| opendevreview | Merged openstack/election master: Update tc-election-summary.py: reflect results https://review.opendev.org/c/openstack/election/+/913918 | 01:49 |
|---|---|---|
| elodilles | hi, ttx suggested ( https://meetings.opendev.org/irclogs/%23openstack-release/%23openstack-release.2024-03-27.log.html#t2024-03-27T08:54:06 ) to let you know that we have 2 repositories where it turned out that a kind of a release blocker ~bug exists: watcher and vitrage is still not updated to the latest oslo.db release (caused issues in previous series as well). fortunately there are two | 10:06 |
| elodilles | volunteers who are trying to fix the issue, but we are close to final RC deadline | 10:06 |
| ttx | Both projects are a bit unmaintained and should be considered for removal in Dalmatian... We might need the TC to force-approve the fixes to fix Caracal release. https://review.opendev.org/c/openstack/watcher/+/914432 is ready for review | 10:46 |
| bauzas | ttx: about watcher, we could move the repo to be on the Nova project but I need first to discuss this with a few folks | 11:00 |
| bauzas | then I could approve some important changes from elodilles | 11:00 |
| bauzas | ttx: tc-members : could you please wait until next week to say whether watcher should be removed ? | 11:01 |
| frickler | bauzas: first step would be to mark the project (team) as inactive, this would automatically be resolved if the repos were moved | 11:04 |
| bauzas | frickler: I see | 11:04 |
| bauzas | I'd then prefer this | 11:04 |
| bauzas | if the watcher project becomes inactive, we could then say that the nova project could help it (or not, depending on a few questions I have) | 11:05 |
| bauzas | at least, Watcher is one of the ol' projects I remember from the 2015 days :) | 11:06 |
| bauzas | if this was happening to Blazar, I'd be super clear that I'd take it on me :) | 11:07 |
| frickler | that's an interesting consolidation path in general which I hadn't thought about before | 11:10 |
| frickler | ttx: the TC doesn't have any special reviewing powers in that context, but I guess they (incl. me with my TC hat on) could authorize the infra team to set some votes. maybe bauzas or some other people like stephenfin with some DB experience could add their review first, though | 11:23 |
| frickler | and then me with my infra hat on could exercise that | 11:24 |
| elodilles | frickler: that is how it worked in some rare case in the past. TC acknowledged that a patch needs to be force merged then infra team used their power to do it. | 12:44 |
| ttx | yeah... and since we are in a bit of a time crunch with RC deadline at the end of this week, it feels like we'll need that ASAP | 13:17 |
| ttx | jamespage has a patch ready for review for Watcher (https://review.opendev.org/c/openstack/watcher/+/914432) and noonedeadpunk is working on one for Vitrage, which hopefully should be finalized before end of week in time for release | 13:18 |
| frickler | I'd be willing to argue that this would be release-critical enough to even allow a last minute release. I still would like to see some technical review first, currently I'd be not sure whether the service would actually work outside of testing, too | 13:20 |
| noonedeadpunk | well, shouldn't tempest be covering that concern? | 13:22 |
| noonedeadpunk | as I guess there's no way to know unless some operator would report back | 13:23 |
| noonedeadpunk | but even then - it might be they just don't have some use-case which is not as well | 13:23 |
| jamespage | I would think so but I would appreciate review from a SME | 13:23 |
| noonedeadpunk | Like concern about service working outside if testing can be applied to literally any project anytime | 13:23 |
| frickler | well usually I would expect developers to have access to some actual deployments where such significant changes could be tested. but maybe we have to live with what we have here | 13:29 |
| frickler | so just waiting to see how much TC feedback we need and how to document that | 13:30 |
