fedora-server
LOGS
17:00:20 <pboy> #startmeeting fedora-server
17:00:20 <zodbot> Meeting started Wed Oct 19 17:00:20 2022 UTC.
17:00:20 <zodbot> This meeting is logged and archived in a public location.
17:00:20 <zodbot> The chair is pboy. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
17:00:20 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:20 <zodbot> The meeting name has been set to 'fedora-server'
17:00:28 <pboy> #topic Welcome / roll call
17:00:35 <pboy> Welcome to our Server WG IRC meeting today!
17:00:42 <pboy> „Same procedure as every year“ We'll give a few minutes for folks to show up
17:00:49 <pboy> I’ll post the agenda in a few minutes.
17:06:13 <cooltshirtguy> hello
17:06:34 <pboy> welcome cooltshirtguy!
17:07:06 <pboy> A bit lonely here today.
17:07:13 <davide> .hello dcavalca
17:07:14 <eseyman> hello, all
17:07:14 <zodbot> davide: dcavalca 'Davide Cavalca' <dcavalca@fb.com>
17:07:20 <eseyman> .hello2 eseyman
17:07:21 <zodbot> eseyman: eseyman 'Emmanuel Seyman' <emmanuel@seyman.fr>
17:07:53 <cooltshirtguy> .hello2 cooltshirtguy
17:07:53 <zodbot> cooltshirtguy: cooltshirtguy 'Jason Beard' <jas_beard@hotmail.com>
17:07:56 <pboy> Welcome together!
17:08:27 <salimma> .hi
17:08:29 <zodbot> salimma: salimma 'Michel Alexandre Salim' <michel@michel-slm.name>
17:09:36 <pboy> Welcome salimma! I have missed you already
17:10:24 <salimma> thanks - been swamped with so many things unfortunately
17:10:28 <pboy> OK. We have achieved our quorum and can start.
17:10:39 <pboy> #topic Agenda
17:10:47 <pboy> #link https://pagure.io/fedora-server/report/Meeting
17:10:54 <pboy> #info 1. Follow up actions
17:11:02 <pboy> #info  2. Release 37 Beta testing
17:11:10 <pboy> #info  3. Documentation update Release 37
17:11:18 <pboy> #info  4.  Server critical path definition proposal
17:11:24 <pboy> #info 5, Open Floor
17:11:33 <pboy> Any additional topic / issue / comment ?
17:11:53 <eseyman> I'm good
17:12:00 <pboy> OK.
17:12:07 <pboy> #topic  1. Follow up actions
17:12:17 <pboy> We had: eseyman and pboy will try hard to replicate and check the certificate issue in the next 2 weeks
17:12:23 <pboy> Unfortunately, I was too busy with f37 beta issues.  eseyman? You too, I suppose?
17:12:37 <eseyman> yes, same here
17:12:56 <pboy> #info eseyman and pboy will try hard to replicate and check the certificate issue in the next 2 weeks ongoing
17:13:04 <eseyman> the good news is that I dropped quite a few TODOs over the last few days
17:13:17 <eseyman> so I'll be more available starting tonight
17:13:42 <pboy> yes, same here. servervm is good, I think release testing is done.
17:14:00 <pboy> #action eseyman and pboy will keep trying hard to replicate and check the certificate issue in the next 2 weeks ongoing
17:14:12 <pboy> As fas as I know, no additional open action.
17:14:19 <pboy> Does anyone have anything to add here?
17:15:02 <pboy> OK. next topic
17:15:10 <pboy> #topic 2. Release 37 Beta testing
17:15:19 <pboy> #link https://pagure.io/fedora-server/issue/91
17:15:32 <pboy> #info Fedora F37 rc 1.2 is available. since yesterday
17:15:40 <pboy> #link  https://fedoraproject.org/wiki/Test_Results:Fedora_37_RC_1.2_Server
17:15:55 <pboy> Current Status as fafr as I know:
17:16:02 <pboy> #info Installation issue with GPT and software RAID resolved. (issue #87)
17:16:10 <pboy> #info Old systemd-nspawn related bugs are basically resolved (bugzilla 1900888,  1900869), but not yet completely
17:16:20 <pboy> Our servervm works fine.
17:16:21 <eseyman> ah, that is nice to hear
17:16:32 <pboy> However, it is unpleasant that a SELinux systemd AVC message immediately before the first login spoils the impression.
