Thursday, 2020-01-16

*** TheJulia has quit IRC00:18
*** tdasilva has quit IRC00:18
*** tdasilva has joined #openstack-tc00:19
*** TheJulia has joined #openstack-tc00:22
*** csatari has joined #openstack-tc00:23
*** rpittau|afk has joined #openstack-tc00:35
*** bauzas has quit IRC00:36
*** bauzas has joined #openstack-tc00:38
*** rpittau|afk has quit IRC00:42
*** tdasilva has quit IRC00:42
*** knikolla has quit IRC00:42
*** csatari has quit IRC00:42
*** ildikov has quit IRC00:43
*** gmann has quit IRC00:43
*** aprice has quit IRC00:43
*** TheJulia has quit IRC00:44
*** tdasilva has joined #openstack-tc00:44
*** mwhahaha has quit IRC00:45
*** mnasiadka has quit IRC00:46
*** jungleboyj has quit IRC00:46
*** coreycb has quit IRC00:46
*** tonyb[m] has quit IRC00:47
*** jrosser has quit IRC00:48
*** masayukig has quit IRC00:49
*** mnaser has quit IRC00:49
*** johnsom has quit IRC00:52
*** jbryce has quit IRC01:01
*** zhurong has joined #openstack-tc01:08
*** jamesmcarthur has joined #openstack-tc01:14
*** jamesmcarthur_ has joined #openstack-tc01:16
*** csatari has joined #openstack-tc01:16
*** jamesmcarthur has quit IRC01:17
*** mnasiadka has joined #openstack-tc01:18
*** zhurong has quit IRC01:18
*** csatari has quit IRC01:22
*** csatari has joined #openstack-tc01:24
*** zhurong has joined #openstack-tc01:27
*** mnasiadka has quit IRC01:29
*** csatari has quit IRC01:31
*** jamesmcarthur_ has quit IRC01:40
*** jamesmcarthur has joined #openstack-tc01:42
*** jamesmcarthur has quit IRC01:47
*** jamesmcarthur has joined #openstack-tc02:12
*** jamesmcarthur_ has joined #openstack-tc02:16
*** jamesmcarthur has quit IRC02:16
*** jamesmcarthur_ has quit IRC02:26
*** jamesmcarthur has joined #openstack-tc02:34
*** ricolin has quit IRC02:49
*** ijolliffe has joined #openstack-tc03:00
*** knikolla has joined #openstack-tc03:18
*** knikolla has quit IRC03:24
*** knikolla has joined #openstack-tc03:39
*** jamesmcarthur has quit IRC03:41
*** jamesmcarthur has joined #openstack-tc03:42
*** knikolla has quit IRC03:44
*** jamesmcarthur has quit IRC03:48
*** jamesmcarthur has joined #openstack-tc03:56
*** jamesmcarthur_ has joined #openstack-tc03:59
*** jamesmcarthur has quit IRC03:59
*** knikolla has joined #openstack-tc04:04
*** knikolla has quit IRC04:09
*** openstackgerrit has quit IRC04:12
*** jungleboyj has joined #openstack-tc04:14
*** jungleboyj has quit IRC04:20
*** masayukig has joined #openstack-tc04:23
*** masayukig has quit IRC04:26
*** jamesmcarthur_ has quit IRC04:55
*** jamesmcarthur has joined #openstack-tc04:57
*** jamesmcarthur has quit IRC05:02
*** jamesmcarthur has joined #openstack-tc05:26
*** jamesmcarthur has quit IRC05:32
*** evrardjp has quit IRC05:34
*** evrardjp has joined #openstack-tc05:34
*** jamesmcarthur has joined #openstack-tc06:00
*** jamesmcarthur has quit IRC06:07
*** tdasilva has quit IRC06:23
*** e0ne has joined #openstack-tc06:34
*** e0ne has quit IRC06:35
*** gmann has joined #openstack-tc06:48
*** mnaser has joined #openstack-tc06:49
*** mwhahaha has joined #openstack-tc06:51
*** csatari has joined #openstack-tc07:00
*** e0ne has joined #openstack-tc07:02
*** johnsom has joined #openstack-tc07:02
*** jamesmcarthur has joined #openstack-tc07:02
*** coreycb has joined #openstack-tc07:03
*** jrosser has joined #openstack-tc07:03
*** TheJulia has joined #openstack-tc07:03
*** masayukig has joined #openstack-tc07:05
*** jamesmcarthur has quit IRC07:09
*** knikolla has joined #openstack-tc07:11
*** jungleboyj has joined #openstack-tc07:11
*** mnasiadka has joined #openstack-tc07:14
*** rpittau|afk has joined #openstack-tc07:19
*** aprice has joined #openstack-tc07:20
*** jbryce has joined #openstack-tc07:21
*** ildikov has joined #openstack-tc07:25
*** soumplis has joined #openstack-tc07:29
*** tonyb[m] has joined #openstack-tc07:31
*** e0ne has quit IRC07:38
*** tdasilva has joined #openstack-tc07:47
*** slaweq has joined #openstack-tc08:07
*** rpittau|afk is now known as rpittau08:31
*** tosky has joined #openstack-tc08:32
*** dmellado has quit IRC08:35
*** iurygregory has joined #openstack-tc08:36
*** dmellado has joined #openstack-tc08:37
*** sapd1 has quit IRC08:55
*** sapd1_ has joined #openstack-tc08:55
*** jamesmcarthur has joined #openstack-tc09:06
*** jamesmcarthur has quit IRC09:10
*** openstackgerrit has joined #openstack-tc09:55
openstackgerritLuigi Toscano proposed openstack/governance master: Propose a new goal to migrate all legacy zuul jobs  https://review.opendev.org/69127809:55
*** e0ne has joined #openstack-tc10:12
*** rpittau is now known as rpittau|bbl10:21
*** lxkong has joined #openstack-tc11:00
*** rpittau|bbl is now known as rpittau12:25
*** ijolliffe has quit IRC12:29
*** tetsuro has quit IRC12:45
*** ijolliffe has joined #openstack-tc13:12
*** jamesmcarthur has joined #openstack-tc13:15
*** ijolliffe has quit IRC13:19
*** lxkong has quit IRC13:20
*** ijolliffe has joined #openstack-tc13:21
*** jamesmcarthur has quit IRC13:24
*** jamesmcarthur has joined #openstack-tc13:25
evrardjpzaneb we don't have enough patches up so it seems...13:25
*** jamesmcarthur has quit IRC13:33
evrardjpI think we are supposed to have a meeting in 20 minutes, did I get that right?13:40
evrardjpthere is no apologies for absence, so I suppose everyone will be there :)13:42
