Horton

Re: Proposal: Save user's original authenticated identity for logging

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Jacob Champion <pchampion@vmware.com>
Cc: "magnus@hagander.net" <magnus@hagander.net>, "stark@mit.edu" <stark@mit.edu>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>, "sfrost@snowman.net" <sfrost@snowman.net>, "tgl@sss.pgh.pa.us" <tgl@sss.pgh.pa.us>
Date: 2021-03-23T05:21:53Z
Lists: pgsql-hackers 
On Mon, Mar 22, 2021 at 07:17:26PM +0000, Jacob Champion wrote:
> v8's test_access lost the in-order log search from v7; I've added it
> back in v9. The increased resistance to entropy seems worth the few
> extra lines. Thoughts?

I am not really sure that we need to bother about the ordering of the
entries here, as long as we check for all of them within the same
fragment of the log file, so I would just go down to the simplest
solution that I posted upthread that is enough to make sure that the
verbosity is protected.  That's what we do elsewhere, like with
command_checks_all() and such.
--
Michael

Commits

  1. Add missing $Test::Builder::Level settings

  2. Add some information about authenticated identity via log_connections

  3. Fix some issues with SSL and Kerberos tests

  4. Refactor all TAP test suites doing connection checks