17:16:59 <pboy> Harmless, but unattracive
17:17:07 <pboy> There were some bugs with Arm SBC, but they were fixed at the last minute.
17:17:30 <pboy> My impression, the server release is fine.
17:17:53 <pboy> What did you find so far?
17:19:41 <pboy> It comes just into my mind:
17:19:45 <eseyman> No Arm SBC here, I'm afraid
17:19:48 <pboy> We have still the problem, that network install and DVD install result in a slightly different installation.
17:20:31 <pboy> Re Arm: I'v just 3 of them, all Rockchip based-
17:20:40 <eseyman> I'm not sure that's a problem
17:21:09 <pboy> eseyman install difference or Rockchip?
17:21:32 <cooltshirtguy> I have some pi3/4. pi4 install went fine
17:21:50 <pboy> cooltshirtguy. Good to know
17:21:52 <eseyman> the fact that the installs are different
17:22:21 <eseyman> this may be just me but I've always thought of them as different things
17:22:44 <pboy> eseyman yes,the difference is small. E.g. in dvd install vim is still default. I don't mind. :-)
17:22:49 <eseyman> granted, we could minimize the different if we thought that was a useful goal
17:24:19 <pboy> I think, most important is to review the overall content of the installation media. The composition is quite old.
17:26:08 <pboy> Well, to finally sum it up, we all have a good impression of the current F37 Server, Nobody found a miss function. We are content with the release.
17:27:19 <eseyman> agreed
17:27:25 <pboy> #agreed  Server Edition rc 1.2 is ready for release.
17:27:44 <pboy> next topic
17:27:54 <pboy> #topic 3. Documentation update Release 37
17:28:04 <pboy> #link https://pagure.io/fedora-server/issue/94
17:28:13 <pboy> Current status, we have some docs updated and ready to publish
17:28:25 <pboy> I have to ad the Cockpit related part of our serverVM doku. So, if someone could check.
17:29:31 <pboy> The most important installation articles are reviewed and updated, I think.
17:29:38 <pboy> But overall, it seems to me that our docs are still a little behind.
17:30:02 <pboy> The eternal problem of documentatin.
17:31:38 <pboy> Any idea how we can improve it in the long run?
17:32:27 <eseyman> it would be nice to have more people
17:32:41 <eseyman> but documentation has never been very appealing to hackers
17:32:49 <cooltshirtguy> haha
17:33:55 <pboy> I have an idea if at the beginning of the next beta we determine 5 items that need to be updated most urgently.
17:34:26 <pboy> And we try to get it done to revise those 5.
17:34:43 <cooltshirtguy> interesting idea
17:35:01 <pboy> That might be more realistic than trying to tackle the whole doc.
17:35:32 <eseyman> I still think there's something to be said to having a documentation day in #fedora-server
17:35:57 <eseyman> writing in a group setting is more stimulating than doing it solo
17:36:21 <pboy> Yeah,  that's  a good idea!
17:36:45 <pboy> I think there are positive experiences for that already.
17:37:21 <pboy> Maybe, we combine the 2 ideas?
17:37:50 <cooltshirtguy> agreed. both ideas
17:37:54 <eseyman> +1
17:38:54 <pboy> #agreed For next release we determine the 5 articles mostly in need of an freview and work on them on a documentation day in #fedora-server
17:39:07 <pboy> #undo
17:39:07 <zodbot> Removing item from minutes: AGREED by pboy at 17:38:54 : For next release we determine the 5 articles mostly in need of an freview and work on them on a documentation day in #fedora-server
17:39:37 <pboy> #agreed For next release we determine the 5 articles mostly in need of an review and work on them on a documentation day in #fedora-server
17:39:50 <pboy> This time without a typo. :-)
17:40:18 <pboy> OK, next topic
17:40:33 <pboy> #topic 4. Server critical path definition proposal
17:40:41 <pboy> No issue yet.
17:40:52 <pboy> #info Mailing List: https://lists.fedoraproject.org/archives/list/server@lists.fedoraproject.org/thread/EVNJXDHQJBDKDXI5DXA37Q7I66EZNOI4/
17:41:13 <pboy> I must admit, I'm a bit lost here.
17:43:38 <pboy> adamw Are you online, can you help us here?
17:44:37 <pboy> copperi  If that is the question, it is not too difficult.
