| *** openstack has joined #openstack-tc | 07:31 | |
| *** ChanServ sets mode: +o openstack | 07:31 | |
| *** tosky has joined #openstack-tc | 07:37 | |
| *** njohnston has quit IRC | 07:46 | |
| *** e0ne has joined #openstack-tc | 08:31 | |
| ttx | It's merged, but not released yet, so there is time to revert it | 08:31 |
|---|---|---|
| *** tetsuro has joined #openstack-tc | 08:58 | |
| *** tetsuro has quit IRC | 09:00 | |
| *** tetsuro has joined #openstack-tc | 09:03 | |
| *** markvoelker has joined #openstack-tc | 09:16 | |
| *** markvoelker has quit IRC | 09:21 | |
| *** tetsuro has quit IRC | 09:22 | |
| *** tetsuro has joined #openstack-tc | 09:25 | |
| *** tetsuro has quit IRC | 09:29 | |
| *** ricolin has quit IRC | 10:18 | |
| *** markvoelker has joined #openstack-tc | 10:27 | |
| *** markvoelker has quit IRC | 10:31 | |
| *** markvoelker has joined #openstack-tc | 11:18 | |
| *** markvoelker has quit IRC | 11:22 | |
| *** iurygregory has quit IRC | 11:49 | |
| *** markvoelker has joined #openstack-tc | 11:50 | |
| *** tkajinam has quit IRC | 11:56 | |
| *** iurygregory has joined #openstack-tc | 11:59 | |
| *** ricolin has joined #openstack-tc | 12:15 | |
| *** ricolin has quit IRC | 12:56 | |
| *** njohnston has joined #openstack-tc | 13:00 | |
| *** njohnston has quit IRC | 13:05 | |
| *** njohnston has joined #openstack-tc | 13:06 | |
| njohnston | o/ | 13:22 |
| *** bnemec has joined #openstack-tc | 13:38 | |
| *** tenbrae has joined #openstack-tc | 13:47 | |
| cloudnull | o/ | 13:51 |
| jungleboyj | o/ | 14:01 |
| *** ricolin has joined #openstack-tc | 14:21 | |
| *** Luzi has quit IRC | 14:34 | |
| *** Luzi has joined #openstack-tc | 14:48 | |
| *** Luzi has quit IRC | 14:50 | |
| knikolla | o/ | 15:06 |
| dansmith | knikolla: since I think you have the torch for this glance thing, | 15:07 |
| dansmith | knikolla: I spent some time last night looking into what the insurmountable challenges are preventing us from running glance in wsgi mode with all their async support in place | 15:07 |
| dansmith | knikolla: and from what I can tell, it was about two lines of code to hack it up and make it work (seriously) | 15:08 |
| knikolla | oh, wow | 15:08 |
| dansmith | knikolla: so I've started work on a proper change: https://review.opendev.org/#/c/742065/ | 15:08 |
| dansmith | knikolla: I know, I'm disappointed too, because I've been going on their word that it was like moving a mountain and pushing devstack changes to make standalone more of a first class citizen | 15:08 |
| knikolla | i'd be happy to help as well with review and code if we're that close to getting it done. | 15:09 |
| knikolla | ha, it got backported to train as well. https://review.opendev.org/#/c/742028/ | 15:10 |
| dansmith | knikolla: yeah, and my patch makes that doc backport even more of a sin, I think | 15:10 |
| dansmith | because now that we've retroactively made a bunch of deployments deprecated, | 15:10 |
| dansmith | we're going to be able to restore that as the default/recommended mode RSN | 15:11 |
| knikolla | yeah. docs change should be reverted while we work on a better solution, which may be yours. | 15:16 |
| knikolla | this aspect of "let me help" is one of the things i love the most about open source and open communities. | 15:16 |
| dansmith | well, fwiw, this help was offered before.. years ago, and they chased them away | 15:17 |
| smcginnis | dansmith: Do you have a link to that? | 15:18 |
| knikolla | that is true, but i think glance was in survival mode at that time. | 15:18 |
| dansmith | smcginnis: link to what? | 15:18 |
| smcginnis | My understanding was that the team didn't know how to approach it and no one else had stepped up to dig into it. | 15:18 |
| smcginnis | The previous attempt to offer help. | 15:18 |
| dansmith | smcginnis: so I remember mtreinish coming to me in Dublin at least, looking for threading help | 15:19 |
| dansmith | I didn't know anything about taskflow, | 15:19 |
| dansmith | and he also seemed like he was at the end of his rope trying to get them to help or work on it | 15:19 |
| dansmith | I don't know about any actual patches or anything like that, but I know he was working on it | 15:19 |
| dansmith | gmann might remember that | 15:19 |
| dansmith | to be clear, treinish was actively trying to make it work for them, and sought me out as someone who knew about eventlet and threading and things | 15:20 |
| smcginnis | I recall mteinish looking at it, not figuring out a way to make it work right, and then the current state. | 15:20 |
| dansmith | yeah | 15:20 |
| dansmith | him doing it completely on his own without help and willingness wasn't likely to succeed, which was my point about a previous offer to help | 15:21 |
| dansmith | smcginnis: I totally get the "they're just two people" thing, which is your point, I know | 15:25 |
| dansmith | here's my concern: | 15:25 |
| dansmith | 1. The -qa team has been kinda kicking this can for them for a long time to keep the WSGI_MODE= cruft available so they can still run standalone | 15:25 |
