15:00:09 <gmann> #startmeeting tc 15:00:09 <openstack> Meeting started Thu May 13 15:00:09 2021 UTC and is due to finish in 60 minutes. The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:12 <openstack> The meeting name has been set to 'tc' 15:00:20 <gmann> #topic Roll call 15:00:22 <ricolin> o/ 15:00:23 <gmann> o/ 15:00:30 <spotz> o/ 15:00:37 <dansmith> o/ 15:02:11 <jungleboyj> o/ 15:02:54 <gmann> yoctozepto6 and Belmiro would not be able to join as per "Apologies for Absence" section 15:02:57 <gmann> so let's start 15:03:03 <gmann> #link https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Agenda_Suggestions 15:03:05 <gmann> today agenda ^^ 15:03:21 <gmann> #topic Follow up on past action items 15:03:39 <gmann> gmann to add SIG chair/co-chair info in sig doc site 15:03:50 <gmann> I pushed the patch #link https://review.opendev.org/c/openstack/governance-sigs/+/790827 15:04:05 <gmann> it is in governance-sig so good to merge with 2nd +2 15:04:30 <gmann> please review and +A accordingly. 15:04:42 <gmann> gmann to prepare the etherpad for draft proposal of PTL+TC periodic meeting 15:05:03 <gmann> I have created the etherpad #link https://etherpad.opendev.org/p/tc-ptl-interaction 15:05:12 <gmann> which we will discuss in later topic 15:05:25 <gmann> #topic Gate health check (dansmith/yoctozepto) 15:05:38 <gmann> dansmith: any news you would like to share 15:05:40 <dansmith> so, I've seen a bunch of cinder fails in the last week 15:05:59 <dansmith> lots of rechecks due to volume related test failures, 15:06:15 <dansmith> so I assume the lull in activity was masking a lot of that, but I think we've still got work to do there 15:06:15 <jungleboyj> :-( 15:06:27 <dansmith> there's also the discussion about decommissioning the ELK infrastructure, 15:06:49 <dansmith> which is well-justified but will significantly reduce our visibility into the frequency of these failures 15:07:02 <dansmith> and limit our ability to work on the health as a system 15:07:21 <dansmith> not sure what to do about it, other than what has been communicated on the list at this point 15:07:24 <gmann> humm, that is valid concern 15:07:26 <jungleboyj> dansmith: Yeah, I was a little concerned with that discussion. 15:07:46 <dansmith> but clearly we either need someone to commit to a serious amount of time to rebuild it, and maintain it going forward, 15:07:52 <dansmith> or we're just going to have to deal without it 15:07:55 <dansmith> jungleboyj: me too 15:08:41 <dansmith> so, that's pretty much it for me..not a particularly happy picture all said 15:08:43 <jungleboyj> I was surprised by that discussion. 15:09:21 <dansmith> I was surprised that more people didn't consider it critical infrastructure 15:09:27 <spotz> If there's issues and no one to maintain it it makes sense. Kinda goes with the frequent call outs for more opendev help needed 15:09:35 <dansmith> but also somewhat surprised at how much resource it consumes 15:09:49 <fungi> it is not a lightweight service, no 15:10:20 <fungi> mostly because it's a massive amount of data our jobs are logging 15:10:25 <gmann> surly it will make hard for maintaining the CI/CD and tracking issue etc 15:11:38 <fungi> it's a classic "big data problem" which needs experienced "big data people" to do it well 15:12:02 <dansmith> fungi: well, that's really your fault 15:12:13 <dansmith> as my wife says "I value X less because you make X look so easy" :) 15:12:48 <dansmith> and ELK is definitely something I've taken for granted because it was just there and easy :) 15:13:47 <gmann> As this is no maintainer issues and kind of must-require for upstream CI/CD smooth maintenance, I feel this is one of the issue we should bring to broader audience. especially company or users. 15:13:59 <fungi> please do 15:14:11 <gmann> and bring this to Board also which should add some resources. 15:14:12 <dansmith> well, they did on the list, and without a deadline 15:14:29 <gmann> we were asking for infra help since many years 15:14:30 <jungleboyj> gmann: ++ I think this is an issue that needs to be highlighted. 15:15:23 <spotz> Yeah it definitely needs attention and some help 15:15:31 <gmann> fungi: any deadline for stopping those service? 15:15:34 <fungi> it's one of many services we're decommissioning, it just happens to be a higher-profile one 15:15:47 <fungi> gmann: not decided yet at least, so no for now 15:16:08 <gmann> ok, 15:16:23 <gmann> so let's take two action here 15:16:36 <fungi> we've in the past month or so taken out limesurvey (survey.o.o), asterisk (pbx.o.o) and our mqtt broker (firehose.o.o) 15:16:41 <spotz> Can we move it to the bottom of the list to give us more time?:) 15:16:43 <fungi> and there will certainly be more 15:17:22 <fungi> and yes, like i said, clark started the discussion early so folks can come up with solutions 15:17:34 <fungi> it's not going away yet, it just needs to 15:18:37 <gmann> 1. ML to especially highlight the ELK services and ask for help explicitly with ref to current thread 15:18:51 <gmann> 2. Bring this to Board meeting which is planned on June 29. 