackup_to_shell' contrib module.