Wednesday, 2017-11-29

*** lbragstad has joined #openstack-tc00:06
*** mriedem has quit IRC00:06
*** hongbin has quit IRC00:13
*** kumarmn has joined #openstack-tc00:42
*** mguiney has quit IRC00:43
*** sdague has quit IRC00:44
*** mguiney has joined #openstack-tc00:45
fungitc-members: commence to officing01:01
harlowjatps reports due on entry01:12
SamYapleyeaaaaa if we could not harlowja, that would be greeeeaaattt01:18
fungiharlowja: cover page optional?01:18
harlowjanope, didn't u get the memo01:19
* fungi has half a mind to give harlowja a taste of his red swingline01:19
* fungi thinks better of it01:20
SamYaplesuch a great movie01:20
SamYaplealways forget jennifer anniston is in that01:20
*** liujiong has joined #openstack-tc01:21
*** kumarmn has quit IRC01:27
*** purplerbot has quit IRC01:56
*** purplerbot has joined #openstack-tc01:56
*** mguiney has quit IRC02:14
*** mguiney has joined #openstack-tc02:14
*** kumarmn has joined #openstack-tc02:17
*** kumarmn has quit IRC02:23
*** liujiong has quit IRC03:26
*** kumarmn has joined #openstack-tc03:51
*** kumarmn has quit IRC04:09
*** kumarmn has joined #openstack-tc04:11
*** kumarmn has quit IRC04:16
*** kumarmn has joined #openstack-tc04:37
*** kumarmn has quit IRC04:41
*** rosmaita has quit IRC04:56
*** dklyle has quit IRC05:15
*** alex_xu has quit IRC06:10
*** alex_xu has joined #openstack-tc06:11
*** mguiney has quit IRC07:45
*** mguiney has joined #openstack-tc07:47
*** jpich has joined #openstack-tc08:17
*** chkumar|404 is now known as chandankumar08:22
*** kumarmn has joined #openstack-tc09:10
*** kumarmn has quit IRC09:14
*** dtantsur|afk is now known as dtantsur09:26
*** gcb has joined #openstack-tc09:39
*** gcb has quit IRC11:05
*** sdague has joined #openstack-tc11:27
openstackgerritIfat Afek proposed openstack/governance master: Update the status of Vitrage goals completion
*** dtantsur is now known as dtantsur|brb12:47
*** rosmaita has joined #openstack-tc12:52
ttxWouldn't mind a quick review of before merging it.12:53
ttx(context: we need a job run to start publishing the ICS calendars in the right places)12:54
*** kumarmn has joined #openstack-tc13:11
*** kumarmn has quit IRC13:15
*** kumarmn has joined #openstack-tc13:20
*** kumarmn has quit IRC13:25
openstackgerritDoug Hellmann proposed openstack/governance master: stop linking to documentation from governance
dims+1 ttx13:39
ttxthanks all13:40
*** openstackstatus has quit IRC13:43
*** openstack has joined #openstack-tc13:44
*** ChanServ sets mode: +o openstack13:44
*** openstackstatus has joined #openstack-tc13:45
*** ChanServ sets mode: +v openstackstatus13:45
*** mriedem has joined #openstack-tc13:51
*** cdent has joined #openstack-tc14:10
*** kumarmn has joined #openstack-tc14:20
*** dtantsur|brb is now known as dtantsur14:24
openstackgerritThierry Carrez proposed openstack/governance master: Tags are applied to deliverables, or teams
*** hongbin has joined #openstack-tc14:40
*** openstack has quit IRC14:46
*** openstack has joined #openstack-tc14:50
*** ChanServ sets mode: +o openstack14:50
ttxefried: +2 on the other one. Thanks for uncovering those14:50
ttxIt's the downside of documenting stuff... you need to keep it up to date14:51
efrieddhellmann Given ttx's desire to keep the project-team-guide non-project-specific, is there any other bug-related wiki content you feel it would be appropriate to move over?14:52
efriedI'll grant you I haven't read the *whole* wiki...14:52
ttxefried: it's a good time travel experience!14:53
efriedLike I was going to hit -- but that has a ton of project-specific links.14:53
efriedmaybe I should quit while I'm ahead...14:54
*** cdent has quit IRC14:59
*** marst has joined #openstack-tc15:06
openstackgerritLance Bragstad proposed openstack/governance master: Update policy goal for manila
*** cdent has joined #openstack-tc15:32
cdentzaneb: just reading your email about tempests tests16:26
zanebcdent: there's a lively discussion going on in #heat as well :)16:26
cdentI sympathize/empathize/*ize with your confusion and concern16:26
cdentjust in case you want more confusion there’s also this:
cdentwhich is about where interop tests are supposed to go16:27
cdentthe fact that we’ve got this same name for thing that does interop testing, cloud verification, and integration tests is chaos16:27
zanebah yes. we really should have figured that out *before* making a project-wide goal of where to put tests :/16:27
smcginniscdent: ++16:28
cdentthe assertion in the project wide goal was that anyone who was using a plugin was supposed to put that plugin in a different repo16:28
cdentit didn’t get much further than that16:29
cdentI tried to assert that if some peopel have to use plugins then everyone should16:29
