| *** bnemec has quit IRC | 00:41 | |
| *** bnemec has joined #openstack-tc | 00:46 | |
| mnaser | lxkong: yeah tbh i see gnocchi as the nova of tsdb | 00:53 |
|---|---|---|
| mnaser | feed in via one api and store/use different backends | 00:53 |
| *** kberger_ has quit IRC | 01:13 | |
| *** bnemec has quit IRC | 01:34 | |
| *** bnemec has joined #openstack-tc | 01:39 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #openstack-tc | 04:33 | |
| *** belmoreira has joined #openstack-tc | 04:53 | |
| *** belmoreira has quit IRC | 05:26 | |
| *** belmoreira has joined #openstack-tc | 05:34 | |
| *** ralonsoh has joined #openstack-tc | 06:14 | |
| *** slaweq has joined #openstack-tc | 06:31 | |
| *** e0ne has joined #openstack-tc | 06:57 | |
| *** e0ne has quit IRC | 07:00 | |
| *** yoctozepto has quit IRC | 07:34 | |
| ttx | lxkong: I thought they dropped the modular storage backend side... but then maybe I'm not current | 07:37 |
| *** e0ne has joined #openstack-tc | 07:39 | |
| ttx | I agree it's more than just a DB, which is the main reason why the situation is so difficult to untangle | 07:41 |
| *** tosky has joined #openstack-tc | 07:47 | |
| *** yoctozepto has joined #openstack-tc | 08:55 | |
| mnaser | ttx: pretty sure it supports ceph, local file and a bunch of other options afaik | 11:50 |
| mnaser | ttx: https://github.com/gnocchixyz/gnocchi/tree/master/gnocchi/storage | 11:51 |
| *** dklyle has quit IRC | 12:24 | |
| knikolla | o/ | 13:25 |
| *** bnemec has quit IRC | 14:11 | |
| *** bnemec has joined #openstack-tc | 14:12 | |
| ricolin | o/ | 14:26 |
| gmann | o/ | 14:27 |
| *** dklyle has joined #openstack-tc | 14:41 | |
| evrardjp | o/ | 15:00 |
| mnaser | so | 15:01 |
| mnaser | how about letting folks who want to maintain gnocchi | 15:01 |
| mnaser | maintain it inside gerrit | 15:01 |
| mnaser | and stop letting that decision be dictated by someone who's not actually maintaining | 15:03 |
| evrardjp | I just read the answer from Zane, very nice educative links :) | 15:05 |
| *** dklyle has quit IRC | 15:05 | |
| ttx | zaneb: heh -- don't let facts get in the way of a good Twitter shaming | 15:06 |
| zaneb | I think JD needs to be clear about what his intentions are | 15:08 |
| evrardjp | ttx: lol | 15:08 |
| mnaser | he has no intentions other than blocking the progress of the project at this point | 15:08 |
| evrardjp | zaneb: agreed | 15:08 |
| zaneb | on GitHub he appears to be saying the project is dead, in which case fork it, but apparently that's not what he meant to say | 15:09 |
| ttx | I guess part of the issue is that JD still wants to paint Gnocchi's move off opendev as the best decision he ever took. Which is increasingly proved wrong by facts | 15:09 |
| evrardjp | I hope there is no shame at being wrong, as long as you learn | 15:09 |
| zaneb | is he going to actively review stuff (along with other existing maintainers)? then forking would be the wrong thing | 15:09 |
| ttx | and yes he can't have it two ways... say "you're on your own" on GitHub and accuse people of theft on Twitter | 15:09 |
| mnaser | he is not actively reviewing stuff | 15:10 |
| mnaser | he has rejected pull requests | 15:10 |
| mnaser | let me find them | 15:10 |
| evrardjp | so, let's just be nice, and ask if he wants to contribute now. | 15:10 |
| mnaser | https://github.com/gnocchixyz/gnocchi/pull/1065 | 15:10 |
| gmann | i agree, if no maintainer for current repo then let it take over by new interested maintainers | 15:10 |
| evrardjp | Maybe things have changed. | 15:10 |
| mnaser | https://github.com/gnocchixyz/gnocchi/pull/1064 | 15:10 |
| mnaser | plenty of unreviewed things | 15:10 |
| evrardjp | Because I interpreted his reaction as a "please help me". | 15:11 |
| zaneb | is he going to allow other people to become maintainers on GitHub but not help out himself, thus forcing them to use his preferred tools without actually contributing himself? Everything about that situation would suck, but if that were the case it'd be hard to put the blame on people who just want to fork | 15:11 |
| mnaser | https://github.com/gnocchixyz/gnocchi/issues/1049 | 15:11 |
| mnaser | it's not a please help me | 15:11 |