| dansmith | but despite that | 15:25 |
| dansmith | 2. They have NO tempest tests for the import API which is reportedly something they're planning to move everyone to instead of the current image-create thing | 15:26 |
| dansmith | so definitely some breakdown in communication | 15:26 |
| smcginnis | Totally agree it's a bad situation all around. I just hope we can focus on what we can do to help get the project to where it needs to be rather than spending much time focusing on past things that could have been handled better. Really really happy to see your patch dansmith. Fingers crossed that this might be a solution so we can finally get past all this. | 15:27 |
| smcginnis | (and then we can start pushing on OSC) ;) | 15:27 |
| knikolla | totally agree. | 15:28 |
| dansmith | yeah I mean, I dunno what to do retroactively, but I definitely think there's some concern over them being very small, very critical and maybe as a result of being a very small team, not a lot of oversight into losing sight of the bigger picture | 15:29 |
| dansmith | so, like, maybe they need a little more hand-holding than other projects that have a larger and more diverse team | 15:29 |
| knikolla | i do see an opportunity for us to improve the channels of communication. | 15:29 |
| smcginnis | Yeah... they've been on the "help wanted" list for a long time now. Doesn't seem to help. | 15:29 |
| smcginnis | They got an intern for a bit, but what the project really needs is someone with a bit more experience. | 15:30 |
| knikolla | we need more dansmiths | 15:30 |
| smcginnis | ++ | 15:30 |
| gmann | dansmith: smcginnis these are mtreinish effort as i remember - https://review.opendev.org/#/c/549743/and all comments in https://review.opendev.org/#/c/545483/5 | 15:30 |
| dansmith | knikolla: careful what you wish for :) | 15:30 |
| knikolla | smcginnis: has there been any success story with the "help needed" list? | 15:31 |
| knikolla | help wanted* | 15:31 |
| dansmith | gmann: ah yeah that first patch is pretty close | 15:31 |
| smcginnis | knikolla: Not that I know of. | 15:32 |
| dansmith | gmann: honestly, *plenty* close for someone to pick it up and run with it, if you ask me | 15:32 |
| gmann | but anyways let's get into what can be done and fix now than history. as dansmith figured out the things and with more testing we can get things forward | 15:32 |
| dansmith | gmann: that's process-based which I don't think is what they'd want for close-to-eventlet behavior, but... | 15:32 |
| smcginnis | And it was a regularly recurring topic for the BoD trying to get some attention and hopefully some companies willing to invest some resources towards it. | 15:32 |
| gmann | smcginnis: yeah but there are many projects in that situation and congress and tricircle are gone because of that. but i agree with you that "someone has to solve the less contributor issues at least for mandatory projects" | 15:34 |
| smcginnis | Yeah... the list is growing, not shrinking, unfortunately. | 15:35 |
| knikolla | it did evolve into the upstream investment opportunities guide https://governance.openstack.org/tc/reference/upstream-investment-opportunities/2020/index.html | 15:36 |
| gmann | knikolla: that is another question if we need to continue that or not as it has no result. | 15:37 |
| smcginnis | Yeah, that renaming made a bug difference. :) | 15:37 |
| smcginnis | *big | 15:37 |
| fungi | yeah, we rewrote them because the board members we tried to approach about the original help wanted list said it wasn't marketable to their organizations and we should instead present them with something they can use to convince business stakeholders | 15:37 |
| knikolla | i'm not too optimistic | 15:38 |
| knikolla | but i'm seeing a resurgence of openstack in academia | 15:38 |
| knikolla | and i think we should tailor it more towards that as well | 15:38 |
| smcginnis | ++ | 15:38 |
| knikolla | maybe do a "research opportunities" | 15:38 |
| smcginnis | There was some effort in Shanghai to get a more formal university program going. | 15:38 |
| fungi | i personally assumed those board members were simply making excuses for why they couldn't help us find experienced full-time long-term headcount | 15:39 |
| smcginnis | That's a great pipeline to get more people involved. | 15:39 |
| knikolla | in our cloud computing course we've always had a few projects related to upstream and had amrith and dims mentor | 15:39 |
| smcginnis | The problem is, we need more experienced people working on some things. But I suppose if we have some newer folks, that might free up some time for the experienced folks to spend on some things. | 15:39 |
| fungi | because i haven't really seen any difference in involvement after the rewrites | 15:39 |
| clarkb | smcginnis: also new people can eventually become experienced peopled | 15:40 |
| smcginnis | fungi: I try to be optimistic about things, but honestly think that's all a lost cause. | 15:40 |
| smcginnis | clarkb: Yep. Long term net positive. | 15:40 |