openstackgerritJean-Philippe Evrard proposed openstack/governance master: Reorder repos  https://review.opendev.org/70167913:43
*** ianychoi_ has quit IRC13:44
openstackgerritJean-Philippe Evrard proposed openstack/governance master: Make mistral-dashboard a separate deliverable  https://review.opendev.org/70036813:44
*** jamesmcarthur has joined #openstack-tc13:44
gmanno/13:47
ttxo/13:48
*** ricolin has joined #openstack-tc13:48
ricolino/13:49
openstackgerritRico Lin proposed openstack/governance-sigs master: Add Tony Breeds as Multi-Arch SIG chair  https://review.opendev.org/70286713:51
ricolintonyb, ^^^13:52
openstackgerritMerged openstack/governance master: Add ideas for OpenStack repository  https://review.opendev.org/70167813:52
tonyb+1'd13:53
tonybThanks13:53
evrardjptc-members meeting in 2 minutes, if my personal NTP isn't wrong13:58
jrollthat is accurate by my personal NTP13:58
ttxI might have to drop before the end, but I'm here right now13:59
evrardjpttx: I put your topics at the start13:59
evrardjp#startmeeting tc14:00
openstackMeeting started Thu Jan 16 14:00:04 2020 UTC and is due to finish in 60 minutes.  The chair is evrardjp. Information about MeetBot at http://wiki.debian.org/MeetBot.14:00
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.14:00
*** openstack changes topic to " (Meeting topic: tc)"14:00
openstackThe meeting name has been set to 'tc'14:00
njohnstono/14:00
evrardjp#topic rollcall14:00
*** openstack changes topic to "rollcall (Meeting topic: tc)"14:00
mnasero/14:00
jungleboyjO/14:00
ttxo/14:00
*** witek has joined #openstack-tc14:00
ricolino/14:01
evrardjpo/14:01
evrardjpone more and we have quorum14:01
evrardjpjroll: still there? :)14:01
gmanno/14:01
evrardjpwoot14:01
evrardjpwe have quorum14:01
njohnstono/14:01
evrardjplet's start!14:02
evrardjp#topic Report on large scale sig14:02
*** openstack changes topic to "Report on large scale sig (Meeting topic: tc)"14:02
evrardjp#link https://media.giphy.com/media/FlOHnTIv1mj2U/giphy.gif14:02
*** diablo_rojo_phon has joined #openstack-tc14:02
evrardjpttx: how does this fly and how/what are the action items?14:02
jroll\o14:02
ttxI post regular updates on the ML about it, it's going well so far14:03
evrardjpyup thanks about that14:03
ttxParticipants are motivated, they held a meeting this week while I could not chair it and the co-chairs stepped up14:03
evrardjpgreat to hear!14:03
ttxSo yes, current signs are pretty positive, now to turn that into results :)14:04
*** diablo_rojo has joined #openstack-tc14:04
evrardjpIf you don't mind let's keep that for next month, so we'll see if the momentum is still there?14:04
*** jamesmcarthur has quit IRC14:04
ttxyes, sure. I'll post a summary of this week meeting later this week14:05
evrardjpthanks!14:05
evrardjp#action ttx post a summary of large scale sig's meeting on the ML this week14:05
evrardjp#topic report about the vision reflection update, that should be started in January14:05
*** openstack changes topic to "report about the vision reflection update, that should be started in January (Meeting topic: tc)"14:05
evrardjpttx, you want to say a few words?14:06
ttxwell the postortem on that was pretty short14:07
evrardjpindeed, so I thought you might want to say a few words14:07
ttxLike few really seemed to care or object to my summary14:07
ttxTL;DR being: glad we did it, no value in doing another immediately again14:08
openstackgerritMerged openstack/governance master: Make mistral-dashboard a separate deliverable  https://review.opendev.org/70036814:08
openstackgerritMerged openstack/governance master: Reorder repos  https://review.opendev.org/70167914:08
ttxI think it was useful to wrap it up with a ML post though14:08
evrardjpI think too14:08
ttxso if we wind down other things in the future it's definitely a good way to close14:08
evrardjp#link http://lists.openstack.org/pipermail/openstack-discuss/2019-December/011566.html14:10
evrardjplet's move on, shall we?14:10
evrardjp#topic report on the analysis of the survey14:10
*** openstack changes topic to "report on the analysis of the survey (Meeting topic: tc)"14:10
evrardjpjungleboyj, a few words?14:10
jungleboyjHey.14:10
jungleboyjSo, a few people have reviewed what I put together.14:11
jungleboyj#link https://review.opendev.org/#/c/698582/14:11
jungleboyjThanks to all who have given input.14:11
jungleboyjIt didn't look like there were strong objections to anything I said.14:11
* asettle scrambles in the front door14:12
evrardjpI guess that we are setting up a framework for the next iterations, do anyone think there is something missing?14:12
jungleboyjzaneb:  Was hoping to have me do an analysis of users who said they used a project as compared to contribution.14:12
evrardjpcan people check on the review and comment there, if not done already?14:12
jungleboyjI could do that analysis.14:13
evrardjpI think it could be valuable14:13
gmanni think after the % update for the users vs contributors. it will good data to publish somewhere to motivate more contribution from users. newsletter or ML ?14:13
jungleboyjI think the other open question is where this document should go.14:13
evrardjpgmann: maybe even collaborate with the board on this14:13
gmann+114:13
evrardjp(after publication)14:13