| ttx | If we forked it on gitHub he would not even react. But if we paint it as a revert of the GitHub move, he feels obliged to object. | 15:11 |
| mnaser | it's a: i don't care about the project anymore | 15:11 |
| mnaser | https://github.com/gnocchixyz/gnocchi/issues/1049#issuecomment-577682676 | 15:11 |
| evrardjp | why tweeting and be vocal about it if you don't care? | 15:11 |
| mnaser | because | 15:11 |
| mnaser | he has something against the openstack community | 15:12 |
| mnaser | and he's forever unhappy with us | 15:12 |
| ttx | evrardjp: because he can't be proven wrong | 15:12 |
| gmann | yeah, history :) | 15:12 |
| mnaser | so honestly, i have no time or patience of the day to deal with someone who just has one goal: bring us down | 15:12 |
| zaneb | also I think we should call a fork a fork | 15:12 |
| jungleboyj | o/ | 15:12 |
| mnaser | yep. it's a fork then. | 15:12 |
| evrardjp | zaneb: +10000 | 15:12 |
| zaneb | the proposal isn't to move it back, it's to fork the existing project | 15:12 |
| ttx | zaneb: agreed, the talk of "moving" has caused the issue | 15:12 |
| gmann | +1 | 15:12 |
| mnaser | fork gnocchi into opendev | 15:12 |
| evrardjp | let's call it pasta | 15:13 |
| evrardjp | it's not quite pasta, it's not quite potato | 15:13 |
| mnaser | how was our telemetry ptl things been like? | 15:13 |
| evrardjp | Folks I am making jokes to chill out the mood | 15:13 |
| zaneb | and this appears to be his main objection: https://twitter.com/juldanjou/status/1298991377875759113 | 15:13 |
| mnaser | zzz | 15:13 |
| mnaser | he just wants to argue, i want to get shit done | 15:13 |
| ttx | evrardjp: I'm chill -- my expectations were very low in the first place | 15:14 |
| mnaser | let me quote him: "As somebody said once: talk is cheap. Show me the code. 😉" | 15:14 |
| mnaser | so, let's fork it and let the fine folks who want to help maintain it do that | 15:14 |
| smcginnis | "We never prevented anyone to send PR to this repository and try to fix the CI or improve things. I just never happened." Isn't that the issue? That it has happened but no one would review or merge the PRs? | 15:14 |
| mnaser | smcginnis: there's open changes to fix the CI and improve things from tobias-urdin since june/july | 15:14 |
| mnaser | they went unreviewed | 15:14 |
| fungi | it seems patently unfair to claim that when new maintainers of a fork would want to host it alongside where they do their other work, that's ignorant of free software dynamics. i could list countless cases of that dating back decades. also it's the height of hypocrisy to complain about that when it's *exactly* what he did when he moved gnocchi out of openstack | 15:14 |
| ttx | JD, unfair? nooooooo | 15:16 |
| mnaser | so then the question remains | 15:16 |
| mnaser | fork into openstack or fork into opendev | 15:16 |
| fungi | but at this point it's probably best to let him rant and get it over with, just ignore him at this point | 15:16 |
| zaneb | CI appears to be the sticking point | 15:17 |
| fungi | also might be a good idea to change the name when forking, maybe tortellini or something | 15:17 |
| mnaser | right, the contributors want to get access to Zuul resources | 15:17 |
| mnaser | and gerrit for reviews/etc | 15:17 |
| zaneb | afaik you can use Zuul with GitHub, I'm not sure that's the problem exactly | 15:18 |
| zaneb | the problem is... | 15:18 |
| evrardjp | fungi: haha that's what I said above :) | 15:18 |
| mnaser | zaneb: opendev doesn't provide ci for projects in github | 15:18 |
| zaneb | 1) no PRs are being reviewed because CI is broken | 15:18 |
| mnaser | or that's not a goal of opendev afaik | 15:18 |
| clarkb | opendev does third party ci so could test gnocchi's ceilometer and other integrations | 15:19 |
| clarkb | but we wouldn't gate gnocchi | 15:19 |
| mnaser | yeah, i think they're looking for gating | 15:19 |
| knikolla | i really don't want to attribute things to malice, but what is really the end goal here in not letting go an unmaintained project? | 15:19 |
| zaneb | 2) people are being asked to fix a CI that doesn't benefit them (they need Zuul), and was hand-rolled by the current maintainers, without the help of the current maintainers | 15:19 |