| fungi | clarkb: problem being we were losing experienced full-time long-term contributors due to changes in business direction from their employers, and in most cases couldn't find them somewhere to go and still be able to continue doing the same things | 15:41 |
| gmann | dansmith: let's test that more (I will help on that) and this is one we can discuss in adhoc meeting also, added your patch in etehrpad https://etherpad.opendev.org/p/tc-glance-wsgi | 15:42 |
| clarkb | fungi: yup, and there is no garuntee that inexperienced people will stick around | 15:42 |
| gmann | knikolla: did we finalize the meeting time? in case i missed | 15:42 |
| dansmith | gmann: "Dan patch" :P | 15:42 |
| dansmith | gmann: sounds good | 15:42 |
| gmann | heh | 15:43 |
| dansmith | gmann: abhi already tested locally and says it works.. I wish we had tempest tests to run against it, but we don't, other than that one we have proposed | 15:43 |
| dansmith | gmann: but yeah, at least devstack in the various configs | 15:43 |
| gmann | dansmith: ok, i will review that patch right after my email checks | 15:44 |
| knikolla | gmann: 3 tc haven't voted yet. | 15:44 |
| gmann | knikolla: ah sorry, i mean glance adhoc meetingh | 15:46 |
| gmann | meeting | 15:46 |
| knikolla | gmann: not yet. i'm holding off on organizing it since it seems we have no blockers and no disagreement yet on path forward. | 15:47 |
| gmann | knikolla: ok. | 15:49 |
| mnaser | shouldn't we have a space toa ctually discuss the next steps | 15:50 |
| mnaser | like what to revert | 15:50 |
| dansmith | knikolla: I think he just said he would not revert the backport, only the doc change in master | 15:51 |
| dansmith | knikolla: to me, the backports are the only ones that really matter, and I don't know how else to express my issues with that | 15:51 |
| knikolla | dansmith: is the async fix going to be backported? | 15:52 |
| knikolla | mnaser: okay. proposing a meeting during one of our office hours. | 15:54 |
| dansmith | knikolla: dunno, that's a glance core question.. so far it could be backported pretty easily, but it's definitely one of those changes where a whole set of things become possible that weren't before, which seems like it makes it not a candidate for stable | 15:55 |
| gmann | but even without async fix also, uwsgi in old stable used to work except import and few more features so doc patch was not at all backport candidate at first place. | 15:55 |
| dansmith | right and lots of people don't care about or even want to use import | 15:56 |
| knikolla | gmann, mnaser: does tomorrow during office hours work? | 15:56 |
| mnaser | i can make it | 15:56 |
| gmann | knikolla: works for me but i think we need to get glance team availability also ? | 15:56 |
| mnaser | we should ping the ml | 15:56 |
| knikolla | it's right after glance's weekly meeting | 15:56 |
| dansmith | that's way too early for me, but you don't need me | 15:56 |
| dansmith | knikolla: no you said thursday to them | 15:57 |
| dansmith | thursday I can make, and that's right after their meeting | 15:57 |
| dansmith | tomorrow is 0100 UTC, which is much earlier | 15:57 |
| knikolla | wait, it's Tuesday? what is wrong with me | 15:57 |
| gmann | :) that whats i was thinking | 15:57 |
| gmann | i am sure you are takling about thursday office hour | 15:57 |
| knikolla | yeah | 15:58 |
| dansmith | knikolla: are you thinking a voice call or irc? | 15:58 |
| knikolla | dansmith: i think we want you in it | 15:58 |
| dansmith | knikolla: happy to join, but I have little to offer I think | 15:58 |
| knikolla | dansmith: i don't know. what's the precedent for adhoc meetings? voice or irc? | 15:59 |
| dansmith | dunno, I'm just a guest here, I'm just curious | 16:01 |
| *** ricolin has quit IRC | 16:02 | |
| gmann | we did on irc in past but i prefer voice call for productivity and easiness but not sure if meeting has to be logged or not ? | 16:05 |
| smcginnis | If things can at least be captured in an etherpad, I think that's fine. | 16:06 |
| gmann | +1 | 16:06 |
| knikolla | the capturing into an auditable format is crucial. | 16:06 |
| gmann | i think summary on ML or etherpad is fine in voice call | 16:07 |
| mnaser | we've done adhoc over chat | 16:09 |
| mnaser | but tbh id prefer an audio discussion with notes | 16:09 |
| knikolla | mnaser: [all][tc][glance] sounds appropriate for tags? | 16:14 |
| *** ralonsoh_ has joined #openstack-tc | 16:41 | |
| *** ralonsoh has quit IRC | 16:43 | |
| *** tenbrae has quit IRC | 16:56 | |
| *** ralonsoh_ has quit IRC | 17:33 | |
| *** ralonsoh has joined #openstack-tc | 17:41 | |
| *** ralonsoh has quit IRC | 18:08 | |
| mnaser | knikolla: late but yes | 20:27 |
| *** e0ne has quit IRC | 20:27 | |
| *** adriant has quit IRC | 20:39 | |
| *** adriant has joined #openstack-tc | 20:40 | |
| *** bnemec has quit IRC | 22:47 | |
| *** tkajinam has joined #openstack-tc | 22:55 | |
| *** iurygregory has quit IRC | 23:42 | |
| *** tosky has quit IRC | 23:47 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!