17:45:09 <pboy> We have a reviewed technical specification that can serve as a guidance
17:45:14 <adamw> what's up
17:45:22 <pboy> Welcome!
17:45:29 <adamw> i replied to martin's mail on the list
17:45:53 <pboy> Yes, you did. Nevertheless,I'm a bit lost.
17:46:07 <pboy> What exactly can we do?
17:46:15 <adamw> the primary intent here is to have the updates gated on openqa tests. but the only mechanism we really have to do that is putting them in the critical path. i explained all this in the devel@ discussion.
17:47:03 <adamw> we can leave things as they are, and updates won't be gated on the openqa tests and may break freeipa, cockpit or postgres without us noticing.
17:47:21 <adamw> or we can do this, and then we will gate updates that break those things, but the updates will have higher requirements to be pushed stable.
17:47:38 <adamw> or else someone can do all the work to create something that's like critpath but not critpath, but i'm not super inclined to do that.
17:48:16 <adamw> (or we can use gating.yaml files and scheduler allowlisting, but again, i already posted to devel@ and in my reply to martin why i don't love that idea)
17:48:58 <adamw> we could look at having openqa +1 updates that pass its tests, but that's kind of a radical step, i guess, and i'd have to see how supportable it is.
17:49:37 <adamw> the problem is it doesn't really test everything. kernel is in the critpath, but openqa's tests will only tell you if the kernel broke anything with booting a VM. if the kernel updates breaks every nvidia machine in the world, openqa isn't gonna find out.
17:50:46 <adamw> i mean...i can look again at the 'create a parallel path' thing if people really want that.
17:51:26 <pboy> Well, I'm not an exxpert here.  I think to have the (most) important updates gated is the best way. I think, that's your original idea, isnn't it?
17:51:41 <pboy> And we have to determine additional manual test?
17:53:32 <pboy> My first question would be, the current tests seem a bit "narrow" to me.
17:53:40 <pboy> Server is more than Postgres, IPA and Cockpit.
17:54:10 <pboy> Virtualization, Container, additional services than just Postgres come into mind
17:54:32 <pboy> But maybe, that's too much?
17:55:07 <eseyman> I'm guessing that someone has to write those tests
17:56:39 <pboy> yeah, that's a 'little' follow up
17:57:25 <eseyman> my gut instinct is that gating updates does more good than harm
17:57:36 <eseyman> but I'm not thrilled about it
17:58:49 <pboy> Yes, at the moment the karma process seems a bit random and unsystematic to me.
17:59:29 <pboy> But, unfortunately, our time is up.
17:59:48 <pboy> We should continue the mailing list discussion.
18:00:17 <adamw> we do have tests for podman also
18:00:20 <eseyman> I'm going to read the devel@ discussion, I missed it the first time round
18:00:58 <adamw> basically, the issue here is that the most convenient way we have to get gating done comes with the possibly-undesirable side effect of raising the push requirements.
18:01:27 <adamw> the decision to be made is whether those side effects really are undesirable (maybe most people do want higher requirements for these critical packages?), and if so, are they undesirable enough to be worth the work of coming up with an alternative way to get gating done.
18:01:36 <adamw> i agree we can continue it on the mailing list
18:02:49 <pboy> OK. let's continue on the list! I learned something from our discussion here  and know better what to decide!
18:03:16 <pboy> #topic 5. Open Floor
18:03:30 <pboy> Anything here to discuss?? (very short)
18:03:37 <eseyman> one short thing
18:03:42 <pboy> go on
18:04:08 <eseyman> I will be at the Open Source Experience event in Paris on the 8th and 9th of november
18:04:36 <eseyman> I'll be hosting the Fedora booth promoting the distribution
18:05:08 <eseyman> I intend to visit the open source companies that have booths there and ask if they're interested in supporting Fedora
18:05:11 <eseyman> EOF
18:05:56 <pboy> Good idea!  Maybe, I should pay Paris a visit.  :-)
18:07:03 <pboy> By the way, I made contact to an SBC maker. They will add Fedora to their recommended list, if we want to.
18:07:42 <pboy> Oh, it's (too) late now.
18:08:06 <pboy> Let's close. Or is there something else?
18:08:12 <cooltshirtguy> nope
18:08:21 <eseyman> don't think so
18:08:30 <eseyman> see you on the mailing lists, folks
18:08:40 <pboy> #endmeeting