jungleboyjgmann: evrardjp ++14:13
jungleboyjI think there is good information in there.14:13
zanebo/14:14
zanebforgot to put this in my calendar14:14
jungleboyj:-)14:14
evrardjpyeah I didn't send an ics or something, maybe I should if it's not on the usual day14:14
jungleboyjSo, are we thinking this would be an ML post or go to a blog or ...14:14
evrardjpI am okay if we put this in goverannce. We can post the results in the ML afterwards.14:15
evrardjp(pointing to governance)14:15
fungiappendices to the project teams guide could be an option too14:15
evrardjpor the other way around, I don't really mind.14:15
*** diablo_rojo has quit IRC14:15
gmannyeah, keeping in governance doc will be good to quickly refer and maintain history14:15
evrardjpI think it's a governance topic, so I would prefer seeing it there14:15
jungleboyjgmann:  ++ Ok.  I was hoping it was something we would hold on it.14:15
*** diablo_rojo has joined #openstack-tc14:16
jungleboyjSo, I if anyone else has comments, please review.  I will make another draft next week.14:17
evrardjpsounds good, thanks jungleboyj!14:17
jungleboyjI will add the user data.14:17
evrardjpLet's move on, and discuss this in the review14:18
evrardjp#topic report on infra liaison and static hosting14:18
*** openstack changes topic to "report on infra liaison and static hosting (Meeting topic: tc)"14:18
jungleboyjIf anyone has strong feelings about where the document should go, please add those in the review.14:18
jungleboyj++14:18
evrardjpmnaser: any progress to report ?14:18
evrardjpis it me, or is the meeting really silent, and boring? Where are all the lively things?14:19
mnaseri haven't done much on that to be honest14:19
mnaseri think andres put all the base infra14:19
mnaserbut i havent followed up with the rest, i'm planning to take a few cycles this weeekend to get through it14:20
evrardjpok -- It might impact the "ideas" project, so I will have to contact you back on this, and help if necessary.14:20
mnasersure, please do14:20
jungleboyjhttps://gph.is/2mWzTNG14:20
evrardjpjungleboyj: :)14:20
fungiyeah, several changes for it merged yesterday14:21
fungi(for the underlying hosting automation)14:21
mnaseropenstack/ideas might be a good poc for that, but yeah, i need to pick up the slack on that14:21
evrardjpwe can discuss that later in the ideas section if you like14:21
evrardjplet's move on to another topic on which jungleboyj's gif applies too14:21
evrardjp#topic report on stable branch policy work.14:21
*** openstack changes topic to "report on stable branch policy work. (Meeting topic: tc)"14:21
evrardjp#link https://media.giphy.com/media/XpYyhn0Gutfaw/giphy.gif14:22
evrardjpmnaser, any extra since the email of smcginnis changing stable reviews from releases perspective? Is that enough? Is there more required? Should we revisit in a few months?14:22
jungleboyj*Laughing*14:22
mnaseri have a drafted change that i plan to push up on what we seem to have converged on the ML14:22
mnaseror at least, what i think we did :)14:22
evrardjpok14:23
evrardjpoh I forgot a few action items in the previous topic, my bad14:23
evrardjp#action mnaser follow up on the static hosting14:24
evrardjp#action jungleboyj update the survey report patch14:24
gmannmnaser: +1. pushing that up for review will be good.14:24
evrardjp#action mnaser push an update to distill what happened on the ML in terms of stable branch policy14:25
evrardjpsorry for the mess there.14:25
njohnstonhttps://media.giphy.com/media/Kun1AbKVIWhRm/giphy.gif14:25
*** ianychoi has joined #openstack-tc14:25
evrardjpI suppose we can go ahead, and we'll wait for your update mnaser14:25
jungleboyjnjohnston:  We all have those days.14:25
evrardjpjungleboyj: for me, it's today :p14:26
evrardjp#topic report on the oslo metrics project14:26
*** openstack changes topic to "report on the oslo metrics project (Meeting topic: tc)"14:26
jungleboyj:-)14:26
evrardjpmnaser jroll: has code appeared since last convo?14:26
mnaseri think ttx had the ocmms with the individuals14:26
jrollnot that I have seen14:26
mnaserand afaik none has, not that i know of14:26
evrardjp#link https://media.giphy.com/media/12NUbkX6p4xOO4/giphy.gif14:26
ttxchecking for progress from last meeting logs14:26
zanebbnemec: o/14:26
evrardjpzaneb: well thought :)14:27
ttxso... spec is coming up, a bit delayed since the author had to deal with an outage in January14:27
ricolinttx large scale SIG will help on tracking one oslo.metrics?14:27
ttxwe large scale SIG are waiting for the first draft of that spec (from masahito @ LINE) before refining it14:28
*** jamesmcarthur has joined #openstack-tc14:28
ttxhopefully more to tell by end of month14:29
evrardjpok I will keep for the next month then14:29
ttxyes please14:29
evrardjpok moving on14:29
ricolinttx that's great14:30
evrardjp#topic report on the community goals for U and V, py2 drop, and goal select process schedule14:30
*** openstack changes topic to "report on the community goals for U and V, py2 drop, and goal select process schedule (Meeting topic: tc)"14:30
evrardjpgmann: you want to say a few words? I saw patches updated very recently and things merging, so that's good! :)14:30
gmannGoals for U: This is latest and final update on selection of Ussuri cycle goals #link http://lists.openstack.org/pipermail/openstack-discuss/2020-January/012019.html14:30