| zaneb | so it's basically a deadlock | 15:20 |
| mnaser | 3) people are being asked to use a tool which they don't want to | 15:20 |
| mnaser | (i.e. github) | 15:20 |
| fungi | important questions on the in-or-out-of-openstack point are probably 1. is there a team ready to adopt it or form around it in an official capacity, and 2. would it want to use openstack-only resources like release management reviews, publication on docs.openstack.org/tarballs.openstack.org | 15:20 |
| zaneb | combined with statements that the project is unmaintained, it's basically like hanging out a big "Go Away" sign | 15:20 |
| zaneb | at least they eventually renewed the domain registration? iirc even that had expired at some point but maybe I am imagining it | 15:21 |
| mnaser | yes it did expire for a while | 15:21 |
| mnaser | and the domain lapsed i think and now is hosted on some oisc or something subdomain | 15:21 |
| mnaser | ok so for the context of openstack vs opendev, we have had to appoint a ptl for telemetry in train | 15:21 |
| zaneb | domain is back and redirecting to that | 15:21 |
| zaneb | but not linked from github anymore | 15:22 |
| evrardjp | knikolla: some people still want that project and are willing to contribute, but they can't... :) | 15:22 |
| smcginnis | Annoying that the community is being put down for trying to work with the team and avoid creating forks. | 15:22 |
| knikolla | evrardjp: i mean from a jd perspective. | 15:22 |
| mnaser | i'd love it if we focused on a progressive solution forward, which the first one being: do we fork into openstack (as a project inside telemetry team) or opendev (ungoverned?) | 15:22 |
| evrardjp | knikolla: got it. This is what I couldn't understand, and why I asked a few questions above. | 15:22 |
| evrardjp | let's see what happens on that code review | 15:23 |
| mnaser | i know ttx had feelings that it could live inside opendev but doesn't feel like it's a good idea in openstack | 15:23 |
| gmann | mnaser: with PTL-less model (after we merge that resolution), PTL role will not be issue | 15:23 |
| evrardjp | I have no strong opinion on opendev vs openstack | 15:23 |
| ttx | I just... want to increase the odds that we ultimately move away from it, rather than double down on it | 15:23 |
| mnaser | ttx: i think that's a very reasonable thing | 15:24 |
| evrardjp | ttx: that's totally fair, and right | 15:24 |
| knikolla | ++ | 15:24 |
| evrardjp | mnaser: : ) | 15:24 |
| mnaser | so IMHO | 15:24 |
| gmann | what interested maintainer thinks? | 15:24 |
| mnaser | we ignore all the noisy, ask the current people looking to maintain it to come up with a cool name, and fork it into its own namespace inside opendev | 15:25 |
| ttx | Forking it on opendev is a short-term workaround. Re-including it as an openstack service is a stronger long-term commitment in my book | 15:25 |
| mnaser | https://review.opendev.org/#/c/744592/ | 15:25 |
| mnaser | lxkong is looking to help maintain | 15:25 |
| knikolla | also, i think forking and renaming, would add more uncertainty to what has already been a complex issue. how do you monitor openstack? monasca, ceilometer, gnocchi fork with a different name | 15:25 |
| mnaser | alongside adriant | 15:25 |
| mnaser | i honestly have 0 interest in finding an amicable resolution with jd, there just is none | 15:25 |
| clarkb | to be clear the proposed change there is a fork. A new name isn't necessary to fork (as github loves to tell you) | 15:26 |
| mnaser | let's not burn out the people who want to write code by having to fight and be called names on public | 15:26 |
| gmann | +1 on ttx message -> "Re-including it as an openstack service is a stronger long-term commitment in my book" | 15:26 |
| clarkb | whether you want a new name is a separate concern | 15:26 |
| mnaser | i've never been around forks so i dont know if like | 15:27 |
| mnaser | using the same name for a fork is 'not an okay thing' | 15:27 |
| ttx | mnaser: it's not an OK thing if the original is still maintained/released | 15:28 |
| clarkb | in this case its probably a good thing to reduce confusion. But if you were to fork on say github you'd likely keep the same name since that is the github default | 15:28 |
