Thursday, 2018-12-20

*** tosky has quit IRC00:43
*** jamesmcarthur has joined #openstack-tc00:53
*** jamesmcarthur has quit IRC00:58
*** jamesmcarthur has joined #openstack-tc01:17
*** jaypipes_ has joined #openstack-tc01:18
*** mriedem has quit IRC01:18
*** jaypipes has quit IRC01:20
*** jamesmcarthur has quit IRC01:22
*** jamesmcarthur has joined #openstack-tc02:15
*** jamesmcarthur has quit IRC02:17
*** jamesmcarthur has joined #openstack-tc02:17
*** gagehugo has quit IRC02:33
*** Bhujay has joined #openstack-tc02:40
*** Bhujay has quit IRC02:47
*** ricolin has joined #openstack-tc02:47
*** Bhujay has joined #openstack-tc02:47
*** Bhujay has quit IRC02:48
*** Bhujay has joined #openstack-tc02:49
*** Bhujay has quit IRC02:50
*** Bhujay has joined #openstack-tc02:50
*** Bhujay has quit IRC02:51
*** Bhujay has joined #openstack-tc02:52
*** Bhujay has quit IRC02:53
*** Bhujay has joined #openstack-tc02:53
*** Bhujay has quit IRC02:54
*** Bhujay has joined #openstack-tc02:55
*** Bhujay has quit IRC02:56
*** Bhujay has joined #openstack-tc02:56
*** Bhujay has quit IRC02:57
*** Bhujay has joined #openstack-tc02:58
*** Bhujay has quit IRC02:59
*** Bhujay has joined #openstack-tc02:59
*** Bhujay has quit IRC03:00
*** Bhujay has joined #openstack-tc03:01
*** Bhujay has quit IRC03:02
*** Bhujay has joined #openstack-tc03:02
*** Bhujay has quit IRC03:03
*** Bhujay has joined #openstack-tc03:04
*** Bhujay has quit IRC03:05
*** diablo_rojo has quit IRC03:05
*** Bhujay has joined #openstack-tc03:05
*** Bhujay has quit IRC03:06
*** Bhujay has joined #openstack-tc03:07
*** Bhujay has quit IRC03:08
*** jamesmcarthur has quit IRC03:08
*** Bhujay has joined #openstack-tc03:08
*** Bhujay has quit IRC03:09
*** Bhujay has joined #openstack-tc03:10
*** Bhujay has quit IRC03:15
*** jamesmcarthur has joined #openstack-tc03:34
*** Bhujay has joined #openstack-tc03:41
*** Bhujay has quit IRC03:46
*** jamesmcarthur has quit IRC04:29
*** Bhujay has joined #openstack-tc05:02
*** jamesmcarthur has joined #openstack-tc06:30
*** mgagne has quit IRC06:33
*** jamesmcarthur has quit IRC06:34
*** mgagne has joined #openstack-tc06:40
*** gagehugo has joined #openstack-tc06:42
*** Bhujay has quit IRC07:13
*** ricolin has quit IRC07:14
*** Bhujay has joined #openstack-tc07:23
*** Bhujay has quit IRC07:24
*** Bhujay has joined #openstack-tc07:24
*** Bhujay has quit IRC07:26
*** Bhujay has joined #openstack-tc07:26
*** Bhujay has quit IRC07:27
*** Bhujay has joined #openstack-tc07:27
*** Bhujay has quit IRC07:29
*** Bhujay has joined #openstack-tc07:29
*** Bhujay has quit IRC07:30
*** Bhujay has joined #openstack-tc07:30
*** jamesmcarthur has joined #openstack-tc07:31
*** Bhujay has quit IRC07:32
*** Bhujay has joined #openstack-tc07:32
*** Bhujay has quit IRC07:33
*** Bhujay has joined #openstack-tc07:33
*** Bhujay has quit IRC07:35
*** jamesmcarthur has quit IRC07:35
*** Bhujay has joined #openstack-tc07:35
*** Bhujay has quit IRC07:36
*** Bhujay has joined #openstack-tc07:36
*** Bhujay has quit IRC07:38
*** Bhujay has joined #openstack-tc07:38
*** Bhujay has quit IRC07:39
*** Bhujay has joined #openstack-tc07:39
*** Bhujay has quit IRC07:41
*** Bhujay has joined #openstack-tc07:41
*** cdent has joined #openstack-tc07:42
*** Bhujay has quit IRC07:42
*** Bhujay has joined #openstack-tc07:42
*** Bhujay has quit IRC07:44
*** Bhujay has joined #openstack-tc07:44
*** Bhujay has quit IRC07:45
*** Bhujay has joined #openstack-tc07:45
*** Bhujay has quit IRC07:47
*** Bhujay has joined #openstack-tc07:47
*** Bhujay has quit IRC07:48
*** Bhujay has joined #openstack-tc07:48
*** Bhujay has quit IRC07:50
*** Bhujay has joined #openstack-tc07:50
*** Bhujay has quit IRC07:51
*** Bhujay has joined #openstack-tc07:51
*** Bhujay has quit IRC07:53
*** Bhujay has joined #openstack-tc07:53
*** Bhujay has quit IRC07:54
*** Bhujay has joined #openstack-tc07:54
*** Bhujay has quit IRC07:56
*** Bhujay has joined #openstack-tc07:56
*** Bhujay has quit IRC07:57
*** Bhujay has