| dougwig | mtreinish: oh, good point. i'll look at it when i get home. | 00:02 |
|---|---|---|
| *** luqas has quit IRC | 00:04 | |
| *** luqas has joined #openstack-infra | 00:04 | |
| *** baoli has quit IRC | 00:05 | |
| *** gyee has quit IRC | 00:05 | |
| *** sabari is now known as zz_sabari | 00:06 | |
| *** salv-orlando has quit IRC | 00:08 | |
| *** luqas__ has joined #openstack-infra | 00:08 | |
| *** ZZelle has quit IRC | 00:09 | |
| *** kumartin has quit IRC | 00:09 | |
| *** luqas has quit IRC | 00:09 | |
| *** hdd has joined #openstack-infra | 00:09 | |
| *** esker has joined #openstack-infra | 00:11 | |
| *** esker has quit IRC | 00:11 | |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/system-config: Added manifest for storyboard-dev.openstack.org https://review.openstack.org/140466 | 00:12 |
| *** emagana has quit IRC | 00:14 | |
| *** emagana has joined #openstack-infra | 00:15 | |
| *** zz_sabari is now known as sabari | 00:15 | |
| HenryG | Are there any downsides to this? https://review.openstack.org/140863 | 00:16 |
| *** sdake has quit IRC | 00:16 | |
| clarkb | HenryG: yes its already disabled by default | 00:16 |
| clarkb | HenryG: it will only rebase you if you need to rebase to address conflicts | 00:16 |
| *** Sukhdev has quit IRC | 00:17 | |
| fungi | or if devs are using older versions of git-review which don't grok that .gitreview setting anyway | 00:17 |
| *** emagana_ has joined #openstack-infra | 00:17 | |
| *** Rockyg has quit IRC | 00:17 | |
| *** carl_baldwin has joined #openstack-infra | 00:17 | |
| clarkb | and if truly depend on parent changes they will have to be uplifted in a rebase to maintain that dependency chain | 00:17 |
| fungi | so the only way to fix them is to get them to use a newer git-review | 00:17 |
| anteaya | fungi: time for a quick review to fix a mistake I made? https://review.openstack.org/#/c/140867/ | 00:18 |
| *** emagana has quit IRC | 00:18 | |
| clarkb | also I do not understand why rebasing and addressing other comments need to be separated. They are essentially bothe -1 comments. One is computer generated the other human generated but both valid to be addressed | 00:18 |
| *** shakamunyi has quit IRC | 00:18 | |
| HenryG | When did it get disabled by default? | 00:18 |
| fungi | HenryG: when that setting got added | 00:18 |
| fungi | checking git log for the version now | 00:19 |
| *** SumitNaiksatam has joined #openstack-infra | 00:19 | |
| clarkb | oh also I think a lot of this stuff is often related to people using git review -x not rebases. I have noticed that more and more people herry pick and that always results in new commits | 00:19 |
| jeblair | clarkb: i'm guessing it means you can review patchset N-2..N-1 and then assume that the diff in N-1..N is solely rebase content. of course, you can't actually assume that. and when you need to diff N-2..N+1 you're screwed anyway. | 00:20 |
| fungi | true, and no setting in git-review is going to solve that | 00:20 |
| clarkb | jeblair: right | 00:20 |
| *** ChuckC has joined #openstack-infra | 00:20 | |
| *** dangers_away is now known as dangers | 00:21 | |
| carl_baldwin | jeblair: I don’t ever assume that a patch set is a pure rebase. But, it is easy to verify. Then, the rest of the changes can be more easily inspected by setting the Reference Version in gerrit. | 00:21 |
| fungi | carl_baldwin: added in 76b9b86 (with a default value of 0) which was included in git-review 1.16 | 00:22 |
| *** annegent_ has joined #openstack-infra | 00:22 | |
| *** emagana_ has quit IRC | 00:22 | |
| clarkb | I think if you want a featuer to make git review say "this will rebase" then abort that is worth doing, though it should do that anyways if there are conflicts | 00:22 |
| clarkb | since git cant resolve those without a human intervening | 00:22 |
| *** emagana has joined #openstack-infra | 00:23 | |
| carl_baldwin | clarkb: It is true that parent changes need to be uplifted in a rebase. But, this needs to be done from bottom up with great care when there are multiple authors. Otherwise, someone’s work in progress could get clobbered. | 00:23 |
| clarkb | carl_baldwin: yes, but humans need to do that | 00:23 |
| clarkb | carl_baldwin: git review can't know when not to do that | 00:23 |
| clarkb | carl_baldwin: and if it never does it then your code will never get past zuul | 00:23 |
| carl_baldwin | clarkb: I agree which is why I think git review should never try it. | 00:24 |
| clarkb | I guess we could just never rebase, have zuul -1 you, then start over | 00:24 |
| carl_baldwin | clarkb: I’m happy with that. | 00:24 |
| clarkb | carl_baldwin: but its only trying when there are conflicts | 00:24 |
| clarkb | carl_baldwin: so you can say at that point "wait let me stop here" | 00:24 |
| *** sdake has joined #openstack-infra | 00:24 | |
| clarkb | and if we have specific examples where that isn't the case we should file bugs and fix them | 00:24 |
| krotscheck | fungi: Bad regex on my part, looks like https://review.openstack.org/#/c/140466/ works now. | 00:25 |
| krotscheck | Or at least it passes. | 00:25 |
| fungi | krotscheck: rereviewing now--thanks | 00:25 |
| clarkb | carl_baldwin: I guess what I am trying to say is git review already does this for the most part | 00:25 |
| clarkb | carl_baldwin: and a global disable will only cause more iterations in gerrit | 00:26 |
| *** andreykurilin_ has quit IRC | 00:26 | |
| carl_baldwin | clarkb: I like the extra iterations to split rebase from other changes. | 00:27 |
| clarkb | carl_baldwin: what git review does not do is git rebase --abort for you | 00:27 |
| *** emagana has quit IRC | 00:27 | |
| clarkb | it will exit and leave you in that state assuming you want to address the conflicts then rerun git review | 00:27 |
| *** avozza is now known as zz_avozza | 00:28 | |
| *** annegent_ has quit IRC | 00:28 | |
| carl_baldwin | clarkb: Since I don’t want git review to rebase for me then I don’t mind. | 00:28 |
| *** dkliban is now known as dkliban_afk | 00:30 | |
| clarkb | carl_baldwin: for the split rebase from other changes thing, you wait to rebase until everyone has +2'd code that jenkins -1s? | 00:30 |
| clarkb | carl_baldwin: trying to understand what the workflow around that is | 00:30 |
| clarkb | (it is not something I do) | 00:30 |
| *** patrickeast has quit IRC | 00:31 | |
| *** annegent_ has joined #openstack-infra | 00:31 | |
| carl_baldwin | clarkb: In my experience, it happens much more in the check queue than in the gate queue but yes, I wait for Jenkins to tell me the patch is in conflict. | 00:32 |
| *** oomichi has joined #openstack-infra | 00:32 | |
| fungi | not to mention, having to address merge conflicts when you rebase has the potential to make the prior votes irrelevant anyway since the code may need more substantial rewriting at that point to remain relevant to the project | 00:32 |
| clarkb | carl_baldwin: and that is preferred to git review telling you it is in conflict? | 00:32 |
| *** jaypipes has quit IRC | 00:32 | |
| clarkb | carl_baldwin: I ask beacuse you essentially push a broken patch that can never merge and that to me seems suboptimal | 00:32 |
| fungi | in general, it's not safe to ignore changes introduced by a rebase, even a gratuitous one, but especially not one with merge conflicts | 00:33 |
| clarkb | keep in mind that git review should not rebase you if there is no conflict so the only case where that option is relevant is where there is a conflict | 00:33 |
| *** teran_ has quit IRC | 00:34 | |
| clarkb | ^^^ is ultimately the source of my confusion. I see git reviews behavior as desired because I know as early aspossible when something needs attention | 00:34 |
| carl_baldwin | clarkb: suboptimal or not, I do push reviews that may fail a merge because I want to split rebase from other changes to help reviewers. | 00:34 |
| clarkb | I then have the choise to either address it or not address and push as is | 00:34 |
| *** teran has joined #openstack-infra | 00:34 | |
| carl_baldwin | If git review only advised me that my patch was in conflict I would like that. | 00:35 |
| clarkb | carl_baldwin: rather than starting a rebase and failing? | 00:35 |
| carl_baldwin | clarkb: Yes. | 00:35 |
| clarkb | are those two things functionally different? | 00:36 |
| *** annegent_ has quit IRC | 00:36 | |
| clarkb | I guess users may feel compeleed to address the conflicts and continue rather than aborting? | 00:36 |
| fungi | so an option like "conflictwarn" which works like the opportunistic rebase but then undoes it and issues a warning instead of pushing | 00:36 |
| clarkb | fungi: sort of except that git review will undo it if there is no conflict. if there is a conflict it will issue a warning in the form of "fix this" and not push anything | 00:37 |
| clarkb | but it does leave you in a state of "you should address these conflicts" which sounds like is a thing carl_baldwin does not like | 00:38 |
| fungi | ahh, yes perhaps | 00:38 |
| carl_baldwin | clarkb: fungi: yes, except that I’d like to be able to push even with the warning. | 00:38 |
| clarkb | carl_baldwin: you can | 00:38 |
| clarkb | carl_baldwin: git rebase --abort && git rebase -R | 00:38 |
| anteaya | clarkb: you have two +2's, self-approve at will I say: https://review.openstack.org/#/c/140867/ | 00:38 |
| clarkb | er && git review -R | 00:38 |
| clarkb | anteaya: thanks | 00:39 |
| *** teran has quit IRC | 00:39 | |
| clarkb | carl_baldwin: I am basically treating the git rebase conflict message as the warning we want and trying to figure out if that warning is not working for some reason | 00:39 |
| clarkb | carl_baldwin: and I think probably because people see a rebase conflict and unsconsciously address it | 00:40 |
| *** yamamoto has joined #openstack-infra | 00:40 | |
| clarkb | in which case an implicit --abort would likely help | 00:40 |
| carl_baldwin | clarkb: I need to find the cases where this was a problem and study them a bit more. | 00:40 |
| *** Masahiro has joined #openstack-infra | 00:40 | |
| clarkb | carl_baldwin: I do think git review -x has become popular and this won't address any of the problems brought up by that | 00:40 |
| fungi | got it, so an option to have it call undo_rebase() followed by a push | 00:40 |
| clarkb | fungi: ya, that is what I am thinking | 00:41 |
| fungi | and warn loudly while doing so | 00:41 |
| *** SumitNaiksatam has quit IRC | 00:41 | |
| carl_baldwin | clarkb: There were cases when we had DVR under development where patch sets got clobbered repeatedly by other authors. | 00:41 |
| carl_baldwin | clarkb: fungi: I’ll do some more research to see if what I thought happened wasn’t quite the case. | 00:42 |
| clarkb | carl_baldwin: yes this should happen only if there are conflicts that need addressing or if they cherry picked or if they have old git review | 00:42 |
| carl_baldwin | clarkb: fungi: Thanks for the discussion. I’m off to dinner now. I’ll look more in to it later. | 00:42 |
| *** salv-orlando has joined #openstack-infra | 00:43 | |
| *** carl_baldwin has quit IRC | 00:43 | |
| openstackgerrit | Merged openstack-infra/project-config: Stop using networking-l2gw group https://review.openstack.org/140867 | 00:43 |
| *** penguinRaider__ has joined #openstack-infra | 00:43 | |
| *** davideagnello has quit IRC | 00:44 | |
| *** Masahiro has quit IRC | 00:45 | |
| *** sabari is now known as zz_sabari | 00:47 | |
| anteaya | clarkb: I'm going out for dinner soon, let me know when gerrit groups are happy | 00:47 |
| anteaya | clarkb: I'll wait | 00:47 |
| clarkb | anteaya: you can check at https://review.openstack.org/#/admin/projects/stackforge/networking-l2gw,access the abandon permission should update | 00:47 |
| clarkb | anteaya: I expcet it to happen within a few minutes | 00:48 |
| * anteaya stares at her browser | 00:48 | |
| *** MaxV has quit IRC | 00:50 | |
| fungi | krotscheck: i think we probably need to not use explicit key material contents for the -dev server, and instead use the snakeoil cert/key pair already on the server. comments left inline | 00:52 |
| *** armax has joined #openstack-infra | 00:52 | |
| krotscheck | kk | 00:52 |
| *** zz_sabari is now known as sabari | 00:52 | |
| fungi | krotscheck: that will match how we do our other dev servers, and is much more convenient for devs trying it out on their own too | 00:52 |
| fungi | (and also it's one more set of things we don't need to stick in hiera) | 00:52 |
| *** bswartz has joined #openstack-infra | 00:53 | |
| *** melwitt has quit IRC | 00:53 | |
| *** Sukhdev has joined #openstack-infra | 00:53 | |
| *** patrickeast has joined #openstack-infra | 00:53 | |
| *** dims has quit IRC | 00:54 | |
| *** gmann has joined #openstack-infra | 00:58 | |
| anteaya | clarkb: yay, done | 00:58 |
| *** dannywilson has joined #openstack-infra | 00:59 | |
| *** otter768 has joined #openstack-infra | 00:59 | |
| anteaya | clarkb: sorry about the mistake and thanks for fixing it | 00:59 |
| *** alexpilotti has quit IRC | 00:59 | |
| clarkb | anteaya: no problem. I missed it in review too | 00:59 |
| *** unicell has joined #openstack-infra | 01:00 | |
| gmann | anteaya: can you please have a look on nova V2.1 API new job. dependent patch is merged - https://review.openstack.org/#/c/138951/ | 01:00 |
| *** salv-orlando has quit IRC | 01:00 | |
| *** alexpilotti has joined #openstack-infra | 01:00 | |
| *** ryanpetrello has joined #openstack-infra | 01:02 | |
| *** dannywilson has quit IRC | 01:03 | |
| *** alexpilotti has quit IRC | 01:03 | |
| *** dannywilson has joined #openstack-infra | 01:03 | |
| *** otter768 has quit IRC | 01:04 | |
