2025-10-12T10:13:22  <acidsys> rebooting discourse
2025-10-12T10:14:33  <acidsys> and done
2025-10-12T13:08:22  *** krop is now known as Guest7385
2025-10-12T13:25:01  *** Guest7385 is now known as krop
2025-10-12T14:02:23  <acidsys>  you can now use seidr1.i.o.o as salt master to target minions regardless of location ( fyi cboltz)
2025-10-12T14:02:57  <cboltz> good to know, thanks
2025-10-12T15:36:01  *** teepee_ is now known as teepee
2025-10-12T20:40:37  <cboltz> acidsys: when you have a minute, can you please have a look at the test failures in !2605 ?
2025-10-12T20:41:05  <cboltz> Is this an issue with the test script, or did I do something wrong in the nftables file?
2025-10-12T20:42:13  <cboltz> (I also tried to use   set site = pillar['grains']['site']   hoping that it would be a bit more secure (since grains could be modified by the minion), and it gave me basically the same error)
2025-10-12T20:50:39  <matrix-o-o> <Emma [it/its]> hm, i wonder what it means when mx returns "Relay access denied"
2025-10-12T20:51:15  <matrix-o-o> <Emma [it/its]> not sure if this means my mail server is misconfigured, or the target address jut isnt set up to receive email lol
2025-10-12T20:51:22  <cboltz> that means it doesn't feal responsible for the domain
2025-10-12T20:51:53  <matrix-o-o> <Emma [it/its]> im guessing nothing is set up to be responsible for email to @code.o.o?
2025-10-12T20:53:09  <acidsys> hi cboltz, no "site" grain is defined. writing 'site: prg2' to /etc/salt/grains makes it work
2025-10-12T20:54:19  <acidsys> for this purpose it would be nice if it tested each tree with the matching site grain ...
2025-10-12T20:55:00  <cboltz> right, I had hoped that the test script cycles through pillar/id/ for each host using nftables ;-)
2025-10-12T20:55:41  <acidsys> test_nftables is not really aware of minions, it only iterates over subdirectories in salt/files/nftables/
2025-10-12T20:56:37  <cboltz> maybe I should simplify the MR to start with only the list of hosts so that we can see if the method works
2025-10-12T20:56:53  <cboltz> if it does, we can still move things that rely on grains
2025-10-12T20:58:31  <acidsys> Emma: https://progress.opensuse.org/issues/178477 volunteers welcome
2025-10-12T21:00:48  <acidsys> I think the backend is configured https://progress.opensuse.org/projects/opensuse-admin/repository/salt/revisions/5805055e948937bb23ac0f4cfc2807982af82f9e/diff but on mx{1,2}.i.o.o something is missing
2025-10-12T21:02:05  <matrix-o-o> <Emma [it/its]> ah, okay, so it's a known issue and not a me isue, good to know :)
2025-10-12T21:02:32  <acidsys> yep
2025-10-12T21:06:53  <acidsys> cboltz: thinking about a solution I wanted to pass the name of the tree to render_file() to then set grains from hosts[$name] but that does not work well for cluster trees because it would need to know when to add {1,2}
2025-10-12T21:07:11  <acidsys> I wonder if macro with argument for site would not be simpler
2025-10-12T21:14:01  <cboltz> we'll see ;-)
2025-10-12T21:14:26  <acidsys> if you want to play, you can easily run it outside of ci
2025-10-12T21:14:26  <cboltz> since my plan for this MR was a POC if symlinking in a common file works, I'll reduce it to the simple parts
2025-10-12T21:14:42  <cboltz> and if it works, we can do the more interesting ones
2025-10-12T21:18:20  <acidsys> sure
2025-10-12T21:45:31  <cboltz> the simplified variant looks good in the test output
2025-10-12T21:48:48  <cboltz> now the only question is if the symlink gets converted to a real file - if I understand the documentation right, the default is   keep_symlinks=False   so that we should get a normal file
2025-10-12T21:49:49  <acidsys> I just tested it to confirm the same
2025-10-12T22:12:11  <cboltz> will you do the deployment, or shall I do it tomorrow^W in ~20 hours?
2025-10-12T22:19:48  <acidsys> will do
2025-10-12T22:21:52  <acidsys> done
2025-10-12T22:22:00  <cboltz> thanks!