| JayF | It seems pretty straightforward to me: watcher does not have a PTL for 2024.2. the easiest path forward, would be for the interested parties to put someone forward as PTL (I would see consolidation into Nova as a similar level of project update that would have a similar effect) | 13:34 |
| frickler | can PTL appointments be approved faster than within a week? | 13:36 |
| frickler | otherwise this may not be able to solve the current release issue | 13:38 |
| frickler | this say standard rules apply, so IMO the answer is: no https://governance.openstack.org/tc/reference/house-rules.html#appointing-project-leaders-liaisons | 13:43 |
| JayF | I was hoping that when I checked the house rules it would be counted as a project update | 13:43 |
| JayF | It's still very early here in the US West so I am operating from mobile | 13:44 |
| JayF | I still think that's a good first step, to document intentions, and then we can take a more timely action in addition, like asking open Dev admins to give TC members the ability to merge patches | 13:45 |
| frickler | I was surprised to see you around this early at all, but you are already on DST, right | 13:45 |
| frickler | JayF: well if you want for the TC to make the +2 themselves, you can submit a change to gerrit ACLs, which could be merged at once. so far I was assuming for the TC to simply ask some infra-root to submit some vote with their super powers | 13:46 |
| fungi | yes, we've (gerrit admins) merged changes before at the tc's request | 13:47 |
| fungi | fwiw, we don't normally have approval permissions on those repositories either, but we have a straightforward process to temporarily elevate our permissions and then downgrade them again as soon as we're done | 13:49 |
| frickler | I also see this as two separate issues mostly. even if nobody steps up as PTL and we mark watcher as inactive, it is IMO now definitely too late to not release it. so we either get the potential fix merge or release it in a broken state | 13:51 |
| JayF | frickler: That makes sense, I agree with you now | 13:58 |
| JayF | I'm OK with either route: asking infra to merge things or updating ACLs | 13:58 |
| JayF | I think asking them to merge things is the most straightforward route then, yeah? | 13:59 |
| frickler | JayF: yes, that avoids the need to change gerrit config, you'd just have to say "make it so" ;) | 14:00 |
| JayF | I think doing the needful to get watcher releasable or removed from the release is something we should make-so :D | 14:01 |
| JayF | I am not sure I have the technical acumen to review a change to watcher, given I didn't know what it did until a few days ago. | 14:01 |
| JayF | So I'm happy to review a specific change and say "land it" based on trust and lack of obvious harm, but that's about all I have to give :D | 14:02 |
| frickler | JayF: yes, I don't have either, so best we can do is trust tox+tempest in this case | 14:02 |
| JayF | fungi: frickler: https://review.opendev.org/c/openstack/watcher/+/914432/2#message-d9f0f0b3017bfe8fb5dcf5153e4746eff16ade07 | 14:04 |
| JayF | I think that's probably what you needed from me? | 14:04 |
| frickler | JayF: fine for me, I can do it after the kolla meeting ends. note we'll also need a cherrypick to 2024.1, but I think we can deal with that in the same way | 14:05 |
| JayF | ack; I can put "and a backport to 2024.1" in that comment so I'm outta the happy path or just ping me with the new patch and I can repost | 14:06 |
| TheJulia | I took a look, and at a minimum level I think it is good, my worry, as we saw elsewhere, things might not actually shake out perfectly, but folks can open or try to fix bugs if they encounter such "fun" issues | 14:08 |
| JayF | It's going to work a heck of a lot better than it would without that patch :) | 14:09 |
| *** freyes_ is now known as freyes | 14:10 | |
| TheJulia | indeed | 14:10 |
| JayF | On a separate topic, the last couple of cycles, I've been on OpenInfra Live and presented on some project-wide bits with a TC hat on as well as Ironic with my PTL hat on. I'll be out of town for this years' OpenInfra Live, I'm not sure we have anything too specific to present this year (unlike timess past with SLURP and Unmaintained), but if someone has something in mind they may want to volunteer. | 14:13 |