| *** pc_m has quit IRC | 01:04 | |
| *** bdpayne has quit IRC | 01:04 | |
| *** armax has quit IRC | 01:04 | |
| *** teran has joined #openstack-infra | 01:05 | |
| *** yamamoto has quit IRC | 01:06 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/project-config: Add experimental Fedora 21 devstack job https://review.openstack.org/140889 | 01:06 |
| anteaya | gmann: +A | 01:06 |
| anteaya | and I'm out see you on Friday | 01:06 |
| *** armax has joined #openstack-infra | 01:06 | |
| gmann | anteaya: Thanks. | 01:06 |
| *** teran has quit IRC | 01:11 | |
| *** dims has joined #openstack-infra | 01:12 | |
| openstackgerrit | Merged openstack-infra/project-config: Adds experimental job for testing Nova V2.1 API https://review.openstack.org/138951 | 01:12 |
| *** luqas___ has joined #openstack-infra | 01:12 | |
| *** luqas__ has quit IRC | 01:12 | |
| *** Longgeek has joined #openstack-infra | 01:13 | |
| *** ryanpetrello has quit IRC | 01:13 | |
| *** Longgeek has quit IRC | 01:14 | |
| *** ayoung_dad_mode has quit IRC | 01:14 | |
| *** Longgeek has joined #openstack-infra | 01:14 | |
| *** dims has quit IRC | 01:16 | |
| *** luqas___ has quit IRC | 01:17 | |
| *** luqas___ has joined #openstack-infra | 01:17 | |
| *** Longgeek has quit IRC | 01:18 | |
| *** hdd has quit IRC | 01:19 | |
| mordred | jeblair, clarkb, mtreinish: yes, sorry - I really do want to review the mysql-proxy thing | 01:19 |
| clarkb | mordred: ok | 01:19 |
| * mordred goes to look | 01:20 | |
| *** zhiwei has joined #openstack-infra | 01:20 | |
| *** davideagnello has joined #openstack-infra | 01:20 | |
| *** reed has quit IRC | 01:21 | |
| *** davideagnello has quit IRC | 01:23 | |
| mordred | clarkb, mtreinish: so, the lua theoretically will only allow select statements and nothing else? | 01:24 |
| mordred | as in, show processlist should be blocked, yeah? | 01:24 |
| *** Masahiro has joined #openstack-infra | 01:24 | |
| clarkb | mordred: that is how I explained it. It splits out read and writes. Then there is a directive saying only allow the reads | 01:25 |
| clarkb | mordred: for that I do not know | 01:25 |
| *** dangers is now known as dangers_away | 01:25 | |
| *** dangers_away is now known as dangers | 01:25 | |
| mordred | ok. we should verify that somehow - because there's a thing running around where you can dos a mysql server with a bunch of show processlist queries | 01:25 |
| *** dangers is now known as dangers_away | 01:25 | |
| clarkb | mordred: well you can dos it with a select too | 01:25 |
| mordred | (all of the SHOW commands in mysql are HORRIBLE HORRIBLE hacks) | 01:25 |
| mordred | sure - but you can dos is real bad with show processlist | 01:26 |
| clarkb | do a full cross join of that 27million row table against itself | 01:26 |
| *** rmcall has joined #openstack-infra | 01:26 | |
| clarkb | but ya I think its reasonable to test these things | 01:27 |
| clarkb | to limit the amount of crazy | 01:27 |
| mordred | clarkb: actually, if your buffers are tuned right, that should just be slow and should not fully kill the entire server | 01:28 |
| mordred | http://bugs.mysql.com/bug.php?id=75155 | 01:29 |
| mordred | there it is | 01:29 |
| *** Sukhdev has quit IRC | 01:29 | |
| clarkb | it will dos it for any other users though iirc | 01:29 |
| clarkb | nwe connections and other queries just won't work | 01:29 |
| mordred | well, read that one - you'll like it | 01:29 |
| clarkb | but its been a long time since I supported a "public" database | 01:30 |
| mordred | yah - that's kinda why I'm being this person | 01:30 |
| *** davideagnello has joined #openstack-infra | 01:30 | |
| clarkb | mordred: at PSU we kept track of how many times we had to kill user's DB connections | 01:30 |
| mordred | every time I've ever talked to folks about runnign mysql with a port open, I've told them they're crazy | 01:30 |
| clarkb | not specific users just a kill tally | 01:30 |
| *** davideagnello has quit IRC | 01:31 | |
| mordred | and I'd honestly really prefer doing this with a slave | 01:31 |
| mordred | but trove | 01:31 |
| clarkb | mordred: even though its completely unsensitive data and an unshared DB host? | 01:31 |
| *** camunoz has quit IRC | 01:31 | |
| clarkb | I guess I don't worry about it too much because I don't mind deleting it and starting over | 01:31 |
| mordred | sure. | 01:31 |
| mordred | but in that case, why bother with mysql-proxy at all? | 01:31 |
| clarkb | mordred: because trove | 01:32 |
| mordred | (this is sort of the crux of my line of questioning) | 01:32 |
| mordred | either we're worried about the database or we're not | 01:32 |
| *** camunoz has joined #openstack-infra | 01:32 | |
| clarkb | no, there is no public IP for the DB | 01:32 |
| clarkb | so we have to proxy it somehow | 01:32 |
| mordred | OH | 01:32 |
| mordred | gotha | 01:32 |
| mordred | I completely misunderstood the problem and solution | 01:32 |
| mordred | (this is kinda like the "are containers for security or ease of install" problem - if you're answering the wrong question, things get dodgy) | 01:33 |
| mordred | sure - if this is for that, then fine | 01:33 |
| clarkb | mordred: yes very similar | 01:33 |
| *** r-daneel has quit IRC | 01:33 | |
| mordred | just like docker, I do not believe this adds any security | 01:34 |
| clarkb | agreed | 01:34 |
| mordred | but it may be useful for other reasons :) | 01:34 |
| fungi | mordred: so basically the options are local replica hosted on a reachable server, tcp socket proxy, or application proxy. mtreinish went with that last one | 01:34 |
| mordred | yah | 01:36 |
| *** annegent_ has joined #openstack-infra | 01:36 | |
| mordred | makes sense | 01:36 |
| mordred | now that I know context :) | 01:36 |
| mordred | this would be easier if we were running our own mysql cluster ... which I occasionally think we should do | 01:37 |
| mordred | but then occasionally I look at my todo list and decide not to do that | 01:37 |
| *** penguinRaider__ has quit IRC | 01:37 | |
| *** armax has quit IRC | 01:38 | |
| *** ChuckC has quit IRC | 01:38 | |
| clarkb | I am going ot go play video games now but can approve that tomorrow if mordred/fungi don't beat me to it | 01:38 |
| *** hdd has joined #openstack-infra | 01:39 | |
| *** penguinRaider has joined #openstack-infra | 01:39 | |
| *** markmcclain has quit IRC | 01:39 | |
| *** yaguang has joined #openstack-infra | 01:41 | |
| *** adalbas has joined #openstack-infra | 01:41 | |
| *** annegent_ has quit IRC | 01:41 | |
| *** yaguang has quit IRC | 01:41 | |
| *** rmcall has quit IRC | 01:42 | |
| *** yamamoto has joined #openstack-infra | 01:43 | |
| *** shakamunyi has joined #openstack-infra | 01:43 | |
| *** ChuckC has joined #openstack-infra | 01:44 | |
| *** ayoung has joined #openstack-infra | 01:45 | |
| *** fandi has quit IRC | 01:46 | |
| *** rmcall has joined #openstack-infra | 01:46 | |
| openstackgerrit | Spencer Krum proposed openstack-infra/puppet-storyboard: Add metadata.json to puppet-storyboard https://review.openstack.org/136591 | 01:47 |
| *** rmcall has quit IRC | 01:47 | |
| *** penguinRaider__ has joined #openstack-infra | 01:48 | |
| *** ChuckC has quit IRC | 01:48 | |
| *** ChuckC has joined #openstack-infra | 01:49 | |
| *** r-daneel has joined #openstack-infra | 01:49 | |
| *** MaxV has joined #openstack-infra | 01:51 | |
| *** rmcall has joined #openstack-infra | 01:51 | |
| *** penguinRaider has quit IRC | 01:51 | |
| *** annegent_ has joined #openstack-infra | 01:52 | |
| openstackgerrit | Ramy Asselin proposed openstack-infra/infra-specs: Update module issues_url & include storyboard https://review.openstack.org/135452 | 01:54 |
| *** sabari is now known as zz_sabari | 01:54 | |
| *** penguinRaider__ has quit IRC | 01:54 | |
| *** amotoki has joined #openstack-infra | 01:55 | |
| *** MaxV has quit IRC | 01:55 | |
| *** penguinRaider has joined #openstack-infra | 01:56 | |
| *** camunoz has quit IRC | 01:57 | |
| *** shashankhegde has quit IRC | 01:57 | |
| *** gmann has left #openstack-infra | 01:58 | |
| *** yaguang has joined #openstack-infra | 02:02 | |
| *** shashankhegde has joined #openstack-infra | 02:03 | |
| *** shayneburgess has quit IRC | 02:04 | |
| *** stevemar has joined #openstack-infra | 02:05 | |
| mordred | nibalizer: don't kill me ... | 02:06 |
| mordred | nibalizer: but I just had an idea that you'll either love or hate | 02:06 |
| mordred | but I may be stupid because there may be things I don't know about how things work | 02:07 |
| *** teran has joined #openstack-infra | 02:07 | |
| mordred | nibalizer: so - is the metadata.json file useful in and of itself in the git repo - or is it useful as an input to the puppet module upload process? | 02:07 |
| *** penguinRaider__ has joined #openstack-infra | 02:08 | |
| nibalizer | hehehe | 02:09 |
| *** camunoz has joined #openstack-infra | 02:10 | |
| nibalizer | it needs to be in the repo for when you build the package and upload it | 02:10 |
| nibalizer | what are you thinking? | 02:10 |
| *** penguinRaider has quit IRC | 02:12 | |
| *** teran has quit IRC | 02:12 | |
| *** luqas__ has joined #openstack-infra | 02:13 | |
| *** dkranz has joined #openstack-infra | 02:13 | |
| *** luqas___ has quit IRC | 02:14 | |
| *** luqas__ has quit IRC | 02:16 | |
| *** luqas has joined #openstack-infra | 02:17 | |
| *** luqas has quit IRC | 02:19 | |
| *** luqas has joined #openstack-infra | 02:19 | |
| mordred | nibalizer: well, no, I mean, other than package building, do tools need it? | 02:23 |
| mordred | nibalizer: because my idea is that, for our other things, we have git tags generate metadata files during build with the appropriate version | 02:24 |
| nibalizer | ooo | 02:24 |
| mordred | nibalizer: so I was thinking that we put this info into our git repos in a nice human yaml file | 02:24 |
| mordred | then we have a wrapper tool thing that can do the git tag + data from yaml file -> output metadata.json and upload | 02:24 |
| *** hdd has quit IRC | 02:25 | |
| mordred | that way our puppet become first class citizens in teh excellent "tag-to-release" workflow that we have for everyone else | 02:25 |
| nibalizer | hrmm | 02:25 |
| nibalizer | so instead of making a git commit that bumps the version number | 02:25 |
| mordred | nibalizer: but - if tools that wat to consume our puppet from git want that file there for some reason, then it breaks | 02:25 |
| mordred | yah | 02:25 |
| nibalizer | the tagging process ninjas in the metadata.json | 02:25 |
| mordred | yup | 02:25 |
| nibalizer | idunno | 02:25 |
| mordred | (this is what pbr does for the python stuff, btw - we don't store versions in files in our python repos) | 02:26 |
| mordred | (same with java and mvn) | 02:26 |
| nibalizer | what about dependencies and stuff? | 02:26 |
| mordred | they all go into the metadata.yaml or whatever we _do_ put into git | 02:26 |
| nibalizer | so we've got this openstackci-httpd module | 02:26 |
| nibalizer | and say we cut a 1.0.0 and then storyboard depends on that | 02:26 |
| nibalizer | what do? | 02:26 |
| mordred | wouldn't that work the same way? | 02:27 |
| nibalizer | i guess | 02:28 |
| nibalizer | and python requirements are loaded from global requirements or? | 02:28 |
| mordred | that's right | 02:28 |
| nibalizer | even for e.g. nodepool? | 02:28 |
| mordred | yah. that's how it works in python land | 02:29 |
| mordred | wait - let's make sure we're talking about the same thing | 02:29 |
| nibalizer | so if i git clone nodepool there isn't a requirements.txt in there? | 02:29 |
| mordred | there is | 02:29 |
| mordred | the only thing that's not in the git repo is the version number of nodepool | 02:30 |
| mordred | because that's calculated from git | 02:30 |
| *** david-lyle is now known as david-lyle_afk | 02:30 | |
| *** dkranz has quit IRC | 02:30 | |
| nibalizer | oh this old problem | 02:30 |
| nibalizer | where git doesn't let you put the git hash in the git repo | 02:30 |
| nibalizer | because paradox | 02:30 |
| mordred | right. | 02:31 |
| mordred | so we've solved it inpython and java | 02:31 |
| mordred | and I _think_ we can solve it here too ... | 02:31 |
| mordred | but I know much less about teh mechanics | 02:31 |
| nibalizer | ya there is no tech reason we cant do that | 02:31 |
| mordred | so - it seems liek perhaps it's not an absurd thing to play with at least | 02:31 |
| *** annegent_ has quit IRC | 02:31 | |
| nibalizer | so long as a metadata.json appears before puppet module build (well, rake blacksmith:build, but thats a different conversation) | 02:32 |
| nibalizer | everything should roll | 02:32 |
| mordred | adn the puppetmodule build is driven from rake? | 02:32 |
| nibalizer | hasn't landed yet but ya | 02:32 |
| mordred | cool | 02:32 |
| nibalizer | we even have rake:upload now | 02:32 |
| mordred | that's even easier then | 02:32 |
| nibalizer | so the pypi node can do it | 02:32 |
| *** baoli has joined #openstack-infra | 02:32 | |
| mordred | we could just make a rake task that does the metadata.json generation | 02:32 |
| nibalizer | ya, but then we'd have to write and maintain ruby | 02:33 |
| mordred | sure. but we could share it with our pupept friends | 02:33 |
| mordred | and maybe one of them would find it useful too | 02:33 |
| mordred | (I mean, I'm with you) | 02:33 |
| nibalizer | ya idunno if this would ... go over well | 02:33 |
| mordred | hehe | 02:33 |
| mordred | it's fine | 02:33 |
| mordred | the python people hate us | 02:33 |
| nibalizer | we used to have the metadata.json file generated from another file | 02:33 |
| nibalizer | and that had problems so now we write the metadata.json file | 02:34 |