cdentbut that didn’t float16:29
cdentbut now it sounds like, zaneb, that you’re getting a mixed message and only some tests need to move?16:29
mugsiezaneb: do you have tests in tempest?16:30
zanebcdent: I'm getting that message, and that was apparently what was agreed at the PTG (I wasn't there)16:30
zanebmugsie: we used to, but I think they were all deleted. we now have them in a tempest plugin in the heat repo16:30
* cdent blinks16:30
mugsiecdent: look at interop meeting logs from yesterday16:31
mugsiesorry, refstack16:31
zanebcdent: more importantly, I'm sending the message that some (most in our case) tests need to *not* move, even if that means they're not in a tempest plugin any more16:31
cdentzaneb: that seems like a reasonable outcome16:33
cdentactually strike that, I misread and don’t actually know what I think16:33
* cdent thinks16:33
zanebcdent: btw a propos of mugsie's level-playing-field-for-plugins crusade, I present
openstackLaunchpad bug 1735016 in python-openstackclient "Plugin documentation is worthless" [Undecided,New]16:33
mugsieI am confused by this line:16:34
mugsieSo from discussion on IRC today I learned that the plan agreed at the PTG was *not* to move all of heat_integrationtests to a separate repo, but just those tests that test the API and are likely to be used in the trademark program for verifying clouds.16:34
mugsiethat is the exact opposite of what I remember16:34
mugsiemy recolection is all API tests should be in a separate repo16:34
zanebmugsie: I wasn't there, so I'm only going on what I'm being told16:34
mugsiezaneb: Yeah, I am just confuised aboiut what has changed16:35
smcginnisI didn't think we were moving anything from openstack/tempest, just that openstack/$PROJECT tempest tests needed to move because of entry points or such.16:35
zanebsmcginnis: right, this thread is only about Heat's in-tree tempest plugin16:36
mugsiesmcginnis: tempest already did a purge of things not in the "core"16:37
smcginnismugsie: Oh? I wasn't aware of that.16:37
zanebmugsie: so the issue for us is that all of our tests use the API, but only a tiny subset are designed to be 'API tests'. we're fine with that subset going in a separate, branchless plugin repo16:38
cdentmugsie: I’ve struggled to extract meaning from the refstack log16:38
cdentzaneb: is heat using gabbi without tempest these days or with or both?16:39
zanebcdent: with16:39
mugsiezaneb: OK, that makes more sense16:39
zanebcdent: and that's the part that makes perfect sense to go into a separate branchless tempest plugin imho16:39
cdentbecause they are effectively api verfication "scripts"?16:40
zanebthey check that the API doesn't change. it makes sense to run those across multiple branches16:41
* cdent nods16:41
zanebI mean, we can and probably should have more and better tests of that stuff. but those are the ones we have at present16:41
zanebthe other tests are designed to test internals for regressions. they just happen to do so by making API calls16:42
mugsiewell, for us, we moved a few out into the repo, and then started increasing coverage16:42
* cdent maintains a little personal fantasy of the trademark tests being written in gabbi16:42
dtroyercdent: ++  is it time to reverse the sense here and have the trademark tests be the primary purpose of whatever code implements it, and _then_ use that in CI alongside tempest?16:43
cdentthat’s kinda what my thinking was with regard to putting them in their own plugin/repo16:45
mugsiedtroyer: not a bad idea.16:45
cdentonce self-contained they can be and mean whatever they want16:45
cdentwhen they are multiply-hosted and multply-meaning that’s harder to do16:45
dtroyerbut totally divorce it from Tempest.  The co=opting of that is playing out in a way that cries for "stop the madness!"16:45
cdentbut the old canard of “who is going to do it” (in the log that mugsie pointed to earlier) is a concern16:46
dtroyerof course, it's easier said than done without people to do the lifting16:46
mugsieI think #3 in the proposal is a good step to divorcing it, without a huge amount of technical work16:47
mugsiea bit of political work though16:47
zaneb+1 imho tempest should just be a test-running framework. there should be an out-of-tree 'openstack-trademark-tests' plugin that implements the trademark tests16:47
dtroyerI'll shut up now, I'm having a hard enough time keeping up with things I want to do, much less everything that self-appointed critics want me to do too16:47
cdentaw, I prefer it when dtroyer is not shut up16:47
mugsiezaneb: could you comment on please ? just so we get a more divirse set of view points16:48
openstackgerritMerged openstack/project-team-guide master: Port wiki bugs pages (1 of ?)