| TheJulia | JayF: just to be on the same page, your asking for the TC portion, yes? Is the Ironic portion going to be covered by someone already? | 14:16 |
| JayF | rpittau is doing the presentation for Ironic; I'm saying for TC/project wide representation or announcements | 14:17 |
| TheJulia | perfect | 14:17 |
| JayF | like I said, I don't think there's anything really beefy to put in there that is coming straight to mind | 14:17 |
| bauzas | I could have done a TC presentation :D | 14:18 |
| bauzas | joking apart, JayF even if I'm not TC, I can add a slide telling some numbers | 14:19 |
| JayF | I think it's useful to have some project-wide perspective; I'm not terribly concerned with who does it :D | 14:20 |
| JayF | Maybe just put a slide together and link it here for feedback? | 14:20 |
| bauzas | I'm saying that because I voluteered for the Nova part | 14:20 |
| * TheJulia suspects people are trying to get JayF to take that day off as well :) | 14:20 | |
| JayF | Oh yeah, I'm completely onboard. It's an easy win | 14:20 |
| bauzas | so if nobody from TC steps up, I can take it | 14:20 |
| bauzas | but I'd indeed prefer some TC representation | 14:21 |
| JayF | TheJulia: lol I'm probably disconnecting from OFTC at 4pm on Friday and not turning back until the next Monday :D | 14:21 |
| TheJulia | JayF: in accordance with the prophecy | 14:21 |
| clarkb | frickler: fungi: if we're just talking about approving things (and not adding verified +2 and submitting) then it is super easy to also add people to groups (like the tc group could go in a watcher core group) and then others can hit the button to make it more official like | 14:58 |
| fungi | true, we can even just temporarily set the tc members group as included in watcher core temporarily | 15:00 |
| frickler | oh, right, I've already approved the patch for now, but it's not unlikely that other projects will require similar actions. not sure if we'd want to add all of the tc though or just tc-chair like for elections repo | 15:03 |
| frickler | JayF: jamespage: change merged, cherry-pick made: https://review.opendev.org/c/openstack/watcher/+/914451 | 15:15 |
| jamespage | thanks frickler | 15:16 |
| JayF | ty, commetned | 15:16 |
| gmann | elodilles: ttx: for vitrage if there is any we have noonedeadpunk maintaining it and he can help | 17:00 |
| gmann | frickler: bauzas JayF for watcher or any such project, it is not mandatory to mark project Inactive before retirement that is different route to retire project but if project is al dead and no maintianer or leader then we can directly start the retirement also | 17:02 |
| noonedeadpunk | I just slightly failed with the timeline due to plenty of other things :( | 17:02 |
| gmann | JayF: I can volunteer for TC presenting in openinfra live | 17:02 |
| bauzas | gmann: ack then I need some time to see if nova is interested in buying the repo | 17:03 |
| JayF | Thanks, just make sure you coordinate with bauzas since he also volutnteered | 17:03 |
| bauzas | JayF: I can leave it to gmann for sure | 17:03 |
| JayF | it's yours for free-fifty :D | 17:03 |
| JayF | hehe | 17:03 |
| bauzas | I'm bad at playing casino and I got an evidence yesterday so I'll pass | 17:04 |
| gmann | bauzas: sure. thanks. I can see note here too. all good on that. after all possibility gone then only we can discuss retirement https://etherpad.opendev.org/p/2024.2-leaderless#L72 | 17:05 |
| bauzas | if we agree next weeks on getting watcher under our umbrella (or I could just volunteer for acting PTL) then we would revisit that in 6 months | 17:07 |
| gmann | ack | 17:07 |
| bauzas | gmann: I'll try to come to a conclusion tomorrow hopefully | 17:08 |
| gmann | no hurrey, we have time, I think PTG is something where we might also try to discuss those mainly to communicate wider | 17:08 |
| gmann | JayF: I think we forgot but we should discuss it in PTG too, added this topic at the end of etherpad https://etherpad.opendev.org/p/apr2024-ptg-os-tc#124 | 17:10 |