joined #openstack-tc07:57
*** Bhujay has quit IRC07:59
*** Bhujay has joined #openstack-tc07:59
*** Bhujay has quit IRC08:00
*** Bhujay has joined #openstack-tc08:00
*** Bhujay has quit IRC08:02
*** Bhujay has joined #openstack-tc08:02
*** Bhujay has quit IRC08:03
*** Bhujay has joined #openstack-tc08:03
*** Bhujay has quit IRC08:05
*** Bhujay has joined #openstack-tc08:05
*** Bhujay has quit IRC08:06
*** Bhujay has joined #openstack-tc08:07
*** tosky has joined #openstack-tc08:24
*** Bhujay has quit IRC08:46
*** jpich has joined #openstack-tc08:54
*** jpich has quit IRC09:04
*** jpich has joined #openstack-tc09:05
*** ricolin has joined #openstack-tc09:12
*** jpich has quit IRC09:21
*** jpich has joined #openstack-tc09:23
*** alexchadin has joined #openstack-tc09:23
*** e0ne has joined #openstack-tc09:53
*** Bhujay has joined #openstack-tc10:23
*** e0ne has quit IRC10:29
*** e0ne has joined #openstack-tc10:30
*** cdent_ has joined #openstack-tc11:44
*** cdent has quit IRC11:46
*** cdent_ is now known as cdent11:46
*** e0ne has quit IRC13:19
*** Bhujay has quit IRC13:31
*** e0ne has joined #openstack-tc13:33
*** mriedem has joined #openstack-tc13:38
*** e0ne has quit IRC14:11
*** e0ne_ has joined #openstack-tc14:11
mriedemare there office hours now or are those dead, or is my calendar wrong?14:28
dhellmannmriedem : top of the hour14:29
dhellmannbut if you want to start early, that's fine too14:29
mriedemDST14:29
mriedemdamn14:29
mriedemjust have some stuff from the public cloud wg meeting, i'll relay during office hours14:31
smcginnisIt doesn't necessarily need to wait, but there may be more TC folks around (or paying more attention) in 30 minutes.14:32
fungiyeah, starts 1500z14:33
fungiso 9am your time i think?14:33
mriedemyeah14:33
mriedemi'm adjusting each calendar entry as i miss each meeting14:33
* fungi has an hour head-start on caffeine14:33
dhellmannmriedem : if you're using google's calendar, using the Reykjavík time zone will give you effectively UTC because they are GMT without DST14:35
dhellmannother tools seem to support UTC directly14:35
mriedemthunderbird14:35
* dhellmann is more of a mad dog 20/20 man, himself14:36
mriedemorange flavored booze, classy14:36
mriedemi also wonder how many times doug has made that exact joke14:37
dhellmannexactly 1 time14:37
dhellmannso thank you for the set up14:37
smcginnisHah14:37
*** zhipeng has joined #openstack-tc14:49
mriedemok so i'll dump this quick, in the public cloud wg meeting howard and tobias were talking about how it'd be nice to have something like https://releases.openstack.org/rocky/highlights.html but for community wide goals, to see progress on a goal within a given release,14:54
*** alexchadin has quit IRC14:55
mriedemwhich i've thought about in the past also - i'm not sure if there is any place that shows completion percentage of a goal14:55
mriedembut i think that would be useful14:55
mriedemi think they might have also been thinking about highlights for maybe non-community wide goals but things tracked within the public cloud sig that could show up in ^; i explained that doc is based on deliverables which the pc sig doesn't really have, so i'm not sure that is the best place for that infor14:56
mriedemi think to date, some projects just highlight individually if they have completed some goal14:57
mriedeme.g. the nova pike release notes just called out support for running the api under uwsgi14:57
zhipengI don't think pc need to add deliverables for that, just related projects add the goal completion status14:59
smcginnisTis time15:01
dhellmanno/15:01
lbragstado/15:02
lbragstadmriedem i like that idea... we had some tooling to that kind of stuff that dhellmann generalized a bit, but it was more focused on project management parts of it15:04
mriedemthe goal-tools repo?15:05
mriedemyeah15:05
mriedemi have to run, but i've dumped the thing so tc people can discuss it15:05
lbragstadyeah15:05
lbragstadhttps://www.lbragstad.com/policy-burndown/15:05
