09:02:17 <jakeyip> #startmeeting magnum
09:02:17 <opendevmeet> Meeting started Wed May  8 09:02:17 2024 UTC and is due to finish in 60 minutes.  The chair is jakeyip. Information about MeetBot at http://wiki.debian.org/MeetBot.
09:02:17 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
09:02:17 <opendevmeet> The meeting name has been set to 'magnum'
09:02:21 <jakeyip> #link https://etherpad.opendev.org/p/magnum-weekly-meeting
09:02:30 <dalees> o/
09:02:49 <jakeyip> #topic Roll Call]
09:02:53 <jakeyip> #topic Roll Call
09:02:55 <jakeyip> o/
09:02:55 <dalees> o/
09:03:00 <dalees> \o
09:03:10 <jakeyip> eager today dalees :)
09:03:30 <dalees> haha, I'm actually only here once not twice.
09:03:41 <jakeyip> #topic review
09:04:50 <jakeyip> I have gone thru some of the magnum-ui. I am not super good at -ui so I won't be able to catch mistakes. I'll just +1.
09:05:18 <jakeyip> I hope mnasiadka will help to +2 and you should be free to merge from there.
09:05:21 <jakeyip> is that ok?
09:05:38 <dalees> thanks jakeyip - we can fix them if they arise but still useful to read and run them to try it out once they all merge
09:06:27 <dalees> I'm ok with that; I feel good about the changes but don't promise 0 bugs. mnasiadka has raised a good point on one, so I split that part out to separate changeset.
09:07:03 <jakeyip> cool. good work there :) . my main aim is not block it.
09:07:25 <dalees> floating_ip vs master_lb_floating_ip_enabled has always confused me a bit, so maybe that's could affect others if they rely on previous behaviour.
09:08:41 <dalees> ricolin was pinged for the kubeconfig action button change, so perhaps he can help review/test there.
09:08:44 <jakeyip> I was under the impression it follows CLI but then again the CLI doesn't make sense
09:09:47 <jakeyip> I find the X-enabled / X-disabled redundant, I think I raised it before but was overruled
09:10:57 <dalees> consistency is more important, but perhaps that's not there either ;)
09:12:23 <dalees> anyway, thanks for the time on the ui reviews. Sorry there are lots, but hopefully things look+work better in that dept :)
09:12:26 <jakeyip> 'master' is a contriversial term in master_lb_floating_ip_enabled though :)
09:13:58 <jakeyip> in the same review you renamed the master -> control plane yet called this master_lb_floating_ip_enabled ... hah
09:14:28 <dalees> I'm only changing UI visible strings, not label names or the API at this stage.
09:15:08 <dalees> those probably need a spec and a deprecation period
09:16:11 <jakeyip> yeah I know...
09:17:07 <jakeyip> I don't think I have any questions at this point
09:17:20 <jakeyip> I'll bring them up in reviews or at next meeting if I have any
09:17:32 <dalees> ok
09:18:51 <jakeyip> I sent up some easy ones, please take a look
09:19:53 <opendevreview> Jake Yip proposed openstack/python-magnumclient master: Update Python runtime for 2024.2  https://review.opendev.org/c/openstack/python-magnumclient/+/918507
09:20:34 <jakeyip> any more reviews?
09:21:16 <jakeyip> https://review.opendev.org/c/openstack/python-magnumclient/+/918507 https://review.opendev.org/c/openstack/magnum/+/918506
09:21:44 <dalees> that's all immediate ones for now, I'll have a look at those others - if you could link into meeting agenda
09:23:16 <jakeyip> ok I want to talk about the versioning for capi though
09:23:21 <jakeyip> #topic capi
09:23:45 <jakeyip> any thoughts about tags / version for magnum-capi-helm and magnum-capi-helm charts?
09:25:01 <dalees> do you mean whether they follow the opendev cycles or their own releases?
