19:00:06 <clarkb> #startmeeting infra 19:00:06 <opendevmeet> Meeting started Tue Jul 23 19:00:06 2024 UTC and is due to finish in 60 minutes. The chair is clarkb. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:06 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:00:06 <opendevmeet> The meeting name has been set to 'infra' 19:00:17 <clarkb> #link https://lists.opendev.org/archives/list/service-discuss@lists.opendev.org/thread/OJTZIQRSLNYPUNSKXK7SNDACXHB675U7/ Our Agenda 19:00:21 <clarkb> #topic Announcements 19:00:49 <clarkb> As frickler mentioned in #opendev earlier today there are some new security patches for nova that we should be careful about and help ensure they merge in a timely manner 19:01:05 <clarkb> Was there anything else to announce? 19:02:26 <clarkb> sounds like no we can move on 19:02:40 <clarkb> I'm actually going to go a bit out of order to get some of these shorter topics out of the way first 19:02:48 <clarkb> #topic AFS Mirror Cleanups 19:03:06 <clarkb> I didn't have anything new here. Most of my focus since getting back has been on gitea and reviewing other people's changes 19:03:46 <clarkb> that said if you have time to review I think topic:drop-ubuntu-xenial has changes up still. CentOS 8 stream doesn't have any changes up right now I don't think but it has been largely removed from zuul on our side of things 19:03:54 <clarkb> * removed from zuul and afs 19:04:09 <clarkb> I think the next step for centos 8 stream is to remove the images compeltely from nodepool /me makes a note to look into doing that today 19:05:19 <clarkb> Anything else to add to this topic? 19:06:11 <clarkb> #topic Testing Rackspace's New Cloud Offering 19:06:45 <clarkb> So I reached out to others at the Foundation for some advice on how to proceed given my emails didn't get a response. Long story short it sounds like july was just a bad time to try and sync up with them so far and the suggestion is to try again in a week or so 19:06:57 <clarkb> So I'll be doing that but otherwise there isn't really anything new to report on this topic 19:07:48 <clarkb> #topic Drop x/* projects with config errors from zuul 19:08:20 <clarkb> I sent an announcement indicating we were going to do these sorts of cleanup. It seems to have generated some feedback on the chagne and we also got x/ospurge handed over to mnaser 19:08:26 <clarkb> #link https://lists.opendev.org/archives/list/service-announce@lists.opendev.org/thread/X5TM537UZFPYXF23H6TJFSKO7M6GVM2W/ Warning that was sent about these planned changes 19:08:32 <clarkb> #link https://review.opendev.org/c/openstack/project-config/+/923509 Proposed cleanup of inactive x/ projects 19:08:46 <frickler> I undropped ospurge from the patch earlier today, so the remainder should be good to go 19:08:53 <clarkb> the change for cleaning things up was recently updated with a current list of project to remove and I think we can proceed with merging that as soon as we get a second review on it 19:09:18 <clarkb> yup thank you for that update. We're past the date we said we would do cleanups too so if anyone else has a momeent to review that I think it is safe to merge 19:09:53 <frickler> yes, that should be no risk to the nova patches either 19:10:03 <clarkb> ++ 19:10:18 <fungi> agreed 19:10:40 <clarkb> #topic Gitea DB Doctoring 19:11:32 <clarkb> The gitea 1.22.1 upgrade was completed last friday. We did hit a minor snag around the formatting of a particular secret. We ended up letting gitea09 generate a new secret then propogated that out via ansible to work around the problem. Its actually a secret for a subsystem that we disable so not a huge deal for us but gitea refused to start without it being in the fomrat it 19:11:34 <clarkb> wants 19:11:45 <clarkb> But that means we are now in a position to run the doctor tool 19:11:53 <clarkb> #link https://etherpad.opendev.org/p/hcz4yMxUIsAWGgyoHKeZ proposed plan for db updates on each gitea node 19:12:25 <clarkb> I've gone ahead and drafted up a plan for the tasks to run on each node. Considering the nova patches are inflight and gitea is used to consume them (though not merge them) I've decided I'll start on this tomorrow morning at the earliest 19:12:42 <fungi> i'm around as early as you want to start on it tomorrow 19:12:50 <clarkb> that also works better for my schedule as it will allow me to dedicate basically all of wednesday to the process and ensure that giteas stay as in sync as possible 19:13:28 <clarkb> a quick review of the plan and posting any concerns to the doc or irc would be appreciated. I think fungi has indicated that it looks good but other feedback is welcome too 19:13:48 <fungi> yep, lgtm 19:14:07 <fungi> i couldn't think of anything else worth testing either 19:15:18 <clarkb> #topic Upgrading Old Servers 19:15:22 <clarkb> is tonyb awake yet? 19:16:02 <clarkb> Looks like no 19:16:38 <clarkb> I haven't managed to catch up on where the mediawiki effort is since I was out. I know the last statement was that things weren't quite ready for review yet but there is/was a held node to test thigns (which fungi has done and may be the most qualified to do so so thank you) 19:17:07 <clarkb> looks like the mediawiki change hasn't updated