15:18:56 <spotz> I'm pretty sure folks saw the email but not the importance 15:19:22 <dansmith> spotz: because they don't know what ELK is doing for us, or because of some other reason? 15:19:25 <dansmith> not enough shouting? 15:19:28 <jungleboyj> Based on the responses I am sure a lot of people are hoping the question will just go away. :-) 15:20:03 <fungi> i can guarantee the question will eventually go away, just maybe not in the way those people are hoping 15:20:06 <spotz> dansmith: don't know what it's doing for us 15:20:11 <dansmith> okay 15:20:13 <jungleboyj> fungi: ++ 15:20:40 <dansmith> unfortunately a minority of people work on gate issues (which is an issue itself), so those are the ones likely to feel strongly 15:20:46 <spotz> Kinda like we're retiring asterisk, well I don't use it so no big deal. But it could be tied into an aleerting system whiichh I'm not aware of... 15:20:56 <dansmith> removing it will raise the bar for difficulty on that already unsung and unpopular task 15:21:04 <gmann> agree 15:21:23 <dansmith> so it's kinda self-fulfilling that it's not a big deal to more people 15:21:23 <gmann> dansmith: do you mind doing 1st one, raising critical alarm on ML for ELK services ? 15:22:05 <fungi> i'm curious how the current discussion on the ml could have reached a broader audience, personally 15:22:17 <dansmith> gmann: well, I can, but I'm a bit confused about what the current email is lacking 15:22:19 <dansmith> right what fungi said 15:22:50 <dansmith> we can send another email with "ELK: why you should care" but I'm not sure it will really garner more attention, but if people think it's important then I can 15:22:54 <fungi> it sounds like, based on dpawlik's recent response, that rdo might be interested in running something 15:22:55 <gmann> I am fine with only current but it has few more services listed also so in case we want to highlight ELK explicitly 15:23:12 <dansmith> fungi: yeah but I'm worried they don't realize the 1TiB/day requirement :P 15:23:22 <dansmith> gmann: ah, okay 15:23:43 <spotz> fungi I can check on the RDO side 15:23:51 <gmann> especially what we must continue for day to day debugging 15:24:17 <gmann> which is what we can highlights in various platform like newsletter, Board etc 15:24:35 <dansmith> okay well, I'll work on a draft and look for some reviews 15:24:46 <gmann> +1 thanks 15:25:20 <gmann> #action dansmith Draft the ML to explicit Alert/highlight the ELK services maintenance help with ref to current thread 15:25:20 <jungleboyj> dansmith: ++ Thanks. 15:25:30 <gmann> I will bring this on Board agenda 15:25:48 <gmann> #action gmann to bring the ELK/infra help to Board 15:26:01 <gmann> any other things to discuss in Gate check topic? 15:26:17 <dansmith> not from me 15:26:42 <gmann> thanks for brining this and keeping eyes on gate things 15:26:45 <gmann> #topic Planning for TC + PTL interaction (gmann) 15:27:01 <gmann> I composed a draft in #link https://etherpad.opendev.org/p/tc-ptl-interaction 15:27:46 <gmann> I would like to ask each members to add your opinion in etherpad and then we can discuss it in next meeting 15:28:56 <gmann> or feel free to add more discussion point too if you want 15:29:09 <gmann> #topic Xena cycle tracker status check 15:29:17 <spotz> Added a comment for if we hold it AT the PTG:) 15:29:30 <gmann> spotz: cool thanks 15:29:34 <gmann> #link https://etherpad.opendev.org/p/tc-xena-tracker 15:29:47 <gmann> this is tracker's tacking week 15:30:51 <gmann> as this is still initial of cycle, we can just discuss the one having progress instead of tracking all item's progress 15:31:28 <gmann> one is from dansmith on charter change on ATC->AC #link https://review.opendev.org/c/openstack/governance/+/790092 15:31:40 <gmann> we have discussion on gerrit on this 15:32:00 <gmann> anything to discuss here or continue on gerrit itself ? 