gmannwe have both goals ready now #link https://governance.openstack.org/tc/goals/selected/ussuri/index.html14:31
evrardjpthanks again for tackling this gmann. It's very hard work.14:31
ttx++14:31
jungleboyj++14:31
gmannV cycle goals: zuulv3 goal is updated by tosky. #link https://review.opendev.org/#/c/691278/14:31
gmannits lgtm, ready for review from other members.14:32
evrardjpDo you need any support outside the reviews?14:32
gmannsupport for implementation ?14:32
evrardjpanything you think it's good to spread the word14:33
gmanntosky is champion to this goal, i would leave the planning and implementation part to him once we select it for V. and yes we can help him for sure.14:33
njohnstonNeutron stadium is already close to done with the zuulv3 migration, we have been doing it in tandem with the py27 drop14:33
gmannonce it is merged in proposed dir, i will move it to selected.14:33
evrardjpyes I think it's fair to say they will be selected :)14:34
gmannnjohnston: +1. nova is also almost done. few experimental jobs are left seems14:34
gmannpy2 drop: This is last update is ent on ML and next will be tomorrow: #link http://lists.openstack.org/pipermail/openstack-discuss/2019-December/011580.html14:35
gmannphase-2 is going on which is common lib and QA tooling to start dropping the py2.14:35
toskynjohnston: good to hear!14:35
gmannpy2 drop schedule is there on main ussuri cycle schedule #link https://releases.openstack.org/ussuri/schedule.html14:35
gmannGoal select process schedule: This is merged and available in goal doc #link https://governance.openstack.org/tc/goals/index.html#goal-selection-schedule14:36
evrardjpI should have removed that one14:36
smcginnisFor the record, I think it's way too late in the cycle to be approving more cycle goals.14:37
ricolintosky, FYI heat jobs (except for grenade job) just migrated to zuul v314:37
gmannmy plan is to start this process for W cycle goals which will be in V cycle. let's see how it goes.14:37
gmannsmcginnis: zuulv3 migration s for V cycle. Ussuri cycle is finalized with 2 goals14:37
smcginnisgmann: I was referring to the newly approved U goal.14:38
gmannsmcginnis: ohk. that is easy one i think in term of projects involvement and effort.14:38
gmannthat's all from my side on goal things.14:38
evrardjpthanks for the report.14:39
evrardjpit's a very hard thing, and we are playing catch-up. I think gmann is doing a tremendous work to make sure we don't have late goals for V and W, so again, thank you gmann.14:39
evrardjplet's move on14:40
evrardjp#topic Report on the convo with Catalyst14:40
*** openstack changes topic to "Report on the convo with Catalyst (Meeting topic: tc)"14:40
ricolingmann, we can create a ics schedule so no step of W cycle goals (including pre-select part) will be forgot14:40
gmannyeah target to streamline for V and W.14:40
evrardjpricolin: evrardjp where are we now?14:40
gmannricolin: +1. I  will check.14:40
ricolinAs promised, Telemetry PTL sended the report out14:41
ricolin#link http://lists.openstack.org/pipermail/openstack-discuss/2019-December/011586.html14:41
ricolinIn general, people from Telemetry team are welcome any one to join the discussion and help with that issue.14:41
ricolinI think we need someone with experience on monitoring services to keep following on next step.14:41
ricolinI can help on checking the status as auto-scaling SIG also affected by it, but will help if someone got more experience stepped in and help.14:41
evrardjpas far as I know there was this idea of forking gnocchi back into openstack/opendev, not really sure where we are14:41
ricolinevrardjp, that's considered as one of solution14:41
* ricolin think we should stop calling convo with Catalyst since one of the co-PTL is from ZTE :)14:43
ttxI need to run now, will catch up with logs14:43
evrardjpricolin: fair. I should have said telemetry in general.14:43
*** jamesmcarthur has quit IRC14:44
evrardjpI am confused about the next steps. What if nobody shows up, we're gonna have things re-merged into telemetry, things that were painfully removed?14:44
ricolinevrardjp, that won't be worst than keep it broken14:45
openstackgerritAkihiro Motoki proposed openstack/governance master: doc nit: remove unnecessary space  https://review.opendev.org/70287914:45
zanebthere were several volunteers to help maintain gnocchi in that thread14:47
evrardjpPeople seemed to have disagreed there, so I am concerned of a split.14:47
zanebI would like to see us go ahead with the fork14:47
evrardjpzaneb: I haven't seen an action from them14:47
zanebbut I don't know that anybody picked up that action14:47
evrardjpzaneb: I agree though14:47
ricolinto fork gnocchi is not the worst idea, but we need to have people to make sure all discussion has been made (including the discussion in Gnocchi part)14:47
zanebyes14:47
evrardjpI think forking gnocchi isn't _the_ solution, but it needs to enter in the mind of people actively looking for one :)14:48
zanebbut I don't think they'll object. they all have other jobs now and they effectively abandoned it14:48
fungijust to be clear, did ceilometer break due to gnocchi ceasing maintenance, or is it merely the cnocern that it might break in the future because gnocchi is unmaintained?14:48
