18:00:07 #startmeeting tc 18:00:07 Meeting started Tue Nov 28 18:00:07 2023 UTC and is due to finish in 60 minutes. The chair is JayF. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:07 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:00:07 The meeting name has been set to 'tc' 18:00:11 #topic Roll Call 18:00:13 o/ 18:00:15 o/ 18:00:19 \o 18:00:24 Welcome to the weekly meeting of the OpenStack Technical Committee. A reminder that this meeting is held under the OpenInfra Code of Conduct available at https://openinfra.dev/legal/code-of-conduct. 18:00:27 Today's meeting agenda can be found at https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee. 18:00:32 There are no noted absenses in the agenda for today's meeting. 18:00:33 o/ 18:00:51 o/ 18:00:51 o/ 18:00:55 o/ 18:00:58 Going to wait until :05 or until all TC members have checked in 18:01:00 o/ 18:04:45 Aight, going to get started. We have 8 that is quorum. 18:04:52 #topic Follow up on tracked action items 18:05:02 o/ 18:05:08 9! 18:06:31 ping? 18:06:41 okay, sorry, wifi network went down locally, but I appear to still have wired connectivity 18:07:19 #info Rosmaita to propose amendment to unmaintained branch resolution allowing a single review group 18:07:24 #info Rosmaita to email mailing list about progress implementing unmaintained branch resolution 18:07:32 rosmaita: I grouped these together since they are related; do you have an update? 18:07:41 * rosmaita actually did all his action items for once 18:07:56 #link https://review.opendev.org/c/openstack/governance/+/900940 18:08:04 #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/ZYAZG43BLJJVXCYZVPYQX5733BYDVVNL/ 18:08:16 so, no response to the email, so i guess everyone is on board 18:08:17 #success rosmaita did all his action items 18:08:20 knikolla: Added success to Success page (https://wiki.openstack.org/wiki/Successes) 18:08:27 :) 18:08:54 That governance change appears to have wide consensus; please make time to review it if you haven't and I will take a pass at landing eligible governance patches this afternoon. 18:08:54 the resolution has achieved ripeness, and i think has enough votes to pass 18:09:12 amendment to resolution looks good to me, thanks rosmaita for preparing that 18:09:13 jsut added one more 18:09:21 \o/ 18:09:45 Is there anything further we should mention about unmaintained branches before moving on? I don't believe we have another agenda item for it. 18:09:48 ditto (and I managed to actually hit the right option this time) 18:09:58 What is our next step, assuming this governance change lands in the next hour or two 18:10:17 i think just to help the release team get this implemented 18:10:42 Is someone on TC doing or helping coordinate that work? 18:10:46 and i guess we need opendev infra team to set up the meta-acl for gerrit 18:11:15 that should be a patch to project-config 18:11:16 yes, if you can get me details on what you need i can propose a patch 18:11:35 great 18:11:37 Instead of hashing those details out here, can someone take the action to work with fungi on this? 18:11:39 it'll just be an edit to the existing gerrit/acls/openstack/meta.config acl file 18:11:40 And we can move on? 18:12:03 #action rosmaita work with fungi on the gerrit meta-acls 18:12:13 #undo 18:12:13 Removing item from minutes: #action rosmaita work with fungi on the gerrit meta-acls 18:12:16 #redo 18:12:24 #action rosmaita work with fungi on the gerrit meta-acls 18:12:34 I didn't think anyone but the person who started the meeting could do that :) 18:12:43 :D 18:12:44 Thanks for that 18:12:50 #topic Gate Health Check 18:12:53 How is the gate? 18:13:21 Things have been mostly calm on the Ironic side. Some failures that we tracked down to real issues in one of our libraries, contained to Ironic-related projects. 18:13:49 one failure there in novaclient fucntional job with neutron new rbac enabled 18:14:19 really not great on the nova side 18:14:24 we have enabled neutron rbac by default in devstack and that broke novaclient job, not sure why but it is disabled in that job and worked around 18:14:30 I'm at 14 rechecks for a patch that has been approved for weeks 18:15:16 many of the failures are related to volume tests, per usual, but some other instabilities have also crept up, but no giant smoking guns that I've seen 18:16:28 Is there anything at a TC level we can or should do? I would say can we raise awareness but if it's bad enough to need >a dozen rechecks; I imagine there's decent awareness. 