since we last met in this meeting 19:17:23 <clarkb> I suspect htat implies it isn't ready for review yet, but tonyb please do say otherwise if that is the case 19:18:04 <clarkb> oh it is also marked WIP so I think that is still the state of things 19:18:44 <clarkb> #topic Reconsider queue depth for integrated gate 19:19:10 <clarkb> I asked about the state of this item yesterday preparing for the meeting and frickler said there doesn't appear to be any interest in implementing the suggestion for early failure detection in openstack jobs 19:19:28 <clarkb> frickler: do you know if the idea has been proposed to the openstack qa team at all? 19:20:00 <clarkb> just thinking if that hasn't happend that would be a good next step just to be sure there isn't going to be any ofrward progress on this proposal before we fallback to other ideas like reducing the queue window floor 19:20:10 <frickler> well essentially there is not much of a qa team, only gmann and kopecmartin, but I didn't talk to them specifically 19:20:46 <clarkb> ok I can volunteer to bring this up over in #openstack-qa and can probably write an initial change for the tempest base job. I probably can't drive the broader adoption across all the jobs we may want to make the change to though but I can let them know 19:21:23 <clarkb> even with the security fixes the gate doesn't seem to be overwhelmed right now so we can probably take a little time to sort these details out 19:21:53 <frickler> well they aren't gating yet due to CI failures 19:22:15 <clarkb> ah the nova change in the gate is indeed unrelated 19:22:27 <frickler> but yeah, not too urgent currently I'd agree 19:23:06 <clarkb> ack I'll bring it up in irc over there tomorrow morning my time as I think that overlaps with kopecmartin and gmann both 19:23:55 <clarkb> #topic Open Discussion 19:24:05 <clarkb> That was all for the agenda I sent out. Definitely feels liek summer :) 19:24:29 <clarkb> when I looked at tonyb's changes for the mediawiki state I noticed that the ansible development testing update stack appears to be getting close 19:24:34 <clarkb> #link https://review.opendev.org/q/topic:%22ansible-devel%22 19:25:34 <clarkb> I think we are approaching service coordinator election time as well 19:25:48 <clarkb> I need to look at old emails and do some date math, but I'll try to have proposals for timing of that at our next meeting 19:28:16 <clarkb> Just trying to come up with any other info. I'll be out august 6-8 and then I'm traveling to the summit in korea from august 31 to september 5. 19:28:47 <clarkb> Other than that I expect to be generally around through the end of summer (I may attend FOSSY as well but tahts mostly over a weekend and I haven't decided yet) 19:29:29 <clarkb> I'll leave things open for a few more minutes but then we can have some time back today 19:30:48 <frickler> can we clean up https://zuul.opendev.org/t/openstack/config-errors?project=opendev%2Fbase-jobs&skip=0 or do we need to support older zuuls? 19:31:03 <frickler> I guess zuul-jobs needs to be more conservative 19:31:51 <clarkb> yes we can clean those items up since opendev/base-jobs is specific to our installation 19:32:11 <clarkb> frickler: is that something you wanted to do or should someone else write the chagne? 19:32:27 <clarkb> (it should be a straightforward mvoe of the playbook into the post-run list with the cleanup attribute set 19:32:56 <frickler> I would be fine with just reviewing the change if someone wants to propose it 19:33:14 <clarkb> ok I can write that up this afternoon. I'm going to look at the nodepool centos 8 stream stuff too 19:33:45 <frickler> mainly wasn't sure if it would be blocked by external dependencies like the rdo stuff 19:34:07 <clarkb> no I don't think it is since each zuul installation should curate their base jobs directly to meet their specific installation needs 19:35:19 <frickler> oh, that's actually zuul-base-jobs, not zuul-jobs https://zuul.opendev.org/t/openstack/config-errors?project=zuul%2Fzuul-base-jobs&skip=0 19:35:45 <frickler> so that's only for the zuul tenant in opendev and could also get updated then? 19:35:53 <clarkb> yes I think so 19:36:10 <clarkb> but I'll double check that one before pushing 19:36:27 <corvus> the zuul-base-jobs repo is like zuul-jobs, but for base jobs 19:36:37 <frickler> also some more in https://zuul.opendev.org/t/zuul/config-errors that might be worth looking at 19:36:42 <corvus> can be used directly, but most people don't; more likely used as a starting point. 19:37:04 <clarkb> oh I see 19:37:06 <corvus> regardless, zuul-base-jobs is not used by opendev 19:37:36 <clarkb> corvus: zuul-base-jobs does still produce the deprecation warning we must be loading enough config for it to generate the warning? 19:37:51 <clarkb> anyway I can update opendev/base-jobs today 19:38:20 <clarkb> Last call I'll end the meeting at 19:40 if we have nothing else 19:38:22 <corvus> clarkb: yeah, we just don't *use* it; i think it's shadowed 19:38:33 <clarkb> ack 19:39:36 <fungi> thanks clarkb! 19:40:05 <clarkb> and thank you to everyone working on OpenDev 19:40:13 <clarkb> We'll be back next week at the same time and location. 19:40:15 <clarkb> #endmeeting