| mordred | yah - but the other file was TERRIBLE | 02:34 |
| mordred | I remember that | 02:34 |
| nibalizer | and going out and saying 'hey guys ive got a great idea! lets generate the metadata.json file' | 02:34 |
| nibalizer | i think thats how you get run out of town | 02:34 |
| mordred | fair | 02:34 |
| nibalizer | but that doesn't mean we cant do it | 02:34 |
| mordred | well, we'll solve our problems with it and if they like it they like it and if they don't they don't | 02:34 |
| mordred | and we can drink alcohol in the mean time | 02:34 |
| nibalizer | we just need to cover in our readme that we dont have a metadata.json and we do fancy things instead | 02:35 |
| nibalizer | so this https://github.com/maestrodev/puppet-blacksmith has a module:bump task i think | 02:35 |
| nibalizer | which i think is interpreting and modifying the metadata.json in place | 02:35 |
| nibalizer | or maybe generating wholesale | 02:35 |
| nibalizer | but we could probably either fork or push our rake tasks into blacksmith | 02:36 |
| nibalizer | that way it is easier for our puppet friends to consume because that gem is kinda the place you put 'release eng' code right? | 02:36 |
| mordred | yah | 02:36 |
| mordred | yah. there is a similar thing in pyton land - the model of "munge my file that's in git" works great if you have push access | 02:36 |
| nibalizer | oh | 02:37 |
| nibalizer | ohhhh | 02:37 |
| mordred | but in a captive code review system, all of the "store version number in source code repo" things break | 02:37 |
| nibalizer | okay now im with you | 02:37 |
| mordred | see what I'm saying? | 02:37 |
| nibalizer | i was like 'whats the advantage' | 02:37 |
| mordred | yah | 02:37 |
| nibalizer | its because that way cores can cut releases | 02:37 |
| mordred | yup | 02:37 |
| *** shakamunyi_ has joined #openstack-infra | 02:37 | |
| mordred | just by pushing a tag | 02:37 |
| nibalizer | without the 'this is the version bump' review having to chug and get stale and be dependent and uggggh | 02:37 |
| mordred | yup | 02:38 |
| nibalizer | i've been getting way into the gerrit api these days tho | 02:38 |
| mordred | I'll bet you have | 02:38 |
| nibalizer | i have 'gerrit wip' on the command line, so i think there could be 'gerrit release' ? | 02:38 |
| nibalizer | but the system is already set up | 02:38 |
| nibalizer | so yea lets make puppet do the thing the python things do | 02:38 |
| nibalizer | and if it burns out around us we can call that experiment over and do something more normal | 02:38 |
| nibalizer | (ofc we should get greater consensus than just us) | 02:39 |
| mordred | (you could add wip support back to git-review, btw ... I used to have the functionality to do a git review --wip but then something changed and we never added it back) | 02:39 |
| mordred | nibalizer: ++ | 02:39 |
| mordred | nibalizer: I basically wanted to make sure that if I hacked up a POC you wouldn't come chasing after me with pointy stcks | 02:39 |
| nibalizer | nope | 02:40 |
| mordred | woot | 02:40 |
| mordred | I don't like pointy sticks | 02:40 |
| nibalizer | but yea you should be aware of my patches to build some of the pipeline because thats probably what you sholud base your work on | 02:40 |
| nibalizer | er maybe it merged! | 02:41 |
| *** Masahiro has quit IRC | 02:42 | |
| *** penguinRaider__ has quit IRC | 02:43 | |
| *** penguinRaider__ has joined #openstack-infra | 02:44 | |
| *** yaguang has quit IRC | 02:45 | |
| *** rmcall has quit IRC | 02:45 | |
| *** radez is now known as radez_g0n3 | 02:46 | |
| *** bdpayne has joined #openstack-infra | 02:47 | |
| mordred | nibalizer: oh, I'm not EVEN close to touching this | 02:47 |
| *** shakamunyi_ has quit IRC | 02:48 | |
| *** Masahiro has joined #openstack-infra | 02:49 | |
| *** yaguang has joined #openstack-infra | 02:49 | |
| ianw | clarkb: finally dug into the nodepool f21 issues properly ... it doesn't like numeric release. changes upcoming | 02:50 |
| heyongli | jogo, anteaya new testcases uploaded: https://review.openstack.org/#/c/139000/ | 02:50 |
| *** shashankhegde has quit IRC | 02:50 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/nodepool: Check env-vars is a dictionary https://review.openstack.org/138250 | 02:51 |
| openstackgerrit | Ian Wienand proposed openstack-infra/nodepool: Handle numeric "release" https://review.openstack.org/140901 | 02:51 |
| *** zz_sabari is now known as sabari | 02:52 | |
| *** hdd has joined #openstack-infra | 02:54 | |
| *** shayneburgess has joined #openstack-infra | 02:55 | |
| *** spzala has quit IRC | 02:57 | |
| *** ivar-lazzaro has quit IRC | 02:57 | |
| *** ildikov_afk has quit IRC | 02:59 | |
| *** tnurlygayanov has quit IRC | 02:59 | |
| openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Bump versions of stdlib, postgres, inifile https://review.openstack.org/140902 | 02:59 |
| *** otter768 has joined #openstack-infra | 03:00 | |
| *** shakamunyi_ has joined #openstack-infra | 03:02 | |
| *** otter768 has quit IRC | 03:04 | |
| *** penguinRaider__ has quit IRC | 03:04 | |
| *** ildikov_afk has joined #openstack-infra | 03:04 | |
| *** tnurlygayanov has joined #openstack-infra | 03:04 | |
| *** penguinRaider has joined #openstack-infra | 03:06 | |
| *** teran has joined #openstack-infra | 03:08 | |
| *** Masahiro has quit IRC | 03:08 | |
| *** fandi has joined #openstack-infra | 03:09 | |
| *** shakamunyi has quit IRC | 03:12 | |
| *** shakamunyi_ has quit IRC | 03:12 | |
| *** Ryan_Lane has quit IRC | 03:12 | |
| *** teran has quit IRC | 03:12 | |
| *** shayneburgess has quit IRC | 03:13 | |
| *** e0ne has quit IRC | 03:16 | |
| *** Masahiro has joined #openstack-infra | 03:16 | |
| *** baoli has quit IRC | 03:19 | |
| *** thedodd has joined #openstack-infra | 03:21 | |
| *** baoli has joined #openstack-infra | 03:23 | |
| *** adalbas has quit IRC | 03:25 | |
| *** penguinRaider__ has joined #openstack-infra | 03:34 | |
| *** xyang1 has joined #openstack-infra | 03:35 | |
| *** signed8bit has joined #openstack-infra | 03:35 | |
| *** signed8bit_ZZZzz has quit IRC | 03:36 | |
| *** penguinRaider has quit IRC | 03:37 | |
| *** ryanpetrello has joined #openstack-infra | 03:40 | |
| *** amotoki has quit IRC | 03:42 | |
| *** amotoki has joined #openstack-infra | 03:42 | |
| *** bdpayne_ has joined #openstack-infra | 03:42 | |
| *** pc_m has joined #openstack-infra | 03:43 | |
| *** bdpayne has quit IRC | 03:46 | |
| *** arosen has quit IRC | 03:46 | |
| *** amotoki has quit IRC | 03:47 | |
| *** ryanpetrello has quit IRC | 03:54 | |
| *** kmartin has quit IRC | 03:57 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/system-config: Add Fedora 21 image build to nodepool https://review.openstack.org/140911 | 03:57 |
| *** yamamoto has quit IRC | 03:59 | |
| openstackgerrit | Jerry Zhao proposed openstack-infra/project-config: Create compass-install project https://review.openstack.org/140912 | 04:01 |
| *** koolhead17 has joined #openstack-infra | 04:01 | |
| *** yamamoto has joined #openstack-infra | 04:02 | |
| *** yamamoto has quit IRC | 04:02 | |
| *** yamamoto has joined #openstack-infra | 04:02 | |
| *** yamamot__ has joined #openstack-infra | 04:02 | |
| *** zz_dimtruck is now known as dimtruck | 04:03 | |
| *** hdd has quit IRC | 04:03 | |
| *** arosen has joined #openstack-infra | 04:04 | |
| *** thedodd has quit IRC | 04:06 | |
| *** yamamoto has quit IRC | 04:06 | |
| *** teran has joined #openstack-infra | 04:09 | |
| *** Masahiro has quit IRC | 04:09 | |
| *** Ryan_Lane has joined #openstack-infra | 04:10 | |
| *** Masahiro has joined #openstack-infra | 04:12 | |
| *** teran has quit IRC | 04:14 | |
| *** mwagner_lap has joined #openstack-infra | 04:14 | |
| *** penguinRaider__ has quit IRC | 04:19 | |
| *** jaypipes has joined #openstack-infra | 04:19 | |
| *** penguinRaider__ has joined #openstack-infra | 04:20 | |
| *** yaguang has quit IRC | 04:20 | |
| *** koolhead17 has quit IRC | 04:22 | |
| *** dims has joined #openstack-infra | 04:23 | |
| *** annegent_ has joined #openstack-infra | 04:24 | |
| *** yaguang has joined #openstack-infra | 04:24 | |
| *** sarob has joined #openstack-infra | 04:25 | |
| *** Masahiro has quit IRC | 04:25 | |
| *** annegent_ has quit IRC | 04:25 | |
| *** yaguang has quit IRC | 04:32 | |
| *** yaguang has joined #openstack-infra | 04:33 | |
| *** Masahiro has joined #openstack-infra | 04:33 | |
| *** hdd has joined #openstack-infra | 04:34 | |
| *** dims_ has joined #openstack-infra | 04:37 | |
| *** amotoki has joined #openstack-infra | 04:38 | |
| *** MaxV has joined #openstack-infra | 04:40 | |
| *** dims has quit IRC | 04:40 | |
| *** esker has joined #openstack-infra | 04:40 | |
| *** mmaglana has joined #openstack-infra | 04:41 | |
| *** signed8b_ has joined #openstack-infra | 04:41 | |
| *** signed8bit has quit IRC | 04:42 | |
| *** baoli has quit IRC | 04:42 | |
| *** amotoki_ has joined #openstack-infra | 04:43 | |
| *** koolhead17 has joined #openstack-infra | 04:43 | |
| *** MaxV has quit IRC | 04:44 | |
| *** sabari is now known as zz_sabari | 04:47 | |
| *** koolhead17 has quit IRC | 04:47 | |
| *** koolhead17 has joined #openstack-infra | 04:47 | |
| *** amotoki_ has quit IRC | 04:48 | |
| openstackgerrit | Spencer Krum proposed openstack-infra/system-config: Bump versions of stdlib, puppetdb , inifile https://review.openstack.org/140902 | 04:50 |
| *** koolhead17 has quit IRC | 04:50 | |
| *** koolhead_ has joined #openstack-infra | 04:50 | |
| *** MaxV has joined #openstack-infra | 04:50 | |
| *** bdpayne_ has quit IRC | 04:54 | |
| *** koolhead_ has quit IRC | 04:54 | |
| *** MaxV has quit IRC | 04:55 | |
| *** achanda has joined #openstack-infra | 04:59 | |
| *** otter768 has joined #openstack-infra | 05:00 | |
| *** dimtruck is now known as zz_dimtruck | 05:02 | |
| *** arosen has quit IRC | 05:03 | |
| *** salv-orlando has joined #openstack-infra | 05:04 | |
| *** otter768 has quit IRC | 05:05 | |
| *** sarob has quit IRC | 05:05 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/requirements: pymemcache and sysv_ipc for tooz https://review.openstack.org/140920 | 05:06 |
| *** penguinRaider__ has quit IRC | 05:06 | |
| *** penguinRaider has joined #openstack-infra | 05:07 | |
| *** teran has joined #openstack-infra | 05:10 | |
| *** Sukhdev has joined #openstack-infra | 05:14 | |
| *** teran has quit IRC | 05:14 | |
| *** achanda has quit IRC | 05:16 | |
| *** teran has joined #openstack-infra | 05:17 | |
| *** nikil2 has joined #openstack-infra | 05:17 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/requirements: tooz was missing from projects.txt https://review.openstack.org/140922 | 05:17 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: Add requirements check for tooz https://review.openstack.org/140923 | 05:20 |
| *** jerryz has joined #openstack-infra | 05:20 | |
| *** achanda has joined #openstack-infra | 05:22 | |
| *** arosen has joined #openstack-infra | 05:25 | |
| *** _nadya_ has joined #openstack-infra | 05:25 | |
| *** stevebaker has quit IRC | 05:25 | |
| *** stevebaker has joined #openstack-infra | 05:25 | |
| *** craigbr has quit IRC | 05:27 | |
| yamamot__ | do i need to create "gerrit groups" before adding a reference in gerrit/acls/stackforge/*.config? | 05:27 |
| *** esker has quit IRC | 05:28 | |
| *** luqas__ has joined #openstack-infra | 05:29 | |
| *** luqas has quit IRC | 05:29 | |
| *** teran has quit IRC | 05:29 | |
| *** teran has joined #openstack-infra | 05:30 | |
| *** teran has quit IRC | 05:30 | |
| *** rmcall has joined #openstack-infra | 05:33 | |
| *** dims has joined #openstack-infra | 05:34 | |
| clarkb | no putting it in there creates it | 05:35 |
| *** stevebaker has quit IRC | 05:35 | |
| *** stevebaker has joined #openstack-infra | 05:36 | |
| *** harlowja is now known as harlowja_away | 05:36 | |
| *** dims__ has joined #openstack-infra | 05:37 | |
| *** dims_ has quit IRC | 05:38 | |
| *** penguinRaider__ has joined #openstack-infra | 05:39 | |
| yamamot__ | clarkb: thank you | 05:39 |
| *** dims has quit IRC | 05:40 | |
| arosen | I have a question... How to you set SERVICE_TIMEOUT through devstack-gate? https://github.com/openstack-dev/devstack/blob/master/lib/glance#L327 It's set to 60 seconds but it seems like it some times doesn't wait long enough though i figure 60 sec should already be enough | 05:40 |
| openstackgerrit | YAMAMOTO Takashi proposed openstack-infra/project-config: Add networking-ofagent project to StackForge https://review.openstack.org/140926 | 05:41 |
| arosen | i.e devstack fails because glance api doesn't respond to the wget request to fail earlier if it can't start. | 05:41 |
| clarkb | arosen I am not sure you can but check the the localrc generation code to be sure | 05:41 |
| *** penguinRaider has quit IRC | 05:43 | |
| nibalizer | mordred: i've been talking with crinkle we have some questions | 05:46 |
| *** andreaf has quit IRC | 05:48 | |
| *** andreaf has joined #openstack-infra | 05:48 | |
| openstackgerrit | YAMAMOTO Takashi proposed openstack-infra/project-config: Add networking-ofagent project to StackForge https://review.openstack.org/140926 | 05:53 |
| *** pc_m has quit IRC | 05:59 | |
| *** rushiagr_away is now known as rushiagr | 06:00 | |
| *** talluri has joined #openstack-infra | 06:00 | |
| *** arosen has quit IRC | 06:00 | |
| *** Masahiro has quit IRC | 06:00 | |
| *** pcrews has quit IRC | 06:01 | |
| *** arosen has joined #openstack-infra | 06:02 | |
| *** Masahiro has joined #openstack-infra | 06:03 | |
| *** achanda has quit IRC | 06:04 | |
| *** talluri has quit IRC | 06:04 | |
| *** salv-orlando has quit IRC | 06:04 | |