18:16:50 well, tbh, I think most people are happy to recheck 20 times to get their patches in 18:16:51 I am sure those are not blind recheck 18:16:55 so awareness is maybe not enough 18:17:32 gmann: mine aren't for sure, except for yesterday when zuul wasn't loading log results, so it was hard to examine for root cause 18:17:34 i think we know what failing like volume tests and need someone to deep dive into those 18:17:52 dansmith: yeah 18:17:54 agree.. before last week, the cinder grenade phase was failing for me a *lot* 18:18:11 haven't seen that yet this week, but the resize vol-backed test has bitten me a few times 18:18:48 there was an issue with cinder, grenade and tooz, I added that as dedicated agenda item 18:18:51 the grenade seems to fail after it creates the third volume and then tries to ssh into the server 18:19:08 at least that's what i've seen a few times 18:19:23 but i don't know what's happening to the server 18:19:33 rosmaita: okay I haven't seen that, it's usually waiting for the volume to become in-use, but that was 2+ weeks ago 18:19:43 it' 18:19:54 it's meaning "what I was seeing" 18:20:16 anyway, agree with gmann that some deep dive is needed for sure 18:20:22 well, there is something bad happening during that phase 18:20:37 I'm pretty burned out on that stuff myself, as I spent months on it after I came back from holiday break this jan, when things were really bad 18:20:42 Unless folks in the TC want to dig this personally, I'd suggest putting some of these research results out onto the list and trying to recruit additional help. 18:20:47 so I hope someone else can really jump in while I'm out 18:21:05 We've gotta try to increase the number of people engaged with these kind of problems, and the only way I know to try and do that is to increase communication around them. 18:21:47 I'm worried that the pain level has to be higher before people will really care, but .. yeah 18:22:06 true 18:22:21 I agree, based on the results of the similar approach I took with eventlet, but ensuring folks know where status is IMO is part of what we're supposed to do 18:22:33 we can only do the work of 9 people, and there's a heck of a lot more to do to keep things going than that 18:22:41 that's not true 18:22:49 I know for a fact that gmann can do the work of three people :P 18:23:09 you should know that I am getting old :) 18:23:09 :) moving on 18:23:13 #topic Leaderless projects 18:23:20 Well, one of those three gmann's gets to lead this topic :D 18:23:34 #link https://etherpad.opendev.org/p/2024.1-leaderless 18:23:39 we have two project PTL appointment proposed for ling time and we should take decision soon i will say in this meeting 18:23:49 sahara: #link https://review.opendev.org/c/openstack/governance/+/899782 18:24:09 sahara is marked as inactive but PTL appointment and give them a chance to make it active does not hurt 18:24:18 2nd is rally: #link https://review.opendev.org/c/openstack/governance/+/898228 18:24:51 it has voting in favor and not in favor too but we should decide 18:25:39 I think they are open in gerrit for enough time, I would like to request tc-members who have not voted to do that 18:25:51 Yes, please do, especially since one of those does have a negative vote 18:26:04 I'd really rather not land something with a plurality if it has a negative vote 18:26:05 other 3 project remaining for PTL appointment is no progress, I will send email today to their previous PTL or any active maintained i get to know 18:26:47 JayF: I will say let's get the remaining vote today and we should either merge them or abandon them this week max 18:27:19 Yep, that's extremely reasonable, and abandoning isn't really a choice unless a project without a PTL is an OK state 18:27:37 I don't think "we take no action" is a good state here; we need to land these *or* take some other action to resolve the project's lack of leadership 18:27:49 but better than not deciding about it and PTL volunteer just wait for us to say yes or not to appointment 18:28:12 Having password issues but will vote if I haven’t 18:28:15 especially there is no other candidate or even maitainers 18:28:16 Yeah, you're right, and I appreciate you pointing it out and will take that advice -- I'll land these if eligible to land by EOW, even if only a plurality vote to land if still eligible 18:28:28 thanks. 18:28:38 Is there anything else on leaderless projects before wee move on? 18:28:41 that is all form my side on this. 18:29:03 Next topic was Implementation of Unmaintained branch statuses -- but I think we covered that during action items 18:29:06 so I'm going to skip it 18:29:11 #topic 2024.1 TC Tracker 18:29:17 #link https://etherpad.opendev.org/p/tc-2024.1-tracker 18:29:27 If there are any updated on TC Tracker items; please give them now. 18:30:50 OK, moving on. 18:31:00 #topic Upgrade issue between tooz and cinder 18:31:06 #link https://review.opendev.org/c/openstack/requirements/+/901131 18:31:11 #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/L72QU3SR2VFVYOFXVYH74V7HGMQ3YJRU/ 18:31:29 Unsure who added this to the agenda; but there's clearly an issue around Tooz upgrades that broke things, it got reverted but there are still issues here to resolve I believe 18:31:58 frickler: ^ looks like this was your topic 18:32:25 yes, that was me, I