| ttx | and you intend to release things as well | 15:28 |
| mnaser | hmm | 15:29 |
| zaneb | one issue with a fork without changing the name is that the PyPI package is still presumably controlled by the people that the TC chair just said he wasn't interested in finding an amicable resolution with | 15:29 |
| mnaser | but gnocchi is currently listed as unmaintained | 15:29 |
| mnaser | i don't think they have interest either | 15:29 |
| knikolla | does gnocchi have widespread usage in production clouds? | 15:29 |
| mnaser | if someone wants to go and drive this, go for it, but not at the interest of a bunch of folks who _actually_ want to fix gnocchi | 15:29 |
| evrardjp | in any case, there should be no conflict. We shouldn't talk like this. We should at the opposite, talk with JD, and ensure the real intentions for him are to not maintain it, so we can make peace and fork appropriately | 15:30 |
| mnaser | i think catalyst does use it hence the names of the people who are putting their names on to help maintain | 15:30 |
| ttx | In our case the issue would likely be PyPI gnocchi namespace. If we have to change that, I'd just rename it | 15:30 |
| evrardjp | we don't want pending conversations and oil on fire | 15:30 |
| evrardjp | else it's gonna come back | 15:30 |
| mnaser | evrardjp: its gonna come back, if we fix it or not | 15:30 |
| evrardjp | we want peace and serenity for the maintainers | 15:30 |
| mnaser | jd has historically and repeatedly came after openstack, time after time | 15:31 |
| mnaser | if we're going to sit and try and go through this, we're going to have people lose interest and be discouraged from gnocchi | 15:31 |
| evrardjp | I didn't say to not work on the fork | 15:31 |
| ttx | Yeah i would not bet money on getting to a common agreement | 15:31 |
| knikolla | I think there are two points here: 1) what is our story wrt monitoring? is gnocchi widespread and the defacto solution? are there other widely used products? 2) if gnocchi is it, how do we move forward with it, if gnocchi is not, how do we move away from it | 15:33 |
| ttx | At best I would try getting JD to clarify whether he intends to review Gnocchi PRs or just let them rot like he has for the last year(s) | 15:33 |
| evrardjp | ttx: this is what I meant | 15:33 |
| evrardjp | but better phrased | 15:34 |
| mnaser | but what if the people pushing code want to be doing it on gerrit? | 15:34 |
| evrardjp | if possible :p | 15:34 |
| mnaser | :\ | 15:34 |
| evrardjp | they can | 15:34 |
| evrardjp | they just fork | 15:34 |
| ttx | But I doubt he would answer a clear "yes" or "no" | 15:34 |
| mnaser | and that's what they want to do | 15:34 |
| evrardjp | so let's go ahead | 15:34 |
| zaneb | ttx: I predict that the answer will be the CI deadlock I mentioned above | 15:34 |
| fungi | usually the reason to consider forking under a different name is to avoid "brand confusion" with the original project (so that someone can't claim we're trying to convince people it's the same when it's possibly changed direction under new management) | 15:34 |
| evrardjp | fungi: this is exactly it | 15:34 |
| evrardjp | new management, possible new future, let's just fork | 15:35 |
| ttx | zaneb: mnaser said there were Ci fixes proposed, and those went unreviewed... So it's not a deadlock, it's just a lock | 15:35 |
| fungi | also it's possible (but perhaps unlikely) that jd has filed for a trademark on the name, which *could* then present additional legal liability for the project/contributors if they knowingly took the same name | 15:35 |
| zaneb | ttx: CI fixes did not pass CI, it was a very preliminary patch | 15:36 |
| evrardjp | I just meant above that we'll have complaints "Now it's forked, they didn't even try to contribute, blabla" which will be obviously wrong, so before having a fire and oil, I just wanted to have JD say that he will maintain or not that software now. | 15:36 |
| ttx | zaneb: ah sigh | 15:36 |
| zaneb | ttx: https://github.com/gnocchixyz/gnocchi/pull/1065 | 15:36 |
| mnaser | evrardjp: he's explicitly said "this project is not maintained" | 15:36 |