| *** wenlock has joined #openstack-infra | 06:05 | |
| *** koolhead17 has joined #openstack-infra | 06:08 | |
| *** koolhead17 has joined #openstack-infra | 06:08 | |
| *** hdd has quit IRC | 06:08 | |
| *** jamespage_ has joined #openstack-infra | 06:09 | |
| *** achanda has joined #openstack-infra | 06:12 | |
| *** zz_sabari is now known as sabari | 06:13 | |
| *** armax has joined #openstack-infra | 06:14 | |
| *** HeOS has quit IRC | 06:15 | |
| *** achanda has quit IRC | 06:15 | |
| *** jkraj has joined #openstack-infra | 06:19 | |
| *** penguinRaider__ has quit IRC | 06:22 | |
| *** luqas__ has quit IRC | 06:23 | |
| *** jkraj has quit IRC | 06:25 | |
| *** luqas__ has joined #openstack-infra | 06:25 | |
| *** penguinRaider__ has joined #openstack-infra | 06:25 | |
| *** Sukhdev has quit IRC | 06:25 | |
| *** jamespage_ has quit IRC | 06:28 | |
| *** teran has joined #openstack-infra | 06:29 | |
| *** achanda has joined #openstack-infra | 06:31 | |
| *** amotoki_ has joined #openstack-infra | 06:33 | |
| *** Masahiro has quit IRC | 06:34 | |
| openstackgerrit | Khai Do proposed openstack-infra/jenkins-job-builder: deprecate postbuildscript onsuccess and onfailure parameter names https://review.openstack.org/139257 | 06:36 |
| *** arosen has quit IRC | 06:36 | |
| *** amotoki_ has quit IRC | 06:37 | |
| *** teran has quit IRC | 06:40 | |
| *** rmcall has quit IRC | 06:43 | |
| *** zz_avozza is now known as avozza | 06:43 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: WIP: sort oslo channels https://review.openstack.org/140933 | 06:45 |
| *** teran has joined #openstack-infra | 06:46 | |
| *** vigneshvar has joined #openstack-infra | 06:46 | |
| *** vigneshvar has quit IRC | 06:46 | |
| *** achanda has quit IRC | 06:47 | |
| *** vigneshvar has joined #openstack-infra | 06:48 | |
| *** mmaglana has quit IRC | 06:49 | |
| *** teran has quit IRC | 06:50 | |
| *** penguinRaider__ has quit IRC | 06:52 | |
| *** amuller has joined #openstack-infra | 06:53 | |
| *** penguinRaider__ has joined #openstack-infra | 06:56 | |
| *** achanda has joined #openstack-infra | 06:59 | |
| *** dims__ has quit IRC | 07:00 | |
| *** otter768 has joined #openstack-infra | 07:01 | |
| *** ildikov_afk has quit IRC | 07:02 | |
| *** tnurlygayanov has quit IRC | 07:02 | |
| *** teran has joined #openstack-infra | 07:03 | |
| *** andreaf has quit IRC | 07:04 | |
| *** salv-orlando has joined #openstack-infra | 07:05 | |
| *** arosen has joined #openstack-infra | 07:05 | |
| *** otter768 has quit IRC | 07:06 | |
| *** teran has quit IRC | 07:07 | |
| *** ildikov_afk has joined #openstack-infra | 07:08 | |
| *** tnurlygayanov has joined #openstack-infra | 07:08 | |
| *** _nadya_ has quit IRC | 07:08 | |
| *** Murad has joined #openstack-infra | 07:08 | |
| *** salv-orlando has quit IRC | 07:10 | |
| *** achanda has quit IRC | 07:11 | |
| openstackgerrit | Merged openstack/requirements: Bump up oslo.utils to 1.1.0 https://review.openstack.org/140573 | 07:11 |
| *** andreaf has joined #openstack-infra | 07:13 | |
| Murad | does anybody faced the issue when zuul catches gerrit change, but dont trigger the job | 07:13 |
| Murad | 2014-12-11 09:06:34,321 INFO zuul.Scheduler: Adding openstack-dev/sandbox, <Change 0x7ff0fc4dbb10 140937,1> to <Pipeline check-mlnx> 2014-12-11 09:06:34,322 INFO zuul.IndependentPipelineManager: Change <Change 0x7ff0fc4dbb10 140937,1> depends on changes [] 2014-12-11 09:06:34,455 INFO zuul.Gerrit: Updating information for 140937,1 2014-12-11 09:06:35,096 INFO zuul.Gerrit: Updating information for 140937,1 2014-12-11 09:06:35,68 | 07:13 |
| *** yamamot__ has quit IRC | 07:14 | |
| *** penguinRaider__ has quit IRC | 07:17 | |
| *** penguinRaider has joined #openstack-infra | 07:17 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config: Check that Gerrit ACL files are normalized https://review.openstack.org/140821 | 07:23 |
| *** camunoz has quit IRC | 07:24 | |
| *** tnovacik has joined #openstack-infra | 07:27 | |
| *** fandi has quit IRC | 07:29 | |
| *** AKPWD has joined #openstack-infra | 07:30 | |
| *** mpaolino has joined #openstack-infra | 07:30 | |
| *** luqas has joined #openstack-infra | 07:30 | |
| *** luqas__ has quit IRC | 07:30 | |
| *** jamielennox is now known as jamielennox|away | 07:33 | |
| *** fandi has joined #openstack-infra | 07:34 | |
| *** koolhead17 has quit IRC | 07:34 | |
| *** koolhead17 has joined #openstack-infra | 07:34 | |
| *** luqas__ has joined #openstack-infra | 07:37 | |
| *** liusheng has quit IRC | 07:37 | |
| *** liusheng has joined #openstack-infra | 07:38 | |
| *** Ryan_Lane has quit IRC | 07:38 | |
| *** luqas has quit IRC | 07:38 | |
| *** koolhead17 has quit IRC | 07:38 | |
| *** Ryan_Lane has joined #openstack-infra | 07:39 | |
| *** Ryan_Lane has quit IRC | 07:39 | |
| *** penguinRaider__ has joined #openstack-infra | 07:39 | |
| *** luqas___ has joined #openstack-infra | 07:40 | |
| *** luqas__ has quit IRC | 07:41 | |
| *** penguinRaider has quit IRC | 07:43 | |
| *** k4n0 has joined #openstack-infra | 07:44 | |
| *** Masahiro has joined #openstack-infra | 07:44 | |
| *** jcoufal has joined #openstack-infra | 07:46 | |
| *** mrunge has joined #openstack-infra | 07:48 | |
| *** cody-somerville has quit IRC | 07:49 | |
| *** zul has quit IRC | 07:49 | |
| *** AKPWD has left #openstack-infra | 07:51 | |
| *** Masahiro has quit IRC | 07:51 | |
| *** cody-somerville has joined #openstack-infra | 07:52 | |
| *** penguinRaider__ has quit IRC | 07:56 | |
| *** Masahiro has joined #openstack-infra | 07:56 | |
| *** penguinRaider__ has joined #openstack-infra | 07:59 | |
| *** jlibosva has joined #openstack-infra | 08:00 | |
| *** jpich has joined #openstack-infra | 08:01 | |
| *** zul has joined #openstack-infra | 08:03 | |
| *** jerryz_ has joined #openstack-infra | 08:03 | |
| *** jerryz has quit IRC | 08:06 | |
| *** e0ne has joined #openstack-infra | 08:09 | |
| *** nelsnelson has joined #openstack-infra | 08:12 | |
| *** wenlock has quit IRC | 08:12 | |
| *** Longgeek has joined #openstack-infra | 08:12 | |
| Murad | anybody here can help regarding jenkins and zuul ? | 08:15 |
| *** koolhead17 has joined #openstack-infra | 08:15 | |
| *** koolhead17 has joined #openstack-infra | 08:15 | |
| *** achanda has joined #openstack-infra | 08:15 | |
| *** penguinRaider has joined #openstack-infra | 08:16 | |
| *** penguinRaider__ has quit IRC | 08:16 | |
| *** patrickeast has quit IRC | 08:18 | |
| *** Longgeek has quit IRC | 08:21 | |
| *** amotoki_ has joined #openstack-infra | 08:21 | |
| *** doude has joined #openstack-infra | 08:22 | |
| *** liusheng has quit IRC | 08:22 | |
| *** liusheng has joined #openstack-infra | 08:23 | |
| *** Longgeek has joined #openstack-infra | 08:23 | |
| *** nfedotov has joined #openstack-infra | 08:24 | |
| *** amotoki_ has quit IRC | 08:26 | |
| *** avozza is now known as zz_avozza | 08:26 | |
| *** mkerrin has quit IRC | 08:29 | |
| AJaeger | Murad, looking at the questions you asked an hour ago: jhesketh might be around, the rest of the infra team is US based. | 08:30 |
| *** mrmartin has joined #openstack-infra | 08:31 | |
| *** stevemar has quit IRC | 08:31 | |
| *** jedimike has joined #openstack-infra | 08:32 | |
| *** arosen has quit IRC | 08:32 | |
| Murad | Thanks AJaeger | 08:33 |
| *** dizquierdo has joined #openstack-infra | 08:33 | |
| *** mpaolino has quit IRC | 08:34 | |
| *** dizquierdo has quit IRC | 08:39 | |
| *** stevemar has joined #openstack-infra | 08:40 | |
| *** zz_avozza is now known as avozza | 08:40 | |
| jhesketh | Murad: hey, I'm around although I don't know much about Jenkins | 08:41 |
| jhesketh | I'll try and help, what's up? | 08:41 |
| Murad | I am trying zuul with openstack-dev/sandbox | 08:42 |
| *** jcoufal has quit IRC | 08:42 | |
| Murad | I am doing a push to gerrit, zuul catches the event but not executing a jenkins job | 08:42 |
| *** jcoufal has joined #openstack-infra | 08:43 | |
| openstackgerrit | Ivan Udovichenko proposed openstack-infra/system-config: Remove hardcoded variables from openstack_project module (template) https://review.openstack.org/140758 | 08:43 |
| jhesketh | Murad: do you have your zuul reporting somewhere (eg back to gerrit or via smtp) | 08:44 |
| *** penguinRaider__ has joined #openstack-infra | 08:45 | |
| jhesketh | if zuul queues the job correctly but there are no workers it will report back "NOT REGISTERED" (as in the job hasn't got a registered gearman worker) | 08:45 |
| *** Ryan_Lane has joined #openstack-infra | 08:45 | |
| jhesketh | if you're getting that the problem is with Jenkins, if you're not getting that your zuul configuration isn't complete | 08:45 |
| *** stevemar has quit IRC | 08:45 | |
| *** Ryan_Lane has quit IRC | 08:46 | |
| *** mpaolino has joined #openstack-infra | 08:46 | |
| *** rushiagr has left #openstack-infra | 08:46 | |
| Murad | I can see the zuul catched the event in /var/logs/zuul/zuul.log only, but not going back to gerrit or smtp | 08:46 |
| Murad | the very same configurations work for me for local gerrit, but not openstack | 08:47 |
| *** arxcruz has joined #openstack-infra | 08:47 | |
| *** vdo has joined #openstack-infra | 08:48 | |
| *** penguinRaider has quit IRC | 08:48 | |
| *** MaxV has joined #openstack-infra | 08:52 | |
| jhesketh | Murad: right, if no reports are going back it's either not registering the event against a trigger or there are no jobs for the project etc | 08:52 |
| *** ildikov_afk is now known as ildikov | 08:52 | |
| *** salv-orlando has joined #openstack-infra | 08:52 | |
| jhesketh | Murad: would you mind pastebin-ing your layout.yaml? | 08:52 |
| Murad | sure, can I paste it here in the chat? | 08:53 |
| jhesketh | Murad: nope, here please: http://paste.openstack.org/ | 08:53 |
| jhesketh | then put the link back in chat ;-) | 08:54 |
| *** HeOS has joined #openstack-infra | 08:54 | |
| Murad | http://paste.openstack.org/show/149210/ | 08:54 |
| jhesketh | Murad: okay, that looks correct on first glance.. Would you mind pasting the relevant parts of the log? | 08:56 |
| *** armax has quit IRC | 08:56 | |
| Murad | http://paste.openstack.org/show/149211/ | 08:57 |
| *** salv-orlando has quit IRC | 08:57 | |
| *** dtantsur|afk is now known as dtantsur | 08:57 | |
| jhesketh | Murad: right, so it's not enquing a job because the patch didn't merge | 08:59 |
| jhesketh | Murad: check your zuul-merger.log | 08:59 |
| Murad | merge:false? | 08:59 |
| jhesketh | Murad: interestingly you should also be seeing a message back to gerrit, have you checked if you're posting on changes | 09:00 |
| Murad | yes there is no posts | 09:00 |
| jhesketh | Murad: umm, you should be running a zuul-merger daemon somewhere which will log to /var/log/zuul/merger.log (or something like that) | 09:00 |
| *** jcoufal has quit IRC | 09:01 | |
| *** jcoufal has joined #openstack-infra | 09:01 | |
| Murad | zuul-merger is running | 09:02 |
| *** otter768 has joined #openstack-infra | 09:02 | |
| Murad | http://paste.openstack.org/show/149213/ | 09:03 |
| jhesketh | Murad: anything interesting in the merger logs? | 09:04 |
| Murad | yes, there is excetions http://paste.openstack.org/show/149213/ | 09:04 |
| jhesketh | Murad: ah, well that's where I'd start | 09:05 |
| jhesketh | looks like your merger isn't configured correctly | 09:05 |
| Murad | but the same merger worked for another gerrit, not openstack gerrit? | 09:05 |
| *** jistr has joined #openstack-infra | 09:06 | |
| *** otter768 has quit IRC | 09:07 | |
| *** bradjones has quit IRC | 09:08 | |
| *** ala_ has joined #openstack-infra | 09:08 | |
| *** doude has quit IRC | 09:09 | |
| *** doude has joined #openstack-infra | 09:09 | |
| AJaeger | jhesketh, btw. could you approve https://review.openstack.org/#/c/128394/ , please? Just a small cleanup and thus everybody ignores it ;( | 09:09 |
| *** bradjones has joined #openstack-infra | 09:10 | |
| Murad | jhesketh, what is the proper conf for git_user_email and git_user_name under [merger] in zuul.conf? | 09:11 |
| *** achanda has quit IRC | 09:11 | |
| *** achanda has joined #openstack-infra | 09:12 | |
| *** derekh has joined #openstack-infra | 09:13 | |
| *** andreaf has quit IRC | 09:13 | |
| *** achanda_ has joined #openstack-infra | 09:13 | |
| *** talluri has joined #openstack-infra | 09:16 | |
| *** achanda has quit IRC | 09:16 | |
| *** mfink has quit IRC | 09:16 | |
| *** hashar has joined #openstack-infra | 09:18 | |
| *** hashar has quit IRC | 09:19 | |
| *** hashar has joined #openstack-infra | 09:20 | |
| *** achanda_ has quit IRC | 09:20 | |
| *** viktors|afk is now known as viktors | 09:21 | |
| *** penguinRaider__ has quit IRC | 09:21 | |
| *** bradjones has quit IRC | 09:24 | |
| *** andreykurilin_ has joined #openstack-infra | 09:24 | |
| jhesketh | AJaeger: done | 09:24 |
| AJaeger | thanks, jhesketh ! | 09:24 |
| *** penguinRaider__ has joined #openstack-infra | 09:25 | |
| jhesketh | Murad: I don't think they are used for authentication | 09:26 |
| jhesketh | I have mine set to the default | 09:26 |
| jhesketh | I think they are used where the merger needs to do a merge commit (it needs an author) | 09:26 |
| jhesketh | so it can be anything | 09:26 |
| *** mkerrin has joined #openstack-infra | 09:27 | |
| openstackgerrit | Merged openstack-infra/nodepool: Remove docutils pin https://review.openstack.org/128394 | 09:28 |
| *** Masahiro has quit IRC | 09:29 | |
| Murad | jhesketh, what do you suggest? | 09:29 |
| *** Masahiro has joined #openstack-infra | 09:30 | |
| *** mrmartin has quit IRC | 09:30 | |
| *** penguinRaider__ has quit IRC | 09:32 | |
| *** _shaps_ has joined #openstack-infra | 09:32 | |