wanted to check whether we need to discuss some general guidance regarding interactions between these projects 18:33:03 naively it seems like tooz could sort out what version of etcd it is talking to then use the appropriate api paths 18:33:47 i.e. support multiple versions of etcd in tooz (noting that coinstalling services that needed those different versions would be challenging still) 18:34:50 yes, that's what https://review.opendev.org/q/topic:%22bug/2043810%22 does 18:35:19 I'm looking some stuff up, Ironic had similar failures and fixed it with devstack configuration 18:35:34 mainly because I want to ensure however it gets fixed that we all align in the same direction 18:36:01 having grenade upgrade also etcd would be another option 18:37:12 but also I added the topic before those tooz patches were proposed, so maybe all is well for now. and the next topic is even more interesting ;) 18:37:59 Hmm https://review.opendev.org/c/openstack/networking-generic-switch/+/895973 is the change Ironic landed 18:38:07 I think it's in the same realm but not the same mechanism 18:38:35 frickler: it sounds like there's not much else to talk abuot here though, things got moving over the holiday? 18:38:42 ack 18:38:56 #topic Declare projects inactive for lack of ...? 18:39:15 #info (list from agenda) Working py311 unit tests, Support for sqla2, Broken CI since 2023.2 release, Do we want/need to look at specific deliverables or only whole projects? 18:39:38 also added by me 18:39:54 definitely not sqla2 yet, IMHO 18:40:03 my main concern is what we can do to get sqla2 into u-c this cycle 18:40:11 the other two seem reasonable 18:41:09 if we have projects that do not support sqla2, but depend on global upper-constraints, what other options do we have? 18:41:12 gate broken even due to py 3.11 or any other reason is reasonable way to mark them inactive 18:41:46 frickler: because it's a lot of work for some projects and 2024.1 won't even be supported on a distro that has 2.0 AFAIK 18:42:10 I think we need to go with community wide goal for sqla2 work and maybe multicycle effort. 18:42:10 it's fine to have it be a requirement or goal or whatever, but marking a project as inactive because it hasn't crossed that line yet is way too large of a hammer, IMHO 18:42:15 yeah 18:42:26 While I mostly agree with frickler that we have to draw a line at some point, and it's getting to be that time, it seems late in the cycle to draw that line TBH. 18:42:34 agree, I think we discussed the same in gerrit in some change 18:42:42 and I suspect if we apply the other criteria, we'll catch clearly inactive projects without catching any that might just be late with SQLA 2.0 migrations 18:42:54 marking project inactive on this is little too strong and inconsistent in our process 18:42:55 yeah 18:43:06 I'm thinking of this like a graduated filter; sqla2.0 is a more fine filter 18:43:10 lets use the larger, coarser filters first 18:43:18 which more clearly say "these should not be in 2024.1" 18:43:36 and perhaps decide (not today; but this cycle) that next cycle is where we draw the SQLA 2.0 line 18:43:43 (it is R-18 according to the schedule) 18:43:48 sqla2.0 is a point in time hurdle.. it's a filter right now, but won't be later, where CI brokenness and py$current unit tests are more of a reasonable timeless filter, IMHO 18:44:55 o.k., so I will propose some inactivity marks based on the latter 18:45:02 * dansmith notes he's going to have to do his 15th recheck on that patch 18:45:15 and then I'll add sqla2 planning as a new topic 18:45:39 ++ 18:45:48 frickler++ Thank you for that, I put up (and abandoned) a governance change in that direction last cycle, it may be useful as a starting point or at least to review preexisting discussion 18:46:03 (I'm also not sure that it'll be sqla2.0 still ;) 18:46:14 Anything else on project inactivity before we move on? 18:47:09 #topic Open Discussion and Review 18:47:11 #link https://review.opendev.org/q/status:open+repo:openstack/governance 18:47:26 as mentioned earlier; please prioritize governance reviews; I will be making a pass in a couple of hours to land things eligible to land 18:47:33 and it's nice if a majority of us have voted on those changes 18:47:46 dansmith: I think you had something for open discussion 18:48:19 Just FYI to all, I'm out for the rest of the year starting tomorrow 18:48:58 dansmith: have a great holiday time :) 18:49:00 Enjoy your holiday! 18:49:15 Are there any other items for open discussion? 18:49:15 I shall endeavor to do so, thanks 18:49:23 Enjoy! and forget about gate :) 18:49:30 Have a good break 18:49:38 gmann: trust me, I shan't be thinking about the gate until Jan :) 18:49:41 just don't break the gate! 18:49:47 dansmith: ++ 18:50:23 Last call for items for open discussion 18:51:21 Thanks for coming to the TC meeting; see you all in 1 week for our monthly video meeting o/ 18:51:23 #endmeeting