mugsieand it looks less like mugsie is a drunk, walking down the road shouting at lamp posts16:50
cdentI think I saw you on the street recently, but not sure, I was probably drunk16:50
*** efried is now known as efried_rollin17:22
*** jpich has quit IRC17:29
zanebmugsie, cdent: added a comment on the review17:29
mugsiezaneb: thanks17:31
*** dtantsur is now known as dtantsur|afk18:10
dhellmannefried_rollin : the triage instructions seem like they'd be project-specific, so I agree it's probably best to quit while you're ahead. if we find more content, we can move it later18:22
*** david-lyle has joined #openstack-tc18:45
*** david-lyle has quit IRC18:54
*** diablo_rojo has quit IRC18:55
*** diablo_rojo has joined #openstack-tc18:55
*** david-lyle has joined #openstack-tc19:01
*** david-lyle has quit IRC19:02
*** david-lyle has joined #openstack-tc19:21
*** david-lyle has quit IRC19:34
*** david-lyle has joined #openstack-tc19:35
*** rosmaita has quit IRC19:37
*** rosmaita has joined #openstack-tc19:43
*** david-lyle has quit IRC19:53
*** david-lyle has joined #openstack-tc20:04
*** efried_rollin is now known as efried20:07
efrieddhellmann Ack.20:07
efrieddhellmann That change set merged; at what point do the changes show up on ?  (So I can update the wiki with the redirect)20:09
efriedttx ^20:12
openstackgerritLance Bragstad proposed openstack/governance master: Update policy goal for aodh
*** david-lyle has quit IRC20:45
*** david-lyle has joined #openstack-tc21:04
*** bbrad has joined #openstack-tc21:14
dhellmannefried : they should have showed up shortly after the merge21:14
efrieddhellmann Here?
dhellmannunless publishing is broken21:15
dhellmannefried : the logs for the publish job are in
dhellmannI don't see a failure, but it doesn't seem to have updated that page21:16
dhellmannthat's something to bring up with the infra team, to figure out what's going on there21:16
dhellmannyour new file does show up in the collected files list:
efriedIs "unreachable" something to worry about?21:18
dhellmannthe infra team is going to be more help with that :-/21:22
efriedWill ask there.  Thanks dhellmann21:23
dhellmannI'm familiar with some of the obvious errors, but I don't see one here21:23
dhellmannsorry I can't be more immediate help21:23
dhellmannlbragstad : if you're OK with this keystoneauth release I can approve it -
dhellmannoops, wrong channel21:27
* dhellmann needs cursor-follows-eyes mode21:27
openstackgerritSean McGinnis proposed openstack/governance master: Mark Cinder policy in code as done
*** bbrad has quit IRC21:58
dtroyer+1 on cursor-follows-eyes, although I worry about what else "they" will do with that ability22:00
*** cdent has quit IRC22:03
EmilienMlbragstad: thanks for all the work on the goals22:53
EmilienMlbragstad: I'm wondering if you tried to get projects updating themselves the governance22:53
EmilienMlbragstad: or if you agreed to do it yourself22:53
lbragstadEmilienM: i've certainly encouraged them to propose the updates but i've defaulted to stepping in when needed22:58
EmilienMlbragstad: for the next champions, we might investigate a workflow where the champion doesn't have to do it and the projects know how to do it themselves in a simple manner22:59
EmilienMit can be an etherpad or a gerrit topic22:59
lbragstadyeah - that'd be easier to tc to instead of having to review all of them22:59
EmilienMand then, once per milestone, we update governance22:59
EmilienMit's not for TC imho23:00
EmilienMit takes me 1 min to review23:00
EmilienMit's more for champions23:00
lbragstadi thought the storyboard integration sounded nice23:00
lbragstadsince we could track everything in a single place23:00
EmilienMimho, a champion should be here for socializing the goal and talk with projects23:00
lbragstadwhen the last commit merges, things are updated in storyboard, and the completion docs are done automatically23:01
EmilienMand projects understand the goal, implement it and update governance or storyboard themselves23:01
lbragstadyeah - that'd be nice23:01
EmilienMdiablo_rojo: hello :-)23:05
EmilienMdiablo_rojo: I was talking with tripleo folks and we were wondering if any project already migrated from Launchpad to StoryBoard, and if yes, who are they, we want to ask them feedback and try ourselves23:05
lbragstadEmilienM: persia was saying the bulk migration is happening earlying 201823:06
EmilienMlbragstad: do we know how that works? which feature are we loosing etc23:06
EmilienMbecause we have tools relying on launchpad now23:06
EmilienMI wasn't aware about a bulk migration23:07
EmilienMwhich projects? all?23:07
lbragstadthat's how i interpreted it when i spoke with persia at the PTG, but i'm sure he'll be able to correct me23:07