| *** Longgeek has quit IRC | 09:33 | |
| *** mrunge has quit IRC | 09:34 | |
| *** sabari is now known as zz_sabari | 09:36 | |
| *** jgallard_ has joined #openstack-infra | 09:37 | |
| *** skolekonov has joined #openstack-infra | 09:39 | |
| jhesketh | Murad: (sorry between things) umm, possibly removing your /var/lib/zuul/git and restarting the services | 09:40 |
| *** e0ne has quit IRC | 09:41 | |
| jhesketh | if that doesn't work, it's complaining about access to git, so figure out the creds it is using and see if you can clone changes from gerrit with it | 09:41 |
| *** Longgeek has joined #openstack-infra | 09:42 | |
| *** bradjones has joined #openstack-infra | 09:43 | |
| *** ihrachyshka has joined #openstack-infra | 09:45 | |
| *** fandi has quit IRC | 09:45 | |
| *** Hal_ has joined #openstack-infra | 09:47 | |
| *** _nadya_ has joined #openstack-infra | 09:50 | |
| *** pblaho has joined #openstack-infra | 09:52 | |
| *** Masahiro has quit IRC | 09:52 | |
| *** ihrachyshka has quit IRC | 09:52 | |
| *** ihrachyshka has joined #openstack-infra | 09:53 | |
| *** hashar has quit IRC | 09:57 | |
| *** fandi has joined #openstack-infra | 09:58 | |
| Murad | jhesketh, I deleted that folder and it worked, thank you very very much | 09:59 |
| Murad | :) | 09:59 |
| *** mrunge has joined #openstack-infra | 10:00 | |
| *** mpaolino has quit IRC | 10:02 | |
| *** hashar has joined #openstack-infra | 10:05 | |
| *** ociuhandu has joined #openstack-infra | 10:07 | |
| *** Masahiro has joined #openstack-infra | 10:08 | |
| *** amotoki_ has joined #openstack-infra | 10:09 | |
| *** mpaolino has joined #openstack-infra | 10:14 | |
| *** amotoki_ has quit IRC | 10:14 | |
| *** andreaf has joined #openstack-infra | 10:15 | |
| *** boris-42 has joined #openstack-infra | 10:18 | |
| *** cnesa has joined #openstack-infra | 10:20 | |
| *** dizquierdo has joined #openstack-infra | 10:21 | |
| *** penguinRaider__ has joined #openstack-infra | 10:22 | |
| *** jp_at_hp has joined #openstack-infra | 10:27 | |
| jhesketh | Murad: woo :-) | 10:27 |
| nikil2 | Hi in my CI system i have configured the jenkisn master with zuul and jobs are ready. It is pointing to my internal gerrit server not to review.openstack.org. Now when any checkin happens the noop-check-communication job is in queu stating " Waiting for next available executer". | 10:30 |
| nikil2 | I have only master as of now and this job "noop-check" is configured with master only | 10:31 |
| *** jamielennox|away is now known as jamielennox | 10:32 | |
| nikil2 | Murad: Now you are able to run the jobs? | 10:33 |
| *** Masahiro has quit IRC | 10:33 | |
| nikil2 | Murad: Because i am at the same step of your setup but for the job is always in queu | 10:34 |
| *** Masahiro has joined #openstack-infra | 10:34 | |
| *** talluri has quit IRC | 10:35 | |
| *** tnovacik is now known as tnovacik|afk | 10:35 | |
| *** penguinRaider__ has quit IRC | 10:35 | |
| *** penguinRaider has joined #openstack-infra | 10:35 | |
| *** amotoki_ has joined #openstack-infra | 10:35 | |
| ttx | anteaya: ironic, tripleo, zaqar, manila, designate, barbican don't have stable-team-maintained stable bracnhes currently | 10:36 |
| *** andreykurilin_ has quit IRC | 10:36 | |
| *** unicell has quit IRC | 10:37 | |
| *** dizquierdo has quit IRC | 10:38 | |
| *** pelix has joined #openstack-infra | 10:39 | |
| *** zhiwei has quit IRC | 10:39 | |
| *** talluri has joined #openstack-infra | 10:39 | |
| *** dtantsur is now known as dtantsur|brb | 10:43 | |
| *** ociuhandu has quit IRC | 10:45 | |
| *** penguinRaider__ has joined #openstack-infra | 10:45 | |
| *** e0ne has joined #openstack-infra | 10:46 | |
| *** penguinRaider__ has quit IRC | 10:47 | |
| *** penguinRaider__ has joined #openstack-infra | 10:47 | |
| *** penguinRaider has quit IRC | 10:49 | |
| jkt | nikil2: check that you see some build jobs (definitions of endpoints, not really running jobs) within gear | 10:50 |
| jkt | nikil2: iirc `echo status | nc localhost 4307` or what's the port#again | 10:50 |
| *** koolhead17 has quit IRC | 10:52 | |
| *** kaeso has left #openstack-infra | 10:53 | |
| nikil2 | jkt: jenkins port? | 10:53 |
| jkt | nikil2: nope, a local port of gearman's server which runs on your zuul server | 10:54 |
| *** ZZelle_ has quit IRC | 10:57 | |
| *** ZZelle_ has joined #openstack-infra | 10:57 | |
| *** unicell has joined #openstack-infra | 10:58 | |
| nikil2 | jkt: in zuul.conf gearman's server=127.0.0.1 is mentioned but not the port number . so the default port is 4304? But i don't get any output for "echo status | nc localhost 4307" | 10:58 |
| nikil2 | i am running jenkins,zuul and gearman all in same server | 10:59 |
| *** Masahiro has quit IRC | 10:59 | |
| jkt | nikil2: you can use netstat with appropriate parameters to see where's it's listening on | 11:01 |
| jkt | nikil2: alternatively, zuul's documentation mentions the default port number as well | 11:02 |
| *** fandi has quit IRC | 11:03 | |
| *** otter768 has joined #openstack-infra | 11:03 | |
| *** jgallard_ has quit IRC | 11:03 | |
| nikil2 | jkt: it uses 4730 this is what i get => http://paste.openstack.org/show/149236/ currently i have stoped the job as always it will got for queu | 11:04 |
| *** penguinRaider__ has quit IRC | 11:05 | |
| nikil2 | jkt: stoped means i checkin the code then the job starts then it was in queu so i stoped it. Iike this 2 times i did | 11:05 |
| jkt | what does it mean that a job was in a queue? | 11:05 |
| jkt | does it mean that you show it in zuul's status output? | 11:06 |
| *** sandywalsh has quit IRC | 11:06 | |
| *** hashar has quit IRC | 11:07 | |
| *** penguinRaider__ has joined #openstack-infra | 11:07 | |
| *** otter768 has quit IRC | 11:07 | |
| *** dizquierdo has joined #openstack-infra | 11:08 | |
| *** sandywalsh has joined #openstack-infra | 11:08 | |
| *** tnovacik|afk has quit IRC | 11:08 | |
| *** adalbas has joined #openstack-infra | 11:08 | |
| *** rlandy has joined #openstack-infra | 11:09 | |
| nikil2 | jkt: No , when i check in the code to gerrit review -> zuul tirggers gerarman -> gearman starts the jenkins job -> Now the job is in queu | 11:10 |
| jkt | nikil2: what sort of a queue is that? A Jenkins queue, maybe? | 11:10 |
| nikil2 | jkt : yes jenkins queue the job name "noop-check-communication" tells "Waiting for an available executor" | 11:13 |
| openstackgerrit | Ricardo Carrillo Cruz proposed openstack-infra/infra-specs: Create Nodepool REST API spec https://review.openstack.org/141016 | 11:14 |
| jkt | nikil2: seems like a configuration problem of Jenkins -- do you have any slaves, or any other executor which can perform your job? | 11:15 |
| nikil2 | jkt: I have not configured any slaves now. I was about to test first all the communication works find for CI system. May be i will configure a new slave and try in that? | 11:16 |
| *** dizquierdo has quit IRC | 11:17 | |
| nikil2 | currently only master node is there and this job will run in master itself i guess. its just for testing the communication purpose i am doing | 11:17 |
| *** penguinRaider__ has quit IRC | 11:17 | |
| *** asselin has quit IRC | 11:19 | |
| *** obondarev has joined #openstack-infra | 11:20 | |
| *** penguinRaider__ has joined #openstack-infra | 11:21 | |
| nikil2 | jkt: could you please explain me what this -> http://paste.openstack.org/show/149236/ output tells | 11:28 |
| *** aysyd has joined #openstack-infra | 11:28 | |
| *** dizquierdo has joined #openstack-infra | 11:29 | |
| *** _nadya_ has quit IRC | 11:30 | |
| jkt | nikil2: http://stackoverflow.com/questions/20750896/gearman-gearadmin-command-line-tool-status-output-meaning | 11:31 |
| *** markus_z has joined #openstack-infra | 11:31 | |
| nikil2 | jkt: thanks i will look into it. Also parally i am trying to set up another slave node and will try to run the job in that | 11:35 |
| *** penguinRaider__ has quit IRC | 11:37 | |
| *** tnovacik|afk has joined #openstack-infra | 11:37 | |
| *** viktors is now known as viktors|afk | 11:38 | |
| *** social has joined #openstack-infra | 11:39 | |
| *** penguinRaider__ has joined #openstack-infra | 11:40 | |
| *** dizquierdo has quit IRC | 11:41 | |
| *** salv-orlando has joined #openstack-infra | 11:43 | |
| social | hi, we managed to break our repo with tags is there way to get two tags removed from packstack ? | 11:44 |
| *** salv-orlando has quit IRC | 11:47 | |
| *** tnovacik|afk is now known as tnovacik | 11:48 | |
| *** cdent has joined #openstack-infra | 11:51 | |
| *** rcarrill` is now known as rcarrillocruz | 11:52 | |
| *** dizquierdo has joined #openstack-infra | 11:54 | |
| *** luqas___ has quit IRC | 11:55 | |
| *** luqas has joined #openstack-infra | 11:55 | |
| *** e0ne has quit IRC | 11:56 | |
| *** penguinRaider__ has quit IRC | 11:58 | |
| *** penguinRaider has joined #openstack-infra | 11:58 | |
| *** penguinRaider__ has joined #openstack-infra | 11:59 | |
| *** dizquierdo has quit IRC | 11:59 | |
| *** Masahiro has joined #openstack-infra | 12:00 | |
| openstackgerrit | Ivan Udovichenko proposed openstack-infra/system-config: Remove hardcoded variables from openstack_project module (thick_slave) https://review.openstack.org/141027 | 12:03 |
| *** penguinRaider has quit IRC | 12:03 | |
| *** amuller is now known as amuller_afk | 12:03 | |
| *** amotoki has quit IRC | 12:03 | |
| *** amotoki_ is now known as amotoki | 12:03 | |
| *** e0ne has joined #openstack-infra | 12:04 | |
| *** Masahiro has quit IRC | 12:04 | |
| *** yamamoto has joined #openstack-infra | 12:05 | |
| *** mattymo has joined #openstack-infra | 12:08 | |
| *** dmsimard_away is now known as dmsimard | 12:09 | |
| *** reed has joined #openstack-infra | 12:09 | |
| * social pings SergeyLukjanov | 12:11 | |
| *** penguinRaider__ has quit IRC | 12:11 | |
| *** penguinRaider has joined #openstack-infra | 12:11 | |
| *** doug-fish has joined #openstack-infra | 12:11 | |
| *** arxcruz has quit IRC | 12:12 | |
| *** arxcruz has joined #openstack-infra | 12:12 | |
| *** dizquierdo has joined #openstack-infra | 12:12 | |
| *** mpaolino has quit IRC | 12:12 | |
| *** baoli has joined #openstack-infra | 12:12 | |
| *** hashar has joined #openstack-infra | 12:12 | |
| *** enikanorov_ has joined #openstack-infra | 12:13 | |
| *** baoli has quit IRC | 12:14 | |
| *** baoli has joined #openstack-infra | 12:15 | |
| *** enikanorov has quit IRC | 12:15 | |
| *** reed has quit IRC | 12:16 | |
| *** dizquierdo has quit IRC | 12:17 | |
| *** penguinRaider__ has joined #openstack-infra | 12:19 | |
| *** jgallard_ has joined #openstack-infra | 12:19 | |
| *** andymaier has joined #openstack-infra | 12:22 | |
| *** yaguang has quit IRC | 12:22 | |
| *** penguinRaider has quit IRC | 12:23 | |
| *** oomichi has quit IRC | 12:23 | |
| nikil2 | jkt : i have used a seprate slave node and now the "noop-communication" job is working fine . Thanks | 12:24 |
| social | fungi: hi, what is the procedure for removing broken git tags from gerrit ? | 12:26 |
| *** julim has joined #openstack-infra | 12:28 | |
| *** dizquierdo has joined #openstack-infra | 12:29 | |
| SergeyLukjanov | social, pong | 12:30 |
| SergeyLukjanov | social, broken? | 12:30 |
| *** penguinRaider__ has quit IRC | 12:33 | |
| *** penguinRaider__ has joined #openstack-infra | 12:33 | |
| *** dprince has joined #openstack-infra | 12:33 | |
| *** penguinRaider__ has quit IRC | 12:35 | |
| *** penguinRaider__ has joined #openstack-infra | 12:35 | |
| *** andymaier has quit IRC | 12:37 | |
| *** andymaier has joined #openstack-infra | 12:38 | |
| *** obondarev has quit IRC | 12:40 | |
| *** dtantsur|brb is now known as dtantsur | 12:43 | |
| *** e0ne has quit IRC | 12:44 | |
| *** obondarev has joined #openstack-infra | 12:46 | |
| *** eharney has quit IRC | 12:46 | |
| *** radez_g0n3 is now known as radez | 12:46 | |
| *** lttrl has joined #openstack-infra | 12:50 | |
| *** jgrimm is now known as zz_jgrimm | 12:53 | |
| *** kgiusti has joined #openstack-infra | 12:55 | |
| *** dizquierdo has quit IRC | 12:58 | |
| *** jedimike has quit IRC | 12:59 | |
| *** ildikov has quit IRC | 12:59 | |
| *** penguinRaider__ has quit IRC | 13:00 | |
| *** penguinRaider has joined #openstack-infra | 13:00 | |
| *** ildikov has joined #openstack-infra | 13:00 | |
| *** zz_jgrimm is now known as jgrimm | 13:01 | |
| *** jedimike has joined #openstack-infra | 13:01 | |
| *** belmoreira has joined #openstack-infra | 13:01 | |
| *** Adri2000 has quit IRC | 13:03 | |
| *** dmakogon_ is now known as denis_makgoon | 13:03 | |
| *** otter768 has joined #openstack-infra | 13:04 | |
| *** denis_makgoon is now known as denis_makogon | 13:05 | |
| openstackgerrit | Dmitry Teselkin proposed openstack-infra/zuul: Whitelist for gerrit events https://review.openstack.org/137634 | 13:07 |
| *** Adri2000 has joined #openstack-infra | 13:08 | |
| *** otter768 has quit IRC | 13:08 | |
| *** jcoufal_ has joined #openstack-infra | 13:09 | |
| *** mpaolino has joined #openstack-infra | 13:09 | |
| *** jcoufal has quit IRC | 13:12 | |
| *** jaypipes has quit IRC | 13:14 | |
| *** AJaeger has quit IRC | 13:16 | |
| *** jaypipes has joined #openstack-infra | 13:17 | |
| *** e0ne has joined #openstack-infra | 13:18 | |
| *** mbacchi has joined #openstack-infra | 13:22 | |
| *** penguinRaider__ has joined #openstack-infra | 13:22 | |
| *** amuller_afk is now known as amuller | 13:23 | |
| *** mbacchi has quit IRC | 13:24 | |
| *** mbacchi has joined #openstack-infra | 13:24 | |
| openstackgerrit | Dmitry Teselkin proposed openstack-infra/zuul: Whitelist for gerrit events https://review.openstack.org/137634 | 13:25 |
| *** penguinRaider has quit IRC | 13:26 | |