| evrardjp | or share the credentials | 15:36 |
| mnaser | adriant, lxkong: can you update your patch to put a new name that you feel suits the projects, put it in a seperate namespace (i.e. `foobar/foobar`), not under `openstack/` ? | 15:36 |
| knikolla | adriant is NZ timezone | 15:37 |
| mnaser | yeah figured that was an async irc message :) | 15:37 |
| mnaser | ok, left a review describing the same | 15:39 |
| knikolla | i am not sure that forking is the right decision long term, but short term we do need to ensure that what we depend on is at the very least tested. | 15:41 |
| knikolla | and that there is a path to merge fixes, be they security or smth else | 15:42 |
| mnaser | this isn't what i want us to d | 15:42 |
| mnaser | do | 15:42 |
| mnaser | it's not the ideal path but it's the best one given the circumstances, IMHO. | 15:43 |
| knikolla | mnaser: what would an ideal path look like? | 15:46 |
| mnaser | jd agreeing that he has no time to maintain the project because he's moved on, handing over the keys to the people who want to maintain the project and letting them move it into opendev | 15:46 |
| knikolla | with the project passing over to someone else's maintainership in opendev, but not be under openstack governance? | 15:49 |
| mnaser | yes knikolla | 15:53 |
| knikolla | okay, makes sense to me | 15:55 |
| tosky | I have a more trivial question (now that the discussion seems to have ended or so) | 16:11 |
| tosky | just to be sure: who is reponsible for the osf/groups repository, and do community goal affect those osf/ repositories? There are a few legacy jobs (not devstack-gate ones, so not too urgent) | 16:11 |
| tosky | (I may have asked it already) | 16:11 |
| gmann | tosky: i do not think community goal scope cover the non openstack namesapce repo which include osf/ repo also | 16:16 |
| gmann | but you can reachout to individual repo maintainer if they want to move to zuulv3 native jobs. | 16:17 |
| tosky | thanks; sure, I may nag them too | 16:19 |
| tosky | I was just not sure whether osf was considered somehow special | 16:19 |
| gmann | tosky: btw which repo, is it interop ? | 16:20 |
| tosky | those osf/groups jobs are not devstack-gate anyway | 16:20 |
| tosky | osf/groups :) | 16:20 |
| gmann | ok | 16:20 |
| *** e0ne has quit IRC | 16:20 | |
| *** bnemec has quit IRC | 16:22 | |
| *** bnemec has joined #openstack-tc | 16:30 | |
| clarkb | tosky: the osf/groups project was used to manage user groups and sunsetted as many user groups were using meetup anyway (the software was built by foundation contractors | 16:30 |
| clarkb | at this point I think its largely in archival mode and jobs can either be ignored or deleted from it | 16:31 |
| clarkb | fungi and ttx ^ can probably confirm that | 16:31 |
| tosky | I see, thanks! | 16:31 |
| ttx | osf/ repositories are under the OSF, not openstack, so not affected by community goals | 16:34 |
| fungi | yes, osf/groups was the code backing the groups.o.o site for coordinating user groups, but osf eventually dropped the development contract on that in favor of just using a meetup pro account, since most of the user groups were already using meetup anyway, this allowed osf to help shoulder some of that | 16:34 |
| ttx | we shoudl probably retire that one | 16:34 |
| fungi | i agree, it could be retired | 16:34 |
| fungi | there are likely at least a few osf repos which could be retired | 16:35 |
| *** tosky has quit IRC | 16:45 | |
| *** belmoreira has quit IRC | 16:52 | |
| *** ralonsoh has quit IRC | 17:12 | |
| *** e0ne has joined #openstack-tc | 18:09 | |
| *** e0ne has quit IRC | 18:27 | |
| *** tosky has joined #openstack-tc | 18:35 | |
| *** belmoreira has joined #openstack-tc | 18:55 | |
| *** slaweq has quit IRC | 21:22 | |
| *** slaweq has joined #openstack-tc | 21:29 | |
| *** slaweq has quit IRC | 21:34 | |
| *** belmoreira has quit IRC | 21:56 | |
| *** bnemec has quit IRC | 22:07 | |
| *** bnemec has joined #openstack-tc | 22:12 | |
| *** bnemec has quit IRC | 23:37 | |
| *** bnemec has joined #openstack-tc | 23:43 | |
| *** tosky has quit IRC | 23:59 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!