dhellmannthe release management team used to try to track the work done in a cycle by looking at blueprints in launchpad, but we gave that up because it was a lot of work for 1-2 people (at the time) and we didn't think anyone was really using it15:05
dhellmannwe're using storyboard for tracking goal completion now, and yeah, lbragstad and I have talked about burndown charts in the past15:06
lbragstadi would think if you render an accurate burndown, building a representation for that with just basic percentages wouldn't be too difficult15:06
lbragstads/representation/view/15:07
ttxo/15:08
dhellmannI do think it makes sense to encourage teams to include info like this in release highlights and to write more summaries in the prelude sections of their release notes15:08
dhellmannand I would be happy to see someone interested in collecting the information more centrally15:08
ttxmriedem: i remember some goal champions doing burn-down charts15:08
ttxah, lbragstad linked to them15:09
dhellmannwe started to try to make the code sdague wrote for those more generic15:09
dhellmannyeah15:09
lbragstadit's ironic, since neutron just merged support for policy in code and glance is in the process of reviewing it15:10
lbragstadbut i think that'd be the kind of thing that is useful to look at over multiple releases15:10
zhipengBoth burndowm and storyboard tracking looks good for our goal :)15:11
zhipengAs long as we could effectively track it15:11
dhellmanneach goal document links to the storyboard story tracking that goal15:11
*** mriedem is now known as mriedem_afk15:11
dhellmannsee the top of the python3 goal for an example: https://governance.openstack.org/tc/goals/stein/python3-first.html15:12
dhellmannthat one had enough steps that there is 1 story per team15:12
dhellmannis anyone else following the OSF OIP guidelines work that the board is doing? https://etherpad.openstack.org/p/BrainstormingOSFProjectConfirmationGuidelines15:16
lbragstadif we keep managing goal state for each project team in something like storyboard - then rendering burndown charts might be a lot easier than in the past (pulling a list of project repos from governance, applying in excludes list, asking gerrit for reviews matching a certain topic, etc..)15:16
dhellmannI guess the question for me is what value doing all of that provides, and whether it is worth investing time and energy into it. What would we say "no" to doing if we say "yes" to doing that?15:18
dhellmannI don't oppose someone doing that. I think the data is all there (or could be added with minimal effort from project teams). But I don't need it, so my motivation to work on better reporting is low.15:20
dhellmanngmann : what's the state of your patch to add feature discovery to the vision? ttx asked for a mailing list post, are you going to wait until after the new year to do that? https://review.openstack.org/#/c/621516/15:22
dhellmannmaybe I missed it15:22
fungidhellmann: i've been playing along from home on the oip guidelines, yes15:23
dhellmannI haven't participated in the working group meetings. Have you joined any of those?15:24
dhellmannI'm curious about how things are progressing15:24
dhellmannmaybe not much at all right now, since it's election season15:24
fungidhellmann: yeah, i haven't been on their pop-up team calls (i don't think it's been delegated as an official wg of the board), though jbryce may be able to give an update if he's around15:25
dhellmannevery time we have a new thing we make up a new way to classify it...15:26
fungithe osf put together a fairly lightweight straw man, and it seems like some of what's going on there is board members are naturally asking clarifying questions, but also maybe looking to put their own stamp on it in some ways15:26
dhellmannI left a question in the etherpad but don't see a response and I'm not sure where else to be looking. I can't tell if the group is using private email threads, meetings I haven't been able to join, or something else to communicate. Or maybe that they're just very quiet right now because of the timing.15:28