| *** bswartz has quit IRC | 13:26 | |
| openstackgerrit | Dmitry Teselkin proposed openstack-infra/zuul: Whitelist for gerrit events https://review.openstack.org/137634 | 13:31 |
| *** vigneshvar has quit IRC | 13:32 | |
| social | SergeyLukjanov: we misunderstood new release cycle and tagged wrong | 13:33 |
| social | SergeyLukjanov: we just need to remove the tags from git to unbreak our build process :) | 13:34 |
| *** ociuhandu has joined #openstack-infra | 13:35 | |
| *** Sukhdev has joined #openstack-infra | 13:37 | |
| *** bknudson has quit IRC | 13:39 | |
| social | SergeyLukjanov: the thing is I don't know what is OK, should we just remove the tags from github git repo or do we have to add forcepush right to gerrit and forcepush removal there? | 13:40 |
| *** penguinRaider__ has quit IRC | 13:42 | |
| *** penguinRaider__ has joined #openstack-infra | 13:42 | |
| *** marcusvrn has joined #openstack-infra | 13:43 | |
| *** amitgandhinz has joined #openstack-infra | 13:44 | |
| *** penguinRaider__ has quit IRC | 13:44 | |
| *** AJaeger has joined #openstack-infra | 13:44 | |
| *** AJaeger has joined #openstack-infra | 13:44 | |
| *** amitgandhinz has quit IRC | 13:45 | |
| *** penguinRaider has joined #openstack-infra | 13:47 | |
| jd__ | any idea on this failure: https://review.openstack.org/#/c/140037/ logs at: http://logs.openstack.org/37/140037/4/check/gate-tempest-dsvm-neutron-src-tooz/aad8a5d/logs/devstacklog.txt.gz | 13:48 |
| *** Masahiro has joined #openstack-infra | 13:48 | |
| *** dkliban_afk is now known as dkliban | 13:49 | |
| *** achuprin_ has joined #openstack-infra | 13:50 | |
| *** eharney has joined #openstack-infra | 13:52 | |
| *** skolekonov has quit IRC | 13:52 | |
| *** Masahiro has quit IRC | 13:53 | |
| *** tnovacik has quit IRC | 13:54 | |
| *** amitgandhinz has joined #openstack-infra | 13:57 | |
| openstackgerrit | Jean-Frédéric proposed openstack-infra/jenkins-job-builder: Add Mercurial plugin feature to jenkins-job-builder https://review.openstack.org/139136 | 13:59 |
| *** wznoinsk has joined #openstack-infra | 14:01 | |
| *** pblaho has quit IRC | 14:03 | |
| openstackgerrit | Merged openstack-infra/storyboard: Converted application to use utcnow() https://review.openstack.org/140148 | 14:04 |
| *** signed8b_ is now known as signed8bit | 14:05 | |
| nikil2 | in system-config/install_modules.sh or other scripts are there to install jenkins and zuul . Is there any script to make a "slave-machine" so that i can add that slave machine in the jobs wich will then install devstack-gate and run the tests | 14:06 |
| *** mrmartin has joined #openstack-infra | 14:08 | |
| *** penguinRaider__ has joined #openstack-infra | 14:08 | |
| openstackgerrit | Merged openstack-infra/storyboard-webclient: First task tracks story title. https://review.openstack.org/140108 | 14:10 |
| *** bswartz has joined #openstack-infra | 14:10 | |
| *** andreykurilin_ has joined #openstack-infra | 14:10 | |
| *** penguinRaider has quit IRC | 14:12 | |
| *** hdd has joined #openstack-infra | 14:12 | |
| *** ryanpetrello has joined #openstack-infra | 14:16 | |
| *** miqui_ has joined #openstack-infra | 14:21 | |
| *** mriedem has joined #openstack-infra | 14:21 | |
| *** salv-orlando has joined #openstack-infra | 14:22 | |
| *** bknudson has joined #openstack-infra | 14:23 | |
| *** cpowell has joined #openstack-infra | 14:25 | |
| *** Sukhdev has quit IRC | 14:26 | |
| *** shayneburgess has joined #openstack-infra | 14:28 | |
| *** belmoreira has quit IRC | 14:29 | |
| *** Jeson has joined #openstack-infra | 14:36 | |
| *** tnovacik has joined #openstack-infra | 14:36 | |
| mtreinish | mordred: I'll spin up a local test to see if the proxy will send show processlist or not | 14:37 |
| *** penguinRaider has joined #openstack-infra | 14:37 | |
| *** obondarev has quit IRC | 14:37 | |
| *** spzala has joined #openstack-infra | 14:37 | |
| *** penguinRaider__ has quit IRC | 14:37 | |
| *** Jeson has quit IRC | 14:38 | |
| *** mattfarina has joined #openstack-infra | 14:38 | |
| *** dkranz has joined #openstack-infra | 14:39 | |
| *** mrmartin has quit IRC | 14:40 | |
| *** mrunge has quit IRC | 14:41 | |
| *** koolhead17 has joined #openstack-infra | 14:43 | |
| *** mrmartin has joined #openstack-infra | 14:44 | |
| *** _nadya_ has joined #openstack-infra | 14:47 | |
| *** achuprin_ has quit IRC | 14:48 | |
| *** achuprin_ has joined #openstack-infra | 14:48 | |
| *** rmcall has joined #openstack-infra | 14:50 | |
| *** ryanpetrello has quit IRC | 14:51 | |
| *** jpich has quit IRC | 14:51 | |
| *** ryanpetrello has joined #openstack-infra | 14:52 | |
| *** carl_baldwin has joined #openstack-infra | 14:53 | |
| *** dizquierdo has joined #openstack-infra | 14:54 | |
| *** signed8bit has quit IRC | 14:54 | |
| *** r-daneel has quit IRC | 14:54 | |
| *** pc_m has joined #openstack-infra | 14:55 | |
| *** jpich has joined #openstack-infra | 14:56 | |
| *** ayoung has quit IRC | 14:56 | |
| *** signed8bit has joined #openstack-infra | 14:57 | |
| *** Murad has quit IRC | 14:58 | |
| *** tnovacik has quit IRC | 14:59 | |
| mordred | nibalizer: talk to me | 14:59 |
| mtreinish | mordred: I just spun up the policy from the patch and running show processlist did not work, the query got dropped | 14:59 |
| mordred | mtreinish: cool | 15:00 |
| *** salv-orlando has quit IRC | 15:00 | |
| *** penguinRaider__ has joined #openstack-infra | 15:00 | |
| *** tnovacik has joined #openstack-infra | 15:02 | |
| *** pradk has joined #openstack-infra | 15:04 | |
| *** penguinRaider has quit IRC | 15:04 | |
| *** dims has joined #openstack-infra | 15:04 | |
| *** otter768 has joined #openstack-infra | 15:05 | |
| *** dustins has joined #openstack-infra | 15:05 | |
| *** Hal_ has quit IRC | 15:06 | |
| *** ayoung has joined #openstack-infra | 15:08 | |
| *** talluri has quit IRC | 15:09 | |
| *** dizquierdo has quit IRC | 15:09 | |
| *** skraynev has left #openstack-infra | 15:09 | |
| *** otter768 has quit IRC | 15:10 | |
| nikil2 | In thirdparty CI the communication part works fine. Instead of noop-communication job when i run "dsvm-tempest-full" it starts to communicate with devstack_slave node and job starts fine but devstack-gate instllation fails due to lots of dependencies . So i guess there is some way to prepare the slave nodes | 15:11 |
| nikil2 | is there any predeploy scripts available to prepare slave nodes | 15:11 |
| *** rmcall has quit IRC | 15:11 | |
| *** obondarev has joined #openstack-infra | 15:11 | |
| *** ddieterly has joined #openstack-infra | 15:13 | |
| *** penguinRaider__ has quit IRC | 15:15 | |
| *** penguinRaider has joined #openstack-infra | 15:15 | |
| *** yamamoto_ has joined #openstack-infra | 15:15 | |
| *** yamamoto has quit IRC | 15:16 | |
| *** obondarev has quit IRC | 15:17 | |
| *** arxcruz has quit IRC | 15:18 | |
| openstackgerrit | Merged openstack-infra/project-config: Run ceilometer rally scenarios https://review.openstack.org/132650 | 15:21 |
| *** dizquierdo has joined #openstack-infra | 15:22 | |
| *** asselin has joined #openstack-infra | 15:22 | |
| openstackgerrit | Jaroslav Henner proposed openstack-infra/jenkins-job-builder: Allow multiple comment-added events in gerrit trig. https://review.openstack.org/124568 | 15:24 |
| openstackgerrit | Ivan Udovichenko proposed openstack-infra/system-config: Remove hardcoded variables from openstack_project module (users_install) https://review.openstack.org/141069 | 15:24 |
| *** timcline has joined #openstack-infra | 15:26 | |
| *** timcline has quit IRC | 15:26 | |
| *** Hal_ has joined #openstack-infra | 15:26 | |
| *** penguinRaider__ has joined #openstack-infra | 15:26 | |
| openstackgerrit | Ivan Udovichenko proposed openstack-infra/system-config: Remove hardcoded variables from openstack_project module (users_install) https://review.openstack.org/141069 | 15:27 |
| *** timcline has joined #openstack-infra | 15:27 | |
| *** mestery has joined #openstack-infra | 15:27 | |
| *** penguinRaider has quit IRC | 15:30 | |
| *** dizquierdo has quit IRC | 15:31 | |
| *** boris-42 has quit IRC | 15:32 | |
| openstackgerrit | Jesse Pretorius proposed openstack-infra/project-config: Change os-ansible-aio-build check settings https://review.openstack.org/141072 | 15:33 |
| openstackgerrit | Ivan Udovichenko proposed openstack-infra/system-config: Remove hardcoded commands https://review.openstack.org/141073 | 15:35 |
| openstackgerrit | Derek Higgins proposed openstack-infra/system-config: Add hp1 back running tripleo ci https://review.openstack.org/126513 | 15:36 |
| *** Masahiro has joined #openstack-infra | 15:37 | |
| openstackgerrit | Thanh Ha proposed openstack-infra/jenkins-job-builder: Add support for new patchset-created exclude fields https://review.openstack.org/140095 | 15:41 |
| openstackgerrit | Thanh Ha proposed openstack-infra/jenkins-job-builder: Add support for Gerrit Trigger Comment Contains Expression https://review.openstack.org/136605 | 15:41 |
| *** Masahiro has quit IRC | 15:41 | |
| *** zz_dimtruck is now known as dimtruck | 15:44 | |
| *** pblaho has joined #openstack-infra | 15:44 | |
| *** markmcclain has joined #openstack-infra | 15:45 | |
| *** boris-42 has joined #openstack-infra | 15:45 | |
| *** shakamunyi has joined #openstack-infra | 15:45 | |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/python-storyboardclient: Addind apiclient from oslo incubator https://review.openstack.org/138091 | 15:45 |
| *** jerryz_ has quit IRC | 15:46 | |
| *** tonytan4ever has joined #openstack-infra | 15:47 | |
| *** hashar_ has joined #openstack-infra | 15:47 | |
| *** shayneburgess has quit IRC | 15:48 | |
| *** MaxV has quit IRC | 15:48 | |
| *** erikmwilson is now known as Guest48377 | 15:48 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/devstack-gate: Revert "Temporarily disable testing lbaas, fwaas, and vpnaas" https://review.openstack.org/140864 | 15:49 |
| openstackgerrit | Matthew Treinish proposed openstack-infra/devstack-gate: Add neutron advanced services to PROJECTS list https://review.openstack.org/141077 | 15:49 |
| *** asselin_ has joined #openstack-infra | 15:49 | |
| *** hashar has quit IRC | 15:50 | |
| *** erikmwilson has joined #openstack-infra | 15:50 | |
| openstackgerrit | Mauro Rodrigues proposed openstack-infra/devstack-gate: Add Error message for max attempts reached on git remote update https://review.openstack.org/141080 | 15:51 |
| mtreinish | mestery, dougwig: ^^^ that should work now. (well after the first 1 lands an a new nodepool image is kicked off) | 15:52 |
| *** wenlock has joined #openstack-infra | 15:52 | |
| mtreinish | sdague, clarkb: if you get a sec can you take a look at: https://review.openstack.org/141077 and https://review.openstack.org/140864 | 15:52 |
| openstackgerrit | Michael Krotscheck proposed openstack-infra/python-storyboardclient: Adding apiclient from oslo incubator https://review.openstack.org/138091 | 15:53 |
| *** asselin has quit IRC | 15:53 | |
| *** AJaeger has quit IRC | 15:53 | |
| mestery | mtreinish: Looking | 15:54 |
| krotscheck | SergeyLukjanov: Can I get a quick on ^^? I updated the commit message and the previous approvals went away. | 15:54 |
| sdague | mtreinish: do we actually need all the PROJECT entries? | 15:54 |
| sdague | I thought at one point it was all cloned | 15:54 |
| SergeyLukjanov | krotscheck, sure | 15:54 |
| mtreinish | sdague: oh, I assumed it was still needed. It might have changed since the last time I did this though | 15:55 |
| mestery | thanks mtreinish, those look good, waiting on Jenkins now :) | 15:55 |
| sdague | yeh, we should confirm with clarkb / fungi / jeblair | 15:55 |
| *** carl_baldwin has quit IRC | 15:55 | |
| clarkb | sdague we do. everything is cloned in the cache but only given to ddvstack if in the projects list iirc | 15:55 |
| sdague | ok | 15:56 |
| sdague | so https://review.openstack.org/#/c/141077/ is required | 15:56 |
| SergeyLukjanov | krotscheck, +2'd | 15:56 |
| *** penguinRaider__ has quit IRC | 15:56 | |
| *** penguinRaider has joined #openstack-infra | 15:56 | |
| *** dimtruck is now known as zz_dimtruck | 15:57 | |
| clarkb | sdague yes I think so. the tests should confirm and if you check setup workspace logs with and without 141077 that should make it clear | 15:57 |
| *** emagana has joined #openstack-infra | 15:57 | |
| *** rmcall has joined #openstack-infra | 15:57 | |
| *** zz_dimtruck is now known as dimtruck | 15:57 | |
| *** pcrews has joined #openstack-infra | 15:58 | |
| mtreinish | clarkb: well, 1410864 failed without it yesterday which is why I added 141077 | 15:58 |
| *** JayJ has joined #openstack-infra | 15:58 | |
| *** atiwari has joined #openstack-infra | 15:59 | |
| *** MaxV has joined #openstack-infra | 16:00 | |
| *** mfink has joined #openstack-infra | 16:01 | |
| *** JayJ has quit IRC | 16:02 | |
| *** JayJ has joined #openstack-infra | 16:03 | |
| *** zz_johnthetubagu is now known as johnthetubaguy | 16:03 | |
| openstackgerrit | Merged openstack-infra/python-storyboardclient: Adding apiclient from oslo incubator https://review.openstack.org/138091 | 16:04 |
| *** hdd has quit IRC | 16:04 | |
| *** changbl has quit IRC | 16:05 | |
| *** marun has joined #openstack-infra | 16:07 | |
| *** JayJ has quit IRC | 16:07 | |
| *** JayJ has joined #openstack-infra | 16:08 | |
| *** thedodd has joined #openstack-infra | 16:08 | |
| *** asselin_ has quit IRC | 16:08 | |
| *** jlibosva has quit IRC | 16:09 | |
| *** asselin has joined #openstack-infra | 16:11 | |
| *** shakamunyi has quit IRC | 16:12 | |
| *** JayJ has quit IRC | 16:12 | |
| *** JayJ has joined #openstack-infra | 16:13 | |