zanebfungi: the latter14:48
evrardjpsadly only the latter afaik fungi14:48
evrardjpbut there are other concerns apparently14:48
evrardjp"mongo works for us"14:49
fungiso there is a scramble for ceilometer to remove a backend which is currently still working and was considered at the time of implementation superior to what they're hoping to switch it out for14:49
zanebtbh I think the issue is more that it doesn't fit with the plans of the folks actually maintaining telemetry14:49
mnaserfwiw from an operator pov14:49
zanebthe people working on it actively are the people who need to keep it how it was 5 years ago14:49
zanebthe people who are happy with gnochhi are not contributing because they're happy with how it is14:50
mnaseri have given up on the telemetry stack and moved on to prometheus based stuff14:50
jungleboyjmnaser:  I get the feeling that a lot of people have gone that way.14:50
mnaseri honestly cant recommend our telemetry stack to attempting to run openstack at decent scale14:50
zanebmnaser: y'all do billing with prometheus?14:51
mnaserand the biggest root of that problem is the "holding onto things" which is manifesting again by adding the ceilometer api14:51
evrardjpit's maybe better to kill a project than trying to revive one with almost no future14:51
ricolinevrardjp, some current member from Telemetry team figure out some way to optimize mongo14:51
mnasermongodb is a document database14:51
mnaserits not meant for metrics14:52
mnaserthat was a mistake we did a long time ago because mongo was WeB sCaLe back then14:52
mnaserthere is better, more reliable, more actively developed options14:52
mnaserzaneb: yes, and we have a bunch of open sourced exporters too (one for libvirt and another one for openvswitch soon)14:52
evrardjpwe have 4 topics remining and 8 minutes14:52
evrardjpshould we continue this in the office hours?14:52
gmann+114:53
zaneb+114:53
ricolinIMO as long as it's just a optional driver, I think bother to block that effort14:53
ricolin+114:53
njohnston+114:53
jungleboyj+114:53
evrardjpI think the last topics to report are fast14:53
evrardjpok let's move on14:53
evrardjp#topic Report on multi-arch SIG14:53
*** openstack changes topic to "Report on multi-arch SIG (Meeting topic: tc)"14:53
ricolinthink so too:)14:53
evrardjpricolin: I see there are updates on that, even just before the meeting14:53
evrardjpanything to report?14:53
ricolinSIG is doing great, found a lot of tasks that this SIG can  works on.14:53
ricolinHere's the initial meeting log from last meeting if anyone would like to see14:54
evrardjpand I suppose people are taking the tasks? :)14:54
ricolin#link http://lists.openstack.org/pipermail/openstack-discuss/2020-January/011894.html14:54
evrardjpthanks!14:54
ricolinWe also keep asking people if they can provide more CI resources to community, or join in anyway. For now, I document all resources in SIG etherpad14:54
jungleboyjThat is good news.14:54
evrardjpgood news indeed14:54
gmannnice.14:54
ricolinIf you know anyone to can help on provide CI resources let us know!:)14:54
evrardjpanything else to report?14:54
ricolinI just propose adding tonyb as SIG chair https://review.opendev.org/7028614:54
ricolinHis experience will definitely help a lot.14:54
mnaserfwiw we have aarch64 nodes at vexxhost14:55
evrardjp#action everyone let the multi-arch SIG know if you have CI resources in different architectures :)14:55
mnaserwe can probably help out wit ha few.14:55
ricolinmnaser, that's great to know!:)14:55
ricolinAnd finally, I will set up StoryBoard and repo for SIG before next Tue. . After that the others can have more clear view on what's going on in Multi-Arch.14:55
evrardjpmnaser: awesome!14:55
mnaseri'm happy to sponsor hosting different archs if a hardware vendor wants to provide hardware14:55
mnaserarm was extremely cool at providing us with hardware to make available, so yeah.14:55
ricolinmnaser, that's definitely helpful14:56
jungleboyjmnaser:  That is great.14:56
evrardjpanything else?14:56
ricolinevrardjp, that's all I got14:56
evrardjpok let's move to next topic14:56
evrardjp#topic report on release naming14:56
*** openstack changes topic to "report on release naming (Meeting topic: tc)"14:56
evrardjpmugsie: Can we remove this, as the policy is changed, and we already started V/W already.14:56
evrardjpcorrect?14:56
tonyb[m]mnaser: I think there are vendors interested but it's a lot slower than I'd like14:56
mugsieevrardjp: sorry, double meeting -_- but yeah, I think os14:57
mugsieso*14:57
evrardjpok14:57
evrardjplet's move to the next topic14:57
evrardjp#topic report on the ideas repo14:57
*** openstack changes topic to "report on the ideas repo (Meeting topic: tc)"14:57
jungleboyjOur naming was Victoriaous ;-)14:57
evrardjpjungleboyj: haha14:57
ricolinevrardjp, what's next step plan for that repo:)14:57
evrardjpok, the repo for crazy^W ideas got created recently14:57
evrardjpI will make sure that the framework is in place14:58
jungleboyjhttps://gph.is/2GiLazc14:58
evrardjpthe idea would be to have it freely available and linked in the governance website14:58
evrardjp(pun intended)14:58
jungleboyj:-)14:59
evrardjpmore seriously, the content is there for anyone to jump and write what they think.14:59