jbryceIt's basically holding until next year15:29
dhellmannmakes sense15:29
dhellmanntc-members: after a more careful scan of our projects list, I found that we added 3 teams this year (Adjutant, Qinling, and PowerVMStackers), so I updated the text I will be submitting for the OSF annual report to reflect that https://etherpad.openstack.org/p/openstack-2018-annual-report15:30
dhellmannline 9615:30
smcginnisLooks good.15:31
ttxhmm15:31
dhellmannQinling was proposed in January, PowerVMStackers in February, and Adjutant in March15:32
ttxyes15:32
fungilgtm15:32
jbrycedhellmann: there will be public calls scheduled next year once the group reconvenes15:33
dhellmannjbryce : ok, that's good to know; I'll keep an eye out for the announcements15:34
dhellmanntc-members: during the discussions about managing python 3 updates, the folks from SUSE expressed interest in having their OS listed in the PTI under https://governance.openstack.org/tc/reference/project-testing-interface.html#linux-distributions15:37
*** evrardjp has quit IRC15:37
dhellmanndo we have someone who wants to help make that happen?15:37
smcginnisI thought one of the SUSE folks were going to pursue that.15:37
smcginnisJP maybe? Or Dirk?15:37
dhellmannmaybe?  I didn't write down a name and didn't remember15:37
smcginnisI would think it would be best being driven by someone from SUSE.15:38
dhellmannoh, maybe JP did say he was going to do that15:38
dirkI can help if you want15:38
dirkI think JP volunteered though15:38
dhellmannyeah, I wasn't sure if we wanted to write up anything about how/why we modify that list15:38
dhellmannbefore actually adding the new entry15:38
fungii have a feeling that's dependent on stabilizing upstream testing on opensuse15:38
dhellmannthanks, dirk15:38
fungii know there have been a bunch of failing non-voting opensuse jobs moved to experimental and similar15:39
dhellmannyeah, do we want to make that a prerequisite? if so, do we want to write that down?15:39
dirkfungi: really?15:40
dhellmannright now the information there has the project teams as an audience15:40
fungi"Each project should run some functional tests on these platforms so we make sure OpenStack works with distros used in production."15:40
fungii take that to imply we need those jobs succeeding and gating too15:40
dirkright, thats the part that I want to have clarified15:41
dirkat some point it was said its devstack and we have that15:41
dirkif its more then thats fine as well, we just need to define what the scope should be15:41
dirkits a bit of a tradeoff on burning extra ressources with little coverage15:41
dirklittle additional coverage15:41
*** evrardjp has joined #openstack-tc15:41
fungi"have" devstack in what sense? like i said, i recall teams moving the opensuse jobs to experimental recently because they were failing and non-voting and didn't seem to have anyone actively engaged in fixing them15:42
fungijust curious where opensuse jobs are currently gating changes for components of openstack, or where they're voting in the check pipeline at least, or if they're non-voting for now then where they're succeeding at all15:45
smcginnisI would think the requirements would be "have devstack" and "passed tempest tests".15:46
fungidigging around, it seems to be non-voting but succeeding in check on changes to the devstack repo at least15:47
fungisee "devstack-platform-opensuse-150" results on https://review.openstack.org/625269 for example, which merged on monday15:48
fungi(note that centos-7 is also non-voting but failing on that change)15:49
fungihonestly, the only good argument i've been able to make that we're testing whether openstack works on centos is that tripleo does ~100% of their testing on centos nodes15:49
fungidevstack has had a very tumultuous past with centos and fedora. it15:50
fungi's broken more often than it works there, from what i've seen over the years15:50