| *** JayJ has quit IRC | 16:13 | |
| *** jlibosva has joined #openstack-infra | 16:13 | |
| *** jcoufal_ has quit IRC | 16:14 | |
| *** asselin has quit IRC | 16:14 | |
| *** denis_makogon has quit IRC | 16:15 | |
| *** ihrachyshka has quit IRC | 16:15 | |
| *** wenlock has quit IRC | 16:15 | |
| annegentle | I'm ready to delete four repos from the Docs Program, and I'm not sure how that happens. Is it an infra request similar to a rename? | 16:15 |
| *** isviridov is now known as isviridov_away | 16:16 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/project-config: sort oslo channels https://review.openstack.org/140933 | 16:16 |
| *** jp_at_hp has quit IRC | 16:16 | |
| annegentle | I'm trying to remember an example of a repo deletion, can't come up with one. | 16:17 |
| mtreinish | jeblair: so I was looking into your comments on https://review.openstack.org/135732 I don't think the explanatory text is in the json payload from gerrit | 16:18 |
| jeblair | annegentle: we never delete anything; but we can move them to openstack-attic/ | 16:18 |
| annegentle | jeblair: oh I like that even better, I was coming up with a backup strategy myself but of course we already have one. | 16:18 |
| jeblair | annegentle: yep! the internet! :) | 16:18 |
| annegentle | jeblair: ok so it's a rename, is that a request that happens weekly or some such? | 16:18 |
| mtreinish | I dumped the dict in sync.py when it creates the label objects and all it had was: {u'Workflow': [u'-1', u' 0', u'+1'], u'Code-Review': [u'-2', u'-1', u' 0', u'+1', u'+2']} | 16:19 |
| jeblair | annegentle: well, periodically, when we accumulate enough that we can't ignore them anymore ;) | 16:19 |
| annegentle | jeblair: so maybe I'll wait until I get two more I can delete :) | 16:19 |
| annegentle | deletions are motivating | 16:20 |
| *** jp_at_hp has joined #openstack-infra | 16:20 | |
| jeblair | annegentle: if you put in a change to project-config to do the rename, the machine will start moving; alternately, add them to the infra meeting wiki page. but yeah, if you expect a few, waiting till they are all ready is nice | 16:20 |
| annegentle | nice. okay thanks jeblair | 16:20 |
| jeblair | annegentle: maybe in the mean time, add something to the readme/contributing saying they are defunct | 16:20 |
| *** jp_at_hp has quit IRC | 16:21 | |
| annegentle | jeblair: yeah I've been communicating "Frozen" but makes sense to update the readme | 16:21 |
| jeblair | annegentle: (that's actually all we do for the infra repos; we leave them in place but update their readmes; there's no infra attic) | 16:21 |
| *** armax has joined #openstack-infra | 16:23 | |
| *** jpich has quit IRC | 16:24 | |
| cody-somerville | Where can one see the official list of current PTLs? | 16:24 |
| mtreinish | jeblair: do you still want me too add colors if it's just the numbers? | 16:25 |
| mtreinish | cody-somerville: http://git.openstack.org/cgit/openstack/governance/tree/reference/programs.yaml | 16:25 |
| clarkb | mtreinish: I want to look at the logs like I suggsted simply beacuse I suggested it and figure I should do that :) but I expect them to all look good and approve once I have a chance to look | 16:26 |
| *** zz_sabari is now known as sabari_ | 16:26 | |
| mtreinish | clarkb: oh, I wasn't pushing to rush anything :) I was just saying it failed before in agreement with your suggestion | 16:27 |
| jeblair | mtreinish: yeah, i think we should add the colors to just the numbers for now. the descriptions are available in the api, it's probably a teensy bit more work for them (maybe a schema change) | 16:29 |
| jeblair | mtreinish: "values":{"-2":"Do not merge","-1":"This patch needs further work before it can be merged"," 0":"No score","+1":"Looks good to me, but someone else must approve","+2":"Looks good to me (core reviewer)"}} | 16:30 |
| jeblair | mtreinish: ... ^ | 16:30 |
| *** penguinRaider__ has joined #openstack-infra | 16:30 | |
| jeblair | that's in the change query response | 16:30 |
| mtreinish | jeblair: ok, I'll respin 135732 with colors, and then look into adding the text in a follow on | 16:31 |
| *** craigbr has joined #openstack-infra | 16:31 | |
| jeblair | mtreinish: groovy thanks. i think either of those will help and both together will be double plus good | 16:31 |
| openstackgerrit | Andrea Frittoli proposed openstack-infra/project-config: Fix neutron jobs for tempest on stable branches https://review.openstack.org/140503 | 16:32 |
| clarkb | jeblair: fyi http://lists.openstack.org/pipermail/openstack-dev/2014-December/052757.html still not qite sure what the intent is there | 16:32 |
| *** pblaho has quit IRC | 16:32 | |
| *** penguinRaider has quit IRC | 16:34 | |
| jeblair | clarkb: wow that message just trails off into nothing | 16:35 |
| *** salv-orlando has joined #openstack-infra | 16:35 | |
| cody-somerville | mtreinish: Thanks! | 16:35 |
| clarkb | mtreinish: looks like ironic pxe ssh may still be broken? | 16:36 |
| *** rmcall has quit IRC | 16:36 | |
| clarkb | jeblair: I think the assertion of the problem is not detailed enough for me to understand the problem | 16:38 |
| clarkb | which made udnerstadning the rest of the email hard. (we have lots of rechecks at feature freeze time not beacuse of drivers but beacuse everyone must get all the code in at all costs and we merge tons of code which leads to instability) | 16:39 |
| mtreinish | clarkb: hmm, I think it's unrelated. It's just an ironic ops job that's failing | 16:39 |
| mtreinish | adam_g: http://logs.openstack.org/64/140864/2/check/check-tempest-dsvm-ironic-pxe_ssh/35a03a9/console.html#_2014-12-11_16_34_50_999 | 16:40 |
| mtreinish | clarkb: it times out waiting for the ironic node it's booting to become active | 16:41 |
| *** bauzas has joined #openstack-infra | 16:41 | |
| *** salv-orlando has quit IRC | 16:41 | |
| bauzas | hi, I have a problem with my local Tempest install, because pbr complains this is not the right version in setup.cfg | 16:41 |
| clarkb | bauzas: is your tempest up to date? | 16:42 |
| bauzas | clarkb: well, I was using RECLONE=yes in my localrc | 16:42 |
| bauzas | clarkb: but I can check, for sure | 16:42 |
| mtreinish | clarkb: oops, I forgot to bump the setup.cfg version again | 16:42 |
| mtreinish | we pushed a new tag on Mon | 16:42 |
| mtreinish | let me fix that, one sec | 16:42 |
| *** markus_z has quit IRC | 16:42 | |
| clarkb | bauzas: mtreinish broke it :) | 16:42 |
| bauzas | mtreinish: oh ok, because I wasn't seeing any bug neither review | 16:43 |
| *** tkelsey has joined #openstack-infra | 16:43 | |
| bauzas | mtreinish: feel free to ping me once you publish the change, I'll cherry-pick it directly on my local branch | 16:43 |
| *** salv-orlando has joined #openstack-infra | 16:43 | |
| mtreinish | I'm not sure why the gate didn't catch that though, we've landed newer patches since the tag | 16:43 |
| *** sabari_ is now known as zz_sabari_ | 16:44 | |
| mtreinish | pbr should have exploded in the gate too | 16:44 |
| bauzas | mtreinish: agreed, hence my wonders | 16:44 |
| clarkb | mtreinish: first hunch is we ma not be installing pbr from source in the gate? | 16:44 |
| clarkb | mtreinish: did we release a version of pbr with the version hammer stuff? | 16:44 |
| bauzas | just to be clear, my local branch is up-to-date | 16:45 |
| bauzas | I just checked this | 16:45 |
| mtreinish | clarkb: hmm, dunno. I assumed it was in a release, because I've hit this before | 16:45 |
| mtreinish | that's why I knew it was my fault again :) | 16:45 |
| mtreinish | dhellmann: ^^^ ? | 16:45 |
| clarkb | mtreinish: well efore we didn't have sdague's lib from pypi stuff | 16:45 |
| * dhellmann reads the scrollback | 16:46 | |
| *** amotoki has quit IRC | 16:46 | |
| *** david-lyle_afk is now known as david-lyle | 16:46 | |
| bauzas | mtreinish: clarkb: here is the error http://paste.openstack.org/show/149461/ | 16:46 |
| bauzas | dhellmann: ^ | 16:46 |
| dhellmann | mtreinish: I haven't done a new pbr release lately. | 16:46 |
| bauzas | is this something PEBKAC ? | 16:47 |
| dhellmann | are you installing pbr from source? | 16:47 |
| bauzas | I can't see where I'm wrong | 16:48 |
| pleia2 | so, more unusual rain here in SF, the power actually went out last night, first time in 5 years living here | 16:48 |
| mtreinish | bauzas: https://review.openstack.org/#/c/141098/ | 16:48 |
| dhellmann | there are some known issues in the master branch of pbr, so I don't think we want it installed from source | 16:48 |
| *** andymaier has quit IRC | 16:48 | |
| pleia2 | should be an interesting day, schools are closed, my husband is working from home | 16:48 |
| bauzas | dhellmann: lemme check | 16:48 |
| clarkb | dhellmann: ya Ithink this may be pbr from source | 16:48 |
| bauzas | pbr==0.11.0.dev44.ge0d2674 | 16:48 |
| bauzas | gotcha | 16:48 |
| mtreinish | clarkb: oh, I see your point we were probably installing pbr from source when I hit these issues in the gate before | 16:48 |
| dhellmann | bauzas: yeah, don't install pbr from source | 16:48 |
| *** andreaf is now known as andreaf_ | 16:49 | |
| mtreinish | but now it's the release so no more gate issues | 16:49 |
| clarkb | pleia2: any garage flooding concerns this time around? | 16:49 |
| bauzas | dhellmann: well, it was pulled from somewhere in my devstack I guess | 16:49 |
| bauzas | :) | 16:49 |
| pleia2 | clarkb: we'll see, turns out Comcast's cage in our building was leaking last time, they said they fixed it, but Comcast, so... | 16:49 |
| bauzas | lemme fix that | 16:49 |
| dhellmann | bauzas: yeah, but by default devstack doesn't do that any more so there must be a setting causing it | 16:49 |
| *** tnovacik has quit IRC | 16:50 | |
| openstackgerrit | garyk proposed openstack/requirements: Upgrade oslo.vmware>=0.8.0 https://review.openstack.org/141099 | 16:50 |
| dhellmann | mtreinish: I'm curious about why tempest sets the version in setup.cfg instead of via a tag? I know that's a supported mode, but I'm curious about why you use it in this case. | 16:50 |
| mordred | clarkb: turns out it's not cinder that's a problem in containers | 16:50 |
| mordred | clarkb: it's nova-compute | 16:50 |
| clarkb | mtreinish: I approved your first d-g change. I compared logs and it did what I wanted to see | 16:50 |
| clarkb | mordred: it is cinder... | 16:50 |
| mordred | clarkb: nope | 16:50 |
| clarkb | mordred: yes... | 16:50 |
| clarkb | mordred: cinder requires iscsi | 16:50 |
| mordred | clarkb: I'm talking to the guys from rax who are doing the opensatck-in-containers thing right now | 16:50 |
| mordred | cinder works fine | 16:50 |
| clarkb | mordred: yes and they are wrong | 16:50 |
| mordred | iscsiadm in nova-compute does not | 16:51 |
| jgriffith | clarkb: mordred haha | 16:51 |
| mordred | I'm pretty sure they're not wrong, since they're running it | 16:51 |
| clarkb | mordred: they describe the problem wrong | 16:51 |
| clarkb | mordred: we are basically saying the same thing but with different words | 16:51 |
| clarkb | mordred: it is not iscsiadm that is problematic | 16:51 |
| mordred | clarkb: nova-compute is the thing that has to run outside of a container | 16:51 |
| mordred | not cinder | 16:51 |
| clarkb | mordred: iscsid cannot run in a container. so when iscsiadm runs it fails | 16:51 |
| mtreinish | dhellmann: because tempest's tag numbering is just consecutive integers. The autoversioning from tags stuff doesn't undertsand that | 16:51 |
| clarkb | mordred: yes not cinder and not nova | 16:51 |
| clarkb | mordred: iscsid... | 16:51 |
| mordred | right | 16:51 |
| mordred | but in terms of services | 16:51 |
| clarkb | mordred: so anywhere you need an iscsid | 16:51 |
| mordred | the nova-computer "service" is the thing that needs to run not in a container | 16:52 |
| clarkb | regardless of nova or cinder, it will not work | 16:52 |
| clarkb | mordred: my biggest gripe is the blaming iscsiadm | 16:52 |
| mtreinish | dhellmann: it assumes the next thing will be a point release. I mean I guess we could drop the version string | 16:52 |
| clarkb | which is a symptom not a cause | 16:52 |
| bauzas | dhellmann: just FYI, installing latest pbr fixed the problem thanks | 16:52 |
| dhellmann | mtreinish: ah, so you don't want to tag 4.0.0 or whatever? | 16:53 |
| dhellmann | bauzas: great! | 16:53 |
| * bauzas now wonders how it came from source | 16:53 | |
| *** eharney has quit IRC | 16:53 | |
| mtreinish | dhellmann: yeah the next one will be just 4 | 16:53 |
| clarkb | bauzas: that was the old way of devstack isntalling. depending on how old your devstack is you may have transitioned across that gap without it doing the lib stuff properly | 16:53 |
| clarkb | mordred: we are bad at describing the problem | 16:54 |
| dhellmann | mtreinish: makes sense I guess | 16:54 |
| jgriffith | mordred: honestly I thought it was perfectly sane to leave Compute Nodes as physical/non-containerized | 16:54 |
| clarkb | mordred: we should all rollback to saying "iscsid cannot run in a container. Anything that needs an iscsid will then be unhappy" | 16:54 |
| mtreinish | dhellmann: if we did drop the version string all that would mean is the intermediate builds will show a 3.1 version string with the git hash (which really isn't a big deal I guess) | 16:54 |
| clarkb | mordred: instead of saying cinder/nova/iscsidadm | 16:54 |
| jgriffith | clarkb: mordred if you don't mind me asking... what is it that you're cooking up at the moment? | 16:55 |
| *** markmcclain has quit IRC | 16:55 | |
| mtreinish | dhellmann: because the tag would still be 4 so that should get created properly | 16:55 |