evrardjpMore news in the next meeting.14:59
evrardjpWe'll probably be able to start drafting ideas at that time, which will be notified on the MLs.14:59
evrardjpnext and last topic14:59
evrardjp#topic charter change15:00
*** openstack changes topic to "charter change (Meeting topic: tc)"15:00
evrardjpwe have charter change that we need to pass very quickly before next elections15:00
evrardjpit went through a few iterations already, and it would be nice if you could all vote15:00
evrardjpor propose a change15:00
gmann#link https://review.opendev.org/#/c/699277/15:00
jungleboyjLink?15:00
evrardjp#action tc-members review the charter change15:01
jungleboyjgmann:  Thx15:01
evrardjpthanks gmann :)15:01
zanebreminder that we need like 9 votes on this one15:01
gmann:)15:01
evrardjpyup, it's a charter-change, not formal vote :p15:01
evrardjpahem ahem15:01
* jungleboyj will vote now.15:01
zanebevrardjp: that wasn't directed at you ;)15:01
evrardjpok that's all we have for the meeting. Thanks everyone!15:01
evrardjpwe can now talk about telemetry in the office hours15:02
zanebjust that we need to round up everyone to vote15:02
evrardjp:)15:02
evrardjpthanks for the reminder zaneb15:02
evrardjpthanks everyone for your presence too, and let's continue to move things forward!15:02
evrardjp#endmeeting15:03
*** openstack changes topic to "OpenStack Technical Committee office hours: Tuesdays at 09:00 UTC, Wednesdays at 01:00 UTC, and Thursdays at 15:00 UTC | https://governance.openstack.org/tc/ | channel logs http://eavesdrop.openstack.org/irclogs/%23openstack-tc/"15:03
openstackMeeting ended Thu Jan 16 15:03:02 2020 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:03
openstackMinutes:        http://eavesdrop.openstack.org/meetings/tc/2020/tc.2020-01-16-14.00.html15:03
openstackMinutes (text): http://eavesdrop.openstack.org/meetings/tc/2020/tc.2020-01-16-14.00.txt15:03
ricolinevrardjp, thanks15:03
openstackLog:            http://eavesdrop.openstack.org/meetings/tc/2020/tc.2020-01-16-14.00.log.html15:03
gmannthanks evrardjp for chair.15:03
diablo_rojoThanks evrardjp!15:03
jungleboyjThanks!15:03
*** jamesmcarthur has joined #openstack-tc15:04
njohnstonThanks!15:06
mnaserso anyone wanna talk telemetry15:07
njohnstonsure15:09
jungleboyjhttps://gph.is/1VmlAcG15:09
zanebjungleboyj++15:10
* jroll is listening but may not have much to say15:11
fungii think using "we got mongodb working really well" as an argument isn't very effective, since we can't really recommend closed-source software for our users15:11
cloudnullo/15:12
mnaserright, so15:13
fungi(for those who may have forgotten, mongodb server releases after october 2018 are distributed under the "sspl" which is not an osi-approved license due to the additional restrictions it imposes)15:13
mnaserwe've moved to using prometheus based monitorng with a few things on top of it like thanos15:13
mnaserthat ecosystem is thriving15:13
fungihttps://www.mongodb.com/licensing/server-side-public-license15:13
gmannjungleboyj: that was good one :)15:14
mnaserand its very composable too, for example we have a libvirt exporter15:14
ricolinfungi, that's a good point15:14
mnaserand i have an openvswitch one that im thinking of putting out soon15:15
zanebI'm gonna repeat what I say every time15:15
mnaserso we can replace all of the ceilometer "collectors" with simple exporters already15:15
zanebTelemetry is a service that serves 3 different purposes: 1. operator monitoring; 2. billing; 3. tenant metrics (for e.g. autoscaling)15:16
zanebin retrospect that was a mistake as those have conflicting requirements15:16
zaneband probably should have been fulfilled by 2 or more different services15:16
zanebBUT15:17
zanebwe have to be careful not to talk about replacing it with the solution to only 1 or 2 of those problems15:17
jrollmnaser: fwiw, I'd love to see a little doc on what your telemetry infra looks like15:17
zanebbecause that isn't a replacement for all of its users15:17
jungleboyjzaneb: Good point.15:17
jungleboyjmnaser:  What do you use for Billing then?15:17
mnaserzaneb: i have gave this thought actually.  my idea was that we should make the data accessible/collectable, maybe several small composable systems that rely on external tooling15:18
mnaserfor example: to do billing for nova, yo uhave to figure out how many hours an instance has been running and what flavor for example.15:18
mnaserthat's also useful for an operator15:18
njohnstonMy concern is that we have something that listens to the notifications queues and makes events and metrics that nova, neutron, etc. push available.  I have worked on a project to create an internal showback system that used gnocchi extensively and should have listened for a few aodh alerts (although I was never able to get that working).15:18
mnaserif you also capture things like cpu usage, then voila, that info is useful for all 315:18
mnasernow all you need panko to do is evaluate queries against a prom server for things like autoscaling15:19
mnaseryour billing system queries that same system15:19
mnaserand your operator monitoring stuff can watch for system state (and again all we care are openstack-specific things)15:19
mnaserjroll: i should probably do a talk about it, we're actually making it a lot more generic so maybe we can share it at some point15:20