fungiand there's been almost nobody interested in keeping it working15:51
fungi(if i were a cynic i would say it's probably because red hat doesn't sell any devstack-based products)15:51
fungiso perhaps we make tempest the actual requirement? stand up openstack *somehow* on the target platform and then show that you can successfully exercise it with tempest tests15:52
fungithe refstack capabilities set seems like it would send a consistent message about our faith in our interoperability efforts, at least15:53
dhellmannyeah, I think that's reasonable15:54
dhellmann"on the target platform, in CI"15:57
fungiright, gating in ci. on which projects is probably a bigger question16:21
TheJuliaAnd projects understanding where the happy path is in terms of if something breaks with one, does it forcibly stall all activities, or are they more at liberty to just change things over to non-voting as needing, as long as there is the velocity to recover it/get it working eventually.16:24
*** mriedem_afk is now known as mriedem16:32
scasi have personally dealt with this, from the peanut gallery and a deployment focus16:33
scasthe solution in that case being consistent, but met with skepticism due to a long-term desire to be conservative with changes16:34
fungiTheJulia: in most cases so far the answer has been that if there's nobody engaged in solving platform-specific testing issues, disabling the jobs for those platforms is entirely on the table16:40
fungiwe've done similarly when testing broke for pypy interpreter jobs, or jobs on aging stable branches became un-runnable, or special jobs for particular optional backends...16:41
fungiwhat we test is, to some extent, a tautology16:42
fungiwe test the things we can successfully test most of the time, and we don't test the things nobody is keeping testable16:42
cdentthere's a sense to which much of the openstack development scene and process is in a tautological love affair with the gate16:43
fungiso the role of the tc here, as i see it, is to reflect what the community is testing and not hold them to task over an inability to test things that there is no support in the community to implement16:44
cdentI think we can go one step further than that: We can help to highlight the gaps between demands (from wherever) on the community, and the resources that the community has.16:45
fungioh, sure, i agree16:45
cdentright now those resources (across the board) are rather terrifyingly low16:45
cdentWe need to make it safe for people like PTLs to be able to say "that didn't happen because you didn't stump people"16:46
cdentthat ought to be safe now, but i'm not sure we (the tc) provide enough informational air cover16:46
cdentbiab16:47
TheJuliaagree, which is in part why we've had to temporarily drop tests or mark things non-voting while we work to resolve the issues out of our direct control or influence as a project team or community.16:47
fungii guess my point was that the distros list in the pti is not for beating teams over the head with and telling them they're bad for not supporting, but rather should serve as a living reflection of what is being actively tested16:51
fungiso if teams are continuously testing that openstack works on those platforms (for yet-to-be-defined "works"), then we list the platforms there. and if they aren't we remove those platforms from the list, not remove the teams who aren't getting that support from the distros in question16:52
scasi have a perspective on this, but i'm apt to go long-form on irc, which is, admittedly, not the best utilization of the medium16:54
smcginnisFeel free to start a ML thread. I think it's relevant for a wider audience.16:55
scasi tend to long-form email, too, which can be as misaligned :)16:57
fungiyes, this is ripe for discussion on the openstack-discuss ml16:57
*** e0ne_ has quit IRC16:59
TheJuliaIndeed16:59
TheJuliafungi: I guess ultimately your perception (which I agree with) forces it to kind of become a two step process no matter what is done.17:00