diablo_rojoEmilienM, Refstack, Octavia, Craton, Monasca.23:21
EmilienMdiablo_rojo: ok, thanks23:22
diablo_rojoEmilienM, getting ready to migrate Kolla too. Just waiting for them to ask final questions and then pick a date.23:22
diablo_rojoEmilienM, no problem :)23:22
EmilienMdiablo_rojo: is the process documented?23:22
EmilienMdiablo_rojo: like, do we close bugs on launchpad during migration?23:22
EmilienMdiablo_rojo: what do we need to change on our side? (is Closes-Bug: still work?)23:22
EmilienMok i need to read that23:22
diablo_rojoWe have a script that copies over all the bugs23:23
diablo_rojoNo problem :)23:23
EmilienMdiablo_rojo: what happens to the bugs migrated? are they closes in Launchpad?23:24
diablo_rojoI am trying to figure out with Launchpad's painful apis to modify our migration script to post links to the new storyboard stories on the old launchpad bugs.23:25
diablo_rojoEmilienM, ^^23:25
EmilienMthat would be awesome23:26
EmilienMafter a migration, what happens to the bugs in Launchpad?23:26
EmilienMthey aren't touched at all?23:26
EmilienMdiablo_rojo: ^23:27
diablo_rojoThe history stays there. The project needs to disable the ability to report new bugs and should also update their docs to point to storyboard for new bugs. Should let users know too.23:28
EmilienMdiablo_rojo: ok23:40
EmilienMdiablo_rojo: and is the status of the bugs updated after a migration?23:40
EmilienMdiablo_rojo: for example, if a bug is "In progress", when migrated, the bug is also "In progress"?23:40
*** kumarmn has quit IRC23:41
diablo_rojoI think originally they weren't but I vaguely remember someone fixing that.. fungi do you remember?23:41
fungino, the import script doesn't authenticate to lp and makes no changes to any bugs23:42
diablo_rojoBut do the statuses persist after migration?23:42
fungiwe've discussed modifying it to leave comments in all imported bugs with links to the equivalent stories in sb, but the lp api is very unreliable so i don't think that would scale terribly well23:43
fungithe import script leaves in progress bugs in lp as in progress, yes23:43
diablo_rojofungi, even figuring out how to use the apis to authenticate to comment on the lp bugs hasn't been fun.23:44
diablo_rojowriting the comment isn't hard, its the step before that isn't great..23:44
fungiwell, yeah, you have to script openid auth23:44
fungiit's a terrible method of authenticating tools23:45
diablo_rojoEmilienM, tldr is yes, the statuses persist after migration in sb23:45
fungifine for authenticating web browsers, but not really anything else23:45
EmilienMdiablo_rojo: ok cool.23:45
EmilienMdiablo_rojo: we have a TON of bugs23:45
EmilienMdiablo_rojo: I don't want to spend a weekend manually updating them ;-)23:46
EmilienM(well we have scripts etc)23:46
diablo_rojoEmilienM, WHAT!? You mean you would rather be out flying a plane than glued to a screen spending quality time with launchpad?23:46
* diablo_rojo is suprised23:46
EmilienMdiablo_rojo: I would actually do the triage from the plane23:47
EmilienMand before mention it's because diablo_rojo told me to do so23:47
diablo_rojoIt wouldn't be a problem if there werent so many bugs EmilienM ;)23:47
fungidiablo_rojo: anyway, if you write something to leave comments or update statuses on lp bugs en masse, expect it to be an order of magnitude slower (at least) than querying it in a read-only fashion, and also expect many, many, many of those api calls to just time out with an incomprehensible error from the api23:48
diablo_rojofungi, that sounds like something I don't really want to touch with a ten foot pole..23:48
*** marst has quit IRC23:49
fungibecause lp seems to block synchronously while attempting to send e-mail notifications to all subscribers of a bug, which for lots of those bugs on more active projects can easily be in the hundreds and hundreds of recipients23:49
diablo_rojoThese all sound likek very compelling reasons to get the hell away from lauunchpad..23:50
fungiinstead we mostly rely on closing out the bug tracker in lp to hide the bug reporting interface and no longer provide an index of bugs for that project. couple that with a url to the new bug tracker in the project description in lp and it's seemed to work fine for the most part23:50
fungipeople can still update bugs for those projects in lp _if_ they can find them (e.g. through web search engines), but that seems to mostly not happen23:51
fungiand they at least can't open new bugs23:51
diablo_rojoThe index disappearing is a really easy way to keep people from finding things.23:52

Generated by 2.15.3 by Marius Gedminas - find it at!