jroll<315:20
mnaserbut its all prometheus now, with small exporters everywhere (and mostly all open sourced too)15:20
mnaserhttps://github.com/vexxhost/libvirtd_exporter15:20
mnaserwe also have an openvswitch exporter that grabs network ids and port ids too in the prometheus labels15:20
evrardjptbh I like what I hear there15:20
jungleboyjevrardjp:  Makes sense.15:21
*** jamesmcarthur has quit IRC15:21
ttxback in business15:22
mnaserrealistically i hate to say it but the solution to our problems is15:22
ttxcatching up on logs15:22
mnaserstart a brand new thing and forget the existing stuff.  it is so difficult to "integrate" that with the existing stack15:22
njohnstonDo you think that we could get enough enthusiasm for a prometheus-based solution to form a next-generation Telemetry v2.0 team?15:22
fungithis also sounds like what the trove maintainers wanted to do a year or two ago, and the tc had objections to leaving users of the old solution behind and making an all new api15:23
mnaseri think next generation telemetry v2.0 might be mostly a reference architecture15:23
zanebmnaser: what's the migration path for existing users? that's the real question15:24
mnaserzaneb: none. unless someone is ready to do all the work of migrating old data in from the 8 different backends we supported over the years15:24
mnaseri think the problem is the fact we're not starting from scratc that makes it impossible15:25
zanebI don't care about old data. I care about e.g. Aodh keeps working15:25
zanebbecause it's integrated into all kinds of places15:25
ricolinI think whatever we plan, proper documentation and CI for most scenario (like autoscaling, etc.) is definitely needed15:26
fungii guess the difference might be that the trove team wanted to abandon support on the old api/solution in favor of a replacement. on the other hand having both solutions available and supported might not be as bad (that's more like what happened with nova-network and neutron)15:26
njohnstonI think a telemetry v2.0 team would really be a SIG, with the understanding that nothing would prevent you from doing both old and new - you could run Aodh and Prometheus at the same time without any interference.15:27
ttxnjohnston: ++15:27
evrardjp++15:27
ricolin++15:27
evrardjpbut then it means no deliverable by cycle, isn't it? :p15:27
evrardjpwhich could be fine15:27
ttxevrardjp: it could be a SIG for integration and project team for developing any missing component15:28
njohnstonEventually it would be nice if aodh/gnocchi/etc. could be changed so they could optionally be a compatibility shim in front of Prometheus, but I don't know how those line up against each other well enough to know if that's feasible15:28
ttxbut it's definitely an integration problem15:28
zanebyes, that's great, but Aodh gets its data from somewhere. nobody wants to run two lots of telemetry stacks15:28
evrardjpzaneb: how would heat integrate without aodh, by using metrics from prom?15:29
evrardjpis that possible?15:29
ttxI feel like if we can find a solution that works for both mnaser and zaneb we cover all our requirements15:29
zanebevrardjp: no, let's be clear. Aodh is a *user-facing* API. most of the integration in Heat (& Senlin) actually happens in userspace. you can't replace that with a reference design15:30
evrardjpoh yes I see15:30
jungleboyjttx ++15:31
njohnstonAodh gets it's data by listening to the notification queues in Rabbit, doesn't it?  I don't think we could have a full-featured solution without some way for Prometheus to get a feed of those events in some way.15:31
zanebfor the record, I think mnaser is right and I think prometheus is the future15:31
mnaserso my idea was actually having a small composable service15:31
mnasersomething that aodh can talk to and understand15:31
mnasereither make aodh understand native prom.. or have some translation/compatiblity layer15:32
zanebthat would be fine15:32
mnaserso we focus on simple code that implements an API as a thin layer.. instead of trying to engineer a time series storage db15:32
njohnston++15:32
zanebyes, please15:32
evrardjpwasn't that the supposed goal of the initial refactor ?15:32
evrardjpso that timeseries got outside (inside gnocchi)15:33
zanebevrardjp: well, they wrote their own time storage database. so... no?15:33
zaneb*series15:33
mnaseryeah gnocchi saying it was "just an api" was a disguise, esp for the ceph based stuff15:33
jrollyes, please no more NIH databases15:33
mnaserit was a full on implemetnation that manages to crash ceph clusters :)15:33
zanebin Python, no less15:34
mnaseranyways, i think the hardest part is there are people already super invested in ceilometer15:35
evrardjpI am confused about what's proposed15:36
mnaserand there isnt much interest in improving it15:36
evrardjpso to make sure I understand what you're proposing is... for 1) (operator monitoring) and 2. (billing), you're proposing your reference implementation, and for 3) (auto-scaling) you're proposing that there need to have an update to aodh to support having info from somewhere else, did I get that right?15:38
evrardjpme and telemetry ...15:39
jungleboyj:-)15:41