| gmann | bauzas: on openinfra live, feel free to take it. I did not realize you volunteer for that and thought you were waiting for other TC members reply first. | 17:12 |
| bauzas | gmann: no, please take it if you can | 17:13 |
| bauzas | I'm not part of TC and I don't want to have more confusion to the problem we had | 17:14 |
| bauzas | to add* | 17:14 |
| gmann | bauzas: I do not think that can be confusion and surly you can cover it all. but let me know, I am ok with either | 17:15 |
| noonedeadpunk | gmann: out of the list - I'm just not in agreement with Freezer :) | 17:17 |
| noonedeadpunk | there was interest raised in ML from some individuals, and I hope/pray getting project back on track | 17:17 |
| noonedeadpunk | for 2024.2 | 17:17 |
| gmann | noonedeadpunk: I see the mail discusion but did not follow up those. | 17:17 |
| gmann | ++, feel free to add your comment in etherpad which will be helpful https://etherpad.opendev.org/p/2024.2-leaderless#L27 | 17:18 |
| noonedeadpunk | yeah, so folk stepped up to check on CI and keep it alive, I did a meeting with them to enroll into the process and we're kinda in contact now | 17:18 |
| noonedeadpunk | And scheduled the next meeting next week where I got 7 positive replies about joining it | 17:19 |
| gmann | cool, thanks. please let me know if any help needed from QA or tempest plugin side. I will be happy to do | 17:19 |
| frickler | noonedeadpunk: so maybe then you can discuss whether someone wants to volunteer as PTL or maybe multiple people want to share the duties as DPL team | 17:20 |
| noonedeadpunk | yup, already raised that | 17:20 |
| gmann | I keep pushing changes in freezer tempet plugin to keep testing and stable jobs up to date but there are other gate failure there https://review.opendev.org/q/project:openstack/freezer-tempest-plugin+status:open | 17:20 |
| noonedeadpunk | let's see outcome from meeting on Apr 2 | 17:21 |
| noonedeadpunk | yeah, it's failing on sqlachemy from what I saw | 17:21 |
| gmann | I will suggest DPL model may help them if any new members is little hesitant to take PTL but yeah please discuss with them | 17:21 |
| noonedeadpunk | ++ | 17:21 |
| noonedeadpunk | and I know some org that is currently considering their migration to openstack, and they were ready to get one of their devs on freezer in case of positive decision for migration (which would be clear closer to summer I guess). But will see. | 17:23 |
| gmann | k | 17:24 |
| noonedeadpunk | I guess main issue/blocker would be indeed to get someone who can land things, but we'll sort that out with PTL/DPL | 17:31 |
| bauzas | gmann: please take it | 17:48 |
| gmann | k | 17:49 |
| gmann | noonedeadpunk: I think we hanled same situation in mistral in past. it is doable where we see committed volunteer coming in community | 17:50 |
| gmann | *handled | 17:50 |
| fungi | i really wish we had some way of identifying when dpl liaisons for a project go stale/walk away. there's no dead-man's switch like for ptl elections | 19:38 |
| fungi | i face similar struggles with security reviewer groups for projects not getting updated by their teams/leaders, and ending up full of former contributors but nobody who's actually active on the project | 19:40 |
| JayF | fungi: I think that's going to be something important to discuss in the ptg items around activity | 19:40 |
| fungi | agreed, it's already on the pad | 19:40 |
| JayF | fungi: Honestly, I'm fairly convinced DPL is an antipattern at this point but I am not sure we'd have consensus to drop it | 19:40 |
| gmann | I do not think we need to drop but find a way to monitor the activeness of projects which can be similar to PTL model projects. if we see many of the inactive projects are triggered via their gate broken and release things. if we do some kind of periodic check of these on DPL model also then it will be fine | 21:24 |
| gmann | DPL just eliminate the election leaderless trigger for any inactive projects and rest all applies the same way to both the model. | 21:28 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!