15:32:41 <ricolin> I think we can keep the discussion in review:) 15:33:12 <ricolin> I mean for https://review.opendev.org/c/openstack/governance/+/790092 15:33:27 <gmann> on fungi's point on Bylaw, i left comment there for a way to comply Bylaws. basically what we did for "OpenStack Technical Committee Approved Release" term 15:33:44 <gmann> hope that should be fine from Bylaws perspective 15:34:51 <gmann> moving next 15:34:54 <gmann> Audit and finish the previous cycle pending community-wide goal work 15:35:01 <gmann> I started few patches for PTL guide 15:35:13 <gmann> #link https://review.opendev.org/q/topic:%2522project-ptl-and-contrib-docs%2522+status:open 15:35:18 <gmann> not many but just to start 15:35:22 <fungi> yes, as long as the tc indicates that whatever it's producing is the atcs list the bylaws requires, then it's fine. the bylaws already has a carve-out for adding people who aren't code contributors, just need to have some way to bridge the naming divide 15:35:38 <spotz> Oh in regards to PTL elections, diablo_rojo mentioned it's based on ATC status. Could it be AAC status plus a contrib in that project? 15:35:40 <gmann> yeah 15:35:51 <spotz> AC not AAC 15:35:59 <fungi> it's technically "apc" for ptl elections, not atc 15:36:08 <gmann> yeah, PTL are APC always 15:36:31 <spotz> Ok so reword:) The APC list build off of AC + project contrib? 15:36:41 <fungi> ptl elections also aren't mentioned in the bylaws, so the tc can do whatever it likes there as long as the requirements for editing the charter are observed 15:36:43 <gmann> means ATC which can be contributor in x project would have vote for project y pTL election 15:37:06 <gmann> *would not 15:37:38 <gmann> and yes, we in TC repo will keep list of ATC (renamed as AC ) 15:38:26 <ricolin> Isn't the ATC less for a year? 15:38:47 <gmann> spotz: APC is project contrib only. AC = APC+ other active contributors 15:38:56 <fungi> atc is actually for "365 days" according to the bylaws, i think 15:39:26 <ricolin> fungi, that's what I remember:) 15:39:33 <gmann> yeah 15:39:34 <gmann> " The term shall be for three hundred and sixty five days after the date of approval of the application by the Technical Committee." 15:40:20 <gmann> section 3.a.ii in https://www.openstack.org/legal/technical-committee-member-policy/ 15:40:39 <fungi> for election purposes we've basically considered that to be interpreted as two release cycles 15:40:50 <gmann> right 15:40:57 <spotz> Ok shoot:( 15:42:57 <gmann> anyways let's review the patch and continue discussion there 15:43:00 <gmann> moving next 15:43:06 <gmann> TC members to drive the Y cycle community wide goal 15:43:12 <gmann> ricolin: any updates after you sent the ML 15:44:11 <ricolin> gmann, No much really, mainly checking around with backlogs for goals ans make some survey 15:44:34 <gmann> great. may be we need to cleanup backlog also 15:44:41 <gmann> we did in Ussuri cycle at some extend 15:44:42 <ricolin> I think so 15:45:01 <gmann> thanks for working on this. 15:45:15 <ricolin> I can make some clean 15:45:35 <ricolin> Personally, I like the container image support one 15:45:36 <gmann> I do not see any other progress in etherpad please bring here if there is any and i missed that 15:45:45 <gmann> ricolin: +1 15:46:21 <ricolin> wondering if mnaser have any idea how can we push that one forward 15:47:01 <gmann> I think there was patch also for proposing it as goal 15:47:24 <ricolin> gmann, if RBAC is proposed, we only need one goal to search for 15:47:39 <gmann> ricolin: #link https://review.opendev.org/c/openstack/governance/+/720107 15:47:42 <ricolin> gmann, that was abandoned for an year 15:47:45 <gmann> yeah 15:48:04 <gmann> ricolin: yeah I will compose RBAC one, most probably next week or so 15:48:07 <dansmith> I think a number of projects are working on rbac stuff, 15:48:14 <dansmith> so that seems like a no-brainer right? 15:48:19 <ricolin> If we allow some project skip from the container goal 15:48:24 <gmann> yeah, many are left with less attention 15:48:36 <ricolin> it might get easier to be accepted 15:48:53 <gmann> as per discussion with lbragstad we could do goal this cycle and close the popup team but as we are not doing any goal in this cycle we could do in Y 15:49:21 <spotz> lbragstad is out on paternity:) 15:49:22 <gmann> and to make sure all project are done and we can try out the default enabling new rbac via oslo or so 15:49:29 <gmann> yeah 15:49:52 <jungleboyj> spotz: Oh wow, Didn't know that. Good for him! 15:50:58 <gmann> anything else on goal or tracker status? 15:51:32 <ricolin> tc-members please help to suggest Y-cycle goal if you see any potential fits 15:51:41 <ricolin> nope 15:51:47 <gmann> +1. 15:51:52 <ricolin> but we need to do it asap 15:52:28 <gmann> #topic Open Reviews 15:52:30 <gmann> #link https://review.opendev.org/q/project:openstack/governance+is:open 15:52:35 <gmann> we have two open review 15:52:55 <gmann> this is eligible for approve #link https://review.opendev.org/c/openstack/governance/+/790582 15:53:14 <gmann> other in discussion. 15:53:25 <gmann> that's all for today I think 15:53:41 <gmann> thanks all for joining 15:53:45 <gmann> #endmeeting