jrollso thinking about "migration paths", do we really need one? anyone that wants to use ceilometer/etc can continue to do so (and maintain it), anyone who doesn't want to can move to prom/etc?15:43
jrollcompetition isn't necessarily a bad thing15:43
evrardjpI am wondering where things overlap and will be causing frictions15:45
evrardjpthis is why I asked some clarifications above15:45
njohnstonWe specifically call out that we should not have two projects that try to do the same thing, but I think (hope) that having this be a SIG would sidestep that rule.15:45
njohnstonIIRC15:45
evrardjpI am not yet understanding the design proposed so I am not yet sure how this impacts the community as a whole15:45
evrardjpnjohnston: I don't think so15:45
*** ricolin has quit IRC15:46
evrardjp"Where it makes sense, the project cooperates with existing projects rather than gratuitously competing or reinventing the wheel" doesn't sound opposed15:47
evrardjp(see also https://governance.openstack.org/tc/reference/new-projects-requirements.html )15:48
njohnstonthanks, I was looking for the chapter and verse on that one15:48
evrardjpIt's something we should evaluate properly, but I think that the scope will need to be  crystal clear before accepting.15:49
jungleboyjSo is the interpretation that the SIG is or isn't gratuitously competing?15:49
njohnstonthe latter IMHO15:50
evrardjpI think it's not gratuitously competing, it's proposing a different kind of service/integration point15:50
jungleboyjevrardjp:  Ok, that was my interpretation as well.15:50
jrollfeels to me like a SIG for this should have an initial goal of experimenting and finding a reasonable solution (which mnaser has done some of), and until then we can't make any decisions about if it can replace or add on to the existing telemetry tooling15:51
evrardjpbut as I said above, that's with my simple limited understanding, so please keep in mind that I might change my mind based on clarifications.15:51
jrolland I don't see that as competing15:51
evrardjpwell I guess the question is "which sig" -- as it can be autoscaling (for point 3), and other sigs (for point 1 and 2)15:52
ttxdefinitely not gratuitously competing. It's to have a sane and future-proof solution15:53
evrardjpbut let's not tackle this, it's not a big problem to have, SIG/popup team, I don't really care about the naming... What I would like to see is the outcome :)15:53
jungleboyjttx: ++15:53
evrardjpttx: agreed15:53
njohnstonevrardjp++15:53
njohnstonSo is the next step for this to be proposed to the mailing list?15:59
jungleboyjMake sense.16:00
njohnstonIt would probably make sense to include a preliminary sketch of the technological direction this will be going, partially as a way to attract like-minded possible contributors.16:02
evrardjpI thin the conversation got lost now, and ppl moved to something else :p16:03
evrardjpthink*16:03
njohnstonhmm, probably so16:03
evrardjplet's see what will be done!16:03
evrardjpI have this in the agenda of next month, so we can keep track16:04
njohnstonsounds good, thanks!16:04
*** ianychoi_ has joined #openstack-tc16:05
*** ianychoi has quit IRC16:08
zanebI agree that there's nothing gratuitous about this competition16:08
openstackgerritMerged openstack/governance-sigs master: Add Tony Breeds as Multi-Arch SIG chair  https://review.opendev.org/70286716:18
openstackgerritMerged openstack/governance-sigs master: Add more details in 'advisory' SIG status  https://review.opendev.org/69895616:19
*** rpittau is now known as rpittau|afk16:27
ttxIt's actually the specific case we planned for when we wrote that sentence. Allow something revolutionary (in contrast to evolutionary) to be created when out toolset evolution drove us to a deadend16:30
evrardjpttx: glad you had it all planned :p16:37
*** dtroyer has quit IRC16:52
jungleboyjttx:  Good planning.17:01
*** slaweq has quit IRC17:01
*** slaweq has joined #openstack-tc17:02
zanebin fact the TC was thinking of this specific case (with Telemetry), except that Monasca was the competitor17:03
*** soumplis has quit IRC17:09
*** slaweq has quit IRC17:09
*** iurygregory has quit IRC17:14
*** e0ne has quit IRC17:17
*** slaweq has joined #openstack-tc17:20
*** slaweq has quit IRC17:25
*** evrardjp has quit IRC17:34
*** evrardjp has joined #openstack-tc17:34
fungithat's very much the case, yes17:47
*** witek has quit IRC18:30
*** slaweq has joined #openstack-tc18:31
*** slaweq has quit IRC18:36
*** e0ne has joined #openstack-tc18:44
*** openstackgerrit has quit IRC18:58
*** e0ne has quit IRC19:57
*** tosky has quit IRC20:04
*** slaweq has joined #openstack-tc20:38
*** slaweq has quit IRC21:09
*** jamesmcarthur has joined #openstack-tc21:11
*** jamesmcarthur has quit IRC21:17
*** jamesmcarthur has joined #openstack-tc21:18
*** jamesmcarthur has quit IRC21:25
*** jamesmcarthur has joined #openstack-tc22:13
*** ijolliffe has quit IRC22:16
*** jamesmcarthur has quit IRC22:19
*** jamesmcarthur has joined #openstack-tc22:28
*** jamesmcarthur has quit IRC22:35
*** jamesmcarthur has joined #openstack-tc22:47
*** jamesmcarthur has quit IRC22:48
*** ijolliffe has joined #openstack-tc22:50
*** jamesmcarthur has joined #openstack-tc22:52
*** smcginnis is now known as smcginnis|PTO22:52
*** jamesmcarthur has quit IRC22:55

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!