09:25:52 <jakeyip> we have marked them as independent. just need to come up with our own versions now
09:26:20 <jakeyip> Add magnum-capi-helm deliverables as independent https://review.opendev.org/c/openstack/releases/+/910299
09:26:29 <dalees> right, cool
09:27:17 <mnasiadka> I'll have a look in the ui changes today
09:27:42 <mnasiadka> The plan for magnum-capi-helm is to add first release notes and prelude and release 1.0.0
09:27:47 <mnasiadka> and start bumping versions from there
09:27:54 <mnasiadka> I'll try to find time for that next week
09:28:54 <jakeyip> mnasiadka: can we tag 0.x for it for now? it will help us branch it and start testing patches on nectar
09:30:28 <jakeyip> FYI for Nectar we carry our own branches with patches. Some repos we branch off a tag and have our patches on them. Then when next tag is done we branch again and cherry-pick patches still needed
09:31:09 <jakeyip> I was planning on following stackhpc tags for now,
09:31:23 <jakeyip> but there is now an issue where magnum-capi-helm has diverged
09:33:10 <jakeyip> https://github.com/stackhpc/magnum-capi-helm vs https://github.com/openstack/magnum-capi-helm
09:33:14 <dalees> we follow a similar model, and we'll be looking to update our branch really soon. Looking forward to 1.0.0!
09:33:39 <jakeyip> dalees: do you now follow openstack/ or stackhpc/ ?
09:34:08 <mnasiadka> Sooner we do 1.0.0 and switch to that, the better - the stackhpc repo will be retired after that
09:34:49 <jakeyip> ok I think maybe not worth solving magnum-capi-helm for now, since 1.0.0 is almost here
09:34:59 <jakeyip> internally we just choose a tag and branch, no big deal
09:35:14 <jakeyip> how about magnum-capi-helm-charts?
09:35:17 <dalees> jakeyip: we havn't synced except to cherry pick in some fixes. so stackhpc for now, but we'll switch this upcoming sync, i think
09:36:22 <jakeyip> dalees: ok, may be helpful if you can make sure your cherry-picks are in openstack/magnum-capi-helm before we cut 1.0.0 :)
09:37:18 <dalees> yep, coming around to do that soon. cut 1.0.0 anyway, ours can come in soon - most are already there.
09:38:33 <jakeyip> ok. for magnum-capi-helm-charts I am working with infra to push
09:38:35 <jakeyip> https://review.opendev.org/c/openstack/project-config/+/918287
09:39:25 <jakeyip> dalees: what do you do for this repo?
09:40:34 <dalees> this is where most of our internal changes are (eg. to ignore keypairs, set defaults), so we do the same thing - branch and cherrypick our carried changes over that are still relevant.
09:41:33 <dalees> but again, we've not sync'd fully in a month or two. Coming back soon to catch up and do this in a couple weeks
09:41:37 <jakeyip> ok. I suspect I shouldn't wait for upstream then, will just branch in nectar
09:42:10 <jakeyip> dalees: did you create any tests? I was trying to get some test working from github actions -> zuul
09:45:37 <dalees> not fully automated on commit, we tag and that triggers a push to an OCI registry - after which we can build a template and run our cluster build and functionality tests.
09:46:06 <dalees> and ours are gitlab; so - yet another CI :D
09:47:03 <jakeyip> dalees: was thinking of lint or chart-testing (what stackhpc uses)
09:47:24 <jakeyip> gate basically
09:47:45 <jakeyip> no, check I mean
09:49:54 <dalees> nah, we just make MR to stackhpc's repo ;) will need them in the zuul tests!
09:50:39 <jakeyip> yeah ok then I guess I'll continue working on that then :)
09:51:58 <jakeyip> dalees: one more question - do you build images or packages?
09:53:19 <dalees> jakeyip: for what?
09:53:42 <jakeyip> for magnum
09:54:00 <dalees> container images, with openstack-loci
09:56:16 <jakeyip> ok. thanks.
09:56:20 <jakeyip> nothing else from me
09:56:30 <jakeyip> dalees / mnasiadka: you have anything else?
09:56:44 <mnasiadka> nope
09:56:45 <dalees> one question
09:57:30 <dalees> at some stage, I want to change the Magnum API response for `nodegroup show` to include server uuids of nodes in that group. Will this need a spec?
09:58:52 <mnasiadka> I assume you're planning to introduce a new API microversion - did we ever have specs for that in the past?
10:01:59 <dalees> yeah, that's probably the way to do it. I'll have a look and see, there aren't many specs created.
10:02:51 <jakeyip> I'm not a fan of paperwork, but spec may help in firming your thoughts and for discussion.
10:03:56 <dalees> ok - it'll also provide a place to explain "why"
10:04:04 <dalees> more than a commit message, anyway
10:04:30 <dalees> that's all then, thanks
10:04:32 <jakeyip> yeah
10:04:41 <jakeyip> cool
10:04:49 <jakeyip> #endmeeting