fungia lot of the earlier push to standardize on devstack-gate is that teams who avoided it and tried to roll their own integration testsuite usually got things like change dependencies and cross-project testing wrong. we extracted all that logic and embedded it into zuul as of v3, so it's now a lot easier for projects to do integration testing with the tools of their choice17:02
scasin my testing, i tend to do a three-step process before anyone else sees it, but it doesn't catch all of the bugs17:02
scasthere's only so much that fits within the defined bounds, and software is not any more resource-friendly17:04
scasi liken a lot of how i test to deploying devstack within context17:05
scasi think i need to spend some time in a text editor putting more detail around that, since it's relevant17:07
*** ricolin has quit IRC17:09
*** jpich has quit IRC17:15
*** cdent has quit IRC17:20
*** jamesmcarthur has joined #openstack-tc17:21
*** jamesmcarthur has quit IRC17:27
notmynamettx: for the "soft dependencies", which direction does that flow? eg swift and storlets. swift has zero need of storlets, hard or soft, but using storlets can enhance what you do with data stored in swift. in the same way, swift would have a "see-also" for qingling, cinder, glance, zun, sahara, etc. this is different that eg swift + barbican where swift doesn't require it but has functionality w17:28
notmynamehere encryption secrets can be stored in barbican17:28
notmynameI'm fine with the former (makes things look better and exposes how thing could work together), but the latter is more technically precise when it comes to deployment dependencies17:29
notmynamewhich is the intent of the map?17:29
ttxnotmyname: i think see-also is more informative than technical. "You may also be interested in..."17:30
notmynameok, great. thanks17:31
ttxI think linking to storlets in that case makes a lot of sense17:31
*** gouthamr has quit IRC17:31
notmynamettx: and these are service-level things to mention, right? eg I would not put liberasurecode, despite the fact it's managed under openstack governance as a separate project17:32
ttxright17:33
* ttx runs17:34
*** ianychoi has quit IRC18:22
*** jaypipes_ is now known as jaypipes18:28
*** gouthamr_ has joined #openstack-tc18:37
*** diablo_rojo has joined #openstack-tc18:41
*** diablo_rojo has quit IRC18:46
*** e0ne has joined #openstack-tc18:49
dhellmannfungi : I wonder if we want to move the linux platform list down into the series-specific pages to make it easier for that to reflect changes over time as things come and go.18:52
fungithat's for sure an idea18:52
dhellmannwe have version info there, anyway, irrc18:53
dhellmanntc-members: Alan has posted the dates for the board meetings for 2019, if you want to update your calendar: https://wiki.openstack.org/w/index.php?title=Governance/Foundation&diff=next&oldid=16679619:00
fungithanks for the heads-up, dhellmann!!!19:00
lbragstadnice - thanks dhellmann19:03
openstackgerritDoug Hellmann proposed openstack/governance master: add a note to chair duties about tracking board work  https://review.openstack.org/62664419:07
*** diablo_rojo has joined #openstack-tc19:22
*** e0ne_ has joined #openstack-tc19:25
*** e0ne has quit IRC19:25
*** jamesmcarthur has joined #openstack-tc19:31
*** gouthamr_ is now known as gouthamr20:13
*** jamesmcarthur has quit IRC20:24
*** jamesmcarthur has joined #openstack-tc20:24
*** diablo_rojo has quit IRC20:25
*** jamesmcarthur has quit IRC21:04
*** jamesmcarthur has joined #openstack-tc21:04
*** e0ne_ has quit IRC21:21
*** diablo_rojo has joined #openstack-tc21:38
*** diablo_rojo has quit IRC21:44
*** diablo_rojo has joined #openstack-tc21:47
*** dangtrinhnt_ has joined #openstack-tc21:48
*** dangtrinhnt has quit IRC21:49
*** jamesmcarthur has quit IRC22:00
*** jamesmcarthur has joined #openstack-tc22:01
*** jamesmcarthur has quit IRC22:04
*** diablo_rojo has quit IRC22:54

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