| clarkb | mordred: beacuse when we do that we are discussing symptooms of a bug and it doesn't help us debug anything | 16:55 |
| clarkb | jhesketh: pleia2 had a while back tried to get openstack running in containers so that we could do tripleo testing | 16:55 |
| clarkb | er jgriffith ^ | 16:55 |
| clarkb | jgriffith: and since then everyone else has discovered what pleia2 discovered. You can't do it if you need iscsid | 16:55 |
| mordred | jgriffith: mostly just curious in how they'd done containerized deploy | 16:56 |
| jgriffith | got ya | 16:56 |
| jgriffith | mordred: clarkb FWIW, I've been working on deploying everything BUT compute in containers | 16:56 |
| mordred | jgriffith: but yes, I also agree that non-containerized nova-compute seems perfectly fine | 16:56 |
| clarkb | mordred: and it is "cinder" as a client facing service that requires it | 16:56 |
| clarkb | mordred: because cinder requires you attach volumes via iscsi | 16:56 |
| jgriffith | mordred: clarkb didn't get as far as the RAX folks... so was pretty excited to see what they did | 16:56 |
| clarkb | mordred: from the computer perspective it is the nova process that needs it | 16:56 |
| mordred | clarkb: right. but as a client facing service you do not care about deployment | 16:56 |
| clarkb | mordred: which makes talking about this even harder | 16:56 |
| mordred | clarkb: so the question is, which bits as a deployer can not be run in a container | 16:57 |
| clarkb | mordred: but you do care because ou can sa "we won't offer cinder in this deployment" | 16:57 |
| clarkb | mordred: you have to grok both sides when making deploment decisions | 16:57 |
| mordred | or you just do what they're doing and run nova-compute bare | 16:57 |
| clarkb | yup I think running nova compute bare is a great way to do it | 16:57 |
| jgriffith | perhaps we can poke at the LXC bug a bit, but at this point I'm likely just getting in the way of your work and conversation so I'll pipe down :) | 16:57 |
| *** kmartin has joined #openstack-infra | 16:57 | |
| pleia2 | I keep hoping someone will actually hop on that bug and fix it, but it's now been year+ | 16:57 |
| mordred | jgriffith: I _hear_ that someone is working on it - but it's not an LXC bug as I understand it | 16:57 |
| *** harlowja_at_home has joined #openstack-infra | 16:58 | |
| pleia2 | maybe serge will end up on it come january :) | 16:58 |
| mordred | it's that iscsi doesn't grok network namespacing in the kernel | 16:58 |
| clarkb | mordred: its a kernel issue | 16:58 |
| mordred | yah | 16:58 |
| jgriffith | clarkb: +1 | 16:58 |
| clarkb | iscsi driver in the kernel needs to grow namespacing | 16:58 |
| *** shashankhegde has joined #openstack-infra | 16:58 | |
| *** KurtMartin has joined #openstack-infra | 16:58 | |
| *** ihrachyshka has joined #openstack-infra | 16:58 | |
| *** shakamunyi has joined #openstack-infra | 16:59 | |
| bauzas | clarkb: ok, thanks for the info, and indeed my devstack is pretty old even if I made many updates to all repos, incl. devstack itself | 16:59 |
| clarkb | mordred: however somtimes run nova compute bare isn't an option eg when testing tripleo or if your cloud gives you containers. so still good to understand this stuff | 16:59 |
| clarkb | and hopefully it will get fixed | 16:59 |
| clarkb | bauzas: ya typically devstack won't know how to upgrade itself iirc | 16:59 |
| mordred | clarkb: yes | 17:00 |
| clarkb | bauzas: so if it lays down an initial state for lib installs you end up stuck with that | 17:00 |
| clarkb | mordred: it is also confusing that people say "we totally deploy openstack in containers" then you have to hunt them down and ask them to clarify for the * | 17:00 |
| *** nikil2 has quit IRC | 17:01 | |
| clarkb | * not services XYZ which don't run in containers because pain | 17:01 |
| clarkb | iirc kolla is also having trouble with neutron but that is a docker specific issue? | 17:01 |
| mordred | I believe so, yes | 17:01 |
| clarkb | something about how docker assumes you onl ever need one interface or something | 17:01 |
| mordred | yup | 17:01 |
| *** sabeen1 has joined #openstack-infra | 17:02 | |
| mordred | and that docker is going to manage your networking for you | 17:02 |
| *** kmartin has quit IRC | 17:02 | |
| pleia2 | there were some other weird bits about running in containers, but the iscsi thing was the only show stopper | 17:02 |
| *** SumitNaiksatam has joined #openstack-infra | 17:02 | |
| pleia2 | mostly they were just openstack bugs though | 17:02 |
| clarkb | down to four stubborn nodes on old nodepool. all in the tripleo region | 17:02 |
| clarkb | pleia2: ya and I think those things have improved over the last year or so | 17:03 |
| *** penguinRaider__ has quit IRC | 17:03 | |
| *** penguinRaider has joined #openstack-infra | 17:03 | |
| * clarkb is grumpy that we have to go through this every couple months | 17:03 | |
| jeblair | clarkb: no more disconnects on zuul so far | 17:04 |
| clarkb | jeblair: that is good news. I should check for timeouts on new nodepool | 17:04 |
| *** dtantsur is now known as dtantsur|afk | 17:04 | |
| *** harlowja_at_home has quit IRC | 17:05 | |
| *** achanda has joined #openstack-infra | 17:05 | |
| clarkb | jeblair: looks like just two in total since the pycrpto changes | 17:05 |
| *** otter768 has joined #openstack-infra | 17:06 | |
| clarkb | jeblair: if this holds up I can update DNS and finish old nodepool cleanup (images, aliens) on monday | 17:06 |
| clarkb | I think monday is actually going to be busy with not work so maybe tuesday | 17:07 |
| *** dangers_away is now known as dangers | 17:07 | |
| clarkb | pleia2: you should just pick up kernel hacking and fix this :) | 17:07 |
| jeblair | noo! zanata! | 17:08 |
| jeblair | pleia2: please don't become a kernel hacker :) | 17:08 |
| *** mpaolino has quit IRC | 17:08 | |
| clarkb | ok of the two old git.o.o servers the older one is still showing a bit of traffic from IBM | 17:10 |
| clarkb | the newer one is being occasionally hit by OVH | 17:10 |
| *** otter768 has quit IRC | 17:10 | |
| jeblair | clarkb: oh have you checked the graphs yet? | 17:10 |
| *** reed has joined #openstack-infra | 17:10 | |
| pleia2 | haha, yeah, I'm not touching that :) | 17:10 |
| *** mpaolino has joined #openstack-infra | 17:11 | |
| *** gyee has joined #openstack-infra | 17:11 | |
| dims | jeblair: clarkb: gate-infra-puppet-apply-centos6 seems to be failing on a totally unrelated change, any clues? https://review.openstack.org/#/c/140933/ (i am just sorting the order of the oslo bot channels) | 17:11 |
| jeblair | clarkb, fungi: btw, let's leave git.o.o in cacti for a while to keep its history | 17:11 |
| clarkb | jeblair: I did check graphs they are happy making | 17:11 |
| clarkb | jeblair: we still get peaks but they are shorter on both axis | 17:11 |
| *** zz_sabari_ is now known as sabari_ | 17:12 | |
| mordred | wow. I literally have no idea how to do the same thing with nova-agent/cloud-init on centos7/systemd | 17:12 |
| jeblair | clarkb: yeah, look at that -- a 450mbps + 350mbps peak | 17:12 |
| jeblair | dims: hrm, i thought i saw some talk from fungi about the centos6 job failing yesterday... | 17:13 |
| *** luqas has quit IRC | 17:13 | |
| dims | jeblair: ah ok, will wait for fungi | 17:13 |
| mordred | because the upstart scripts just use upstart for ordering and call out to the init.d scripts ... but on centos7, the init.d scripts themselves defer to systemctl in some way that I really don't fully understand | 17:13 |
| jeblair | 2014-12-10T19:00:04 <fungi> clarkb: what's the possibility that the repeat gate-infra-puppet-apply-centos6 failures are an issue in module dependency ordering introduced when i approved adding puppet-httpd to the modules array yesterday? | 17:13 |
| jeblair | 2014-12-10T19:00:27 <clarkb> fungi: I think that must be what it is. I noticed that mrmartin's fix for the module ordering didn't seem affected | 17:13 |
| jeblair | dims: ^ ? | 17:13 |
| dims | jeblair: that sounds like the same thing i see. thanks | 17:14 |
| clarkb | I haven't looked into it further since I said that but I should probably go review mrmartins patch for real now | 17:14 |
| jeblair | clarkb: link if you need another eye; otherwise i'm on backlog duty | 17:14 |
| dims | thanks clarkb | 17:14 |
| mrmartin | clarkb, jeblair: my patch not a solution yet :) | 17:14 |
| clarkb | jeblair: ok | 17:14 |
| mrmartin | but the problem exists, and generates a lot of noise | 17:15 |
| clarkb | mrmartin: its a partial solution and fixes the immediate problem aiui | 17:15 |
| mrmartin | clarkb, it solves when you redeploy from scatch. becuase in this specific case, some downgrade required in puppet modules, and it is not working | 17:15 |
| clarkb | nibalizer: so I completely disagree with your statement | 17:16 |
| mrmartin | stdlibs 4.4.0 -> 4.3.2 | 17:16 |
| *** luqas has joined #openstack-infra | 17:16 | |
| clarkb | nibalizer: do not be unhappy if we merge it with your -1 :P the key thing being that with tools like pip and puppet module order 100% matters because of how they evaluate dependencies | 17:16 |
| clarkb | nibalizer: so it isn't ok to pretend that dependency resolution can be done properly | 17:16 |
| mrmartin | clarkb, but it requires a real double check, because module versions are upraded in that patch | 17:17 |
| mrmartin | upgraded. | 17:17 |
| clarkb | nibalizer: also its nice if things fail the same way everywhere rather than semi random failures like we get | 17:17 |
| *** bhunter71 has joined #openstack-infra | 17:18 | |
| clarkb | mrmartin: roger. Is it just libs like puppet-stdlib that get upgraded? | 17:18 |
| clarkb | mrmartin: I think that as long as we keep modules like postgres/mysql/apache/etc stable then their dependencies can change and that is ok | 17:18 |
| *** salv-orlando has quit IRC | 17:18 | |
| clarkb | mrmartin: oh I see that we manually upgrade some of the versions there too | 17:19 |
| mrmartin | no, https://review.openstack.org/#/c/140440/2/modules.env puppetlabs-puppetdb is not satisfying the dependencí requirements | 17:19 |
| *** salv-orlando has joined #openstack-infra | 17:19 | |
| clarkb | mrmartin: gotcha | 17:19 |
| mrmartin | and upgrading that brings puppetlabs-firewall and puppetlabs-inifile changes also. | 17:19 |
| clarkb | let me compare against our master to see what we have | 17:19 |
| *** nfedotov has quit IRC | 17:20 | |
| *** vdo has quit IRC | 17:20 | |
| mrmartin | but it cannot handle the downgrade, so without an rm -rf /etc/puppet/modules/* this won't work. | 17:21 |
| clarkb | mrmartin: we have inifile 1.1.3 so that change is fine but we do have firewall 0.0.4 and puppetdb 3.0.1 so those changes will affect us | 17:21 |
| clarkb | mrmartin: I think we can just upgrade right? maybe I misunderstand that | 17:21 |
| mrmartin | clarkb: no sir. the problem here, that we have newer version deployed on nodes, that we have defined in modules.env | 17:22 |
| clarkb | mrmartin: because they are transitive dependencies, but do those newer version work? if so I think what we can do is use those newer versions | 17:22 |
| mrmartin | and if you try in an empty box, with a puppet module install / puppet module upgrade, the upgrade part will also fail, because it not supports the downgrade. | 17:22 |
| clarkb | mrmartin: that is what our tests do and they pass on your change | 17:23 |
| mrmartin | yeap, in an ideal world we could use the newer versions, but they have different version requirements, and it is very easy to get a broken dependency tree | 17:23 |
| jeblair | mordred: did you actually review 122311? | 17:23 |
| *** salv-orlando has quit IRC | 17:24 | |
| mrmartin | check this: http://paste.openstack.org/show/148078/ | 17:24 |
| clarkb | mrmartin: so maybe we should backup here. Why are we updating puppetdb? | 17:24 |
| mordred | jeblair: 122311 does not show up for me | 17:24 |
| clarkb | mrmartin: because of postgres right? | 17:24 |
| jeblair | mordred: i don't know what you mean by that | 17:25 |
| mrmartin | clarkb: exactly, because puppetdb require < 4.0.0 of pgsql | 17:25 |
| jeblair | mordred: it's the mysql-proxy change we've been talking about for 2 days | 17:25 |
| mrmartin | but if you upgrade the puppetdb, you must do the same with firewall and inifile | 17:25 |
| mordred | jeblair: sorry - browser fat-finger | 17:25 |
| mrmartin | and it can broke things. | 17:25 |
| clarkb | mrmartin: gotcha | 17:25 |
| mordred | jeblair: yes, I did - one sec, let me go vote | 17:26 |
| *** Masahiro has joined #openstack-infra | 17:26 | |
| mordred | done | 17:26 |
| jeblair | mordred: thx | 17:26 |
| clarkb | mrmartin: lets assume that updating puppetdb, inifile, and firewall will not break anything because of their puppet and only consider version requirements | 17:26 |
| clarkb | mrmartin: updating those three modules as you did in your change makes the module instalation work? | 17:26 |
| mrmartin | than in that case, this patch will work on new nodes only. because existing deployments have newer modules, which requires a downgrade | 17:27 |
| clarkb | mrmartin: if that is the case I can work on making sure those module updates won't break our puppetdb installation, then we apply your patch as fixing us at a point in time | 17:27 |
| clarkb | mrmartin: which packages require a downgrade? | 17:27 |
| mrmartin | stdlib | 17:27 |
| mrmartin | check this log carefully: http://logs.openstack.org/11/138011/2/check/gate-infra-puppet-apply-precise/ce13ed2/console.html | 17:28 |
| mrmartin | Error: Could not install module 'puppetlabs-stdlib' (v4.3.2) / Module 'puppetlabs-stdlib' (v4.4.0) is already installed | 17:28 |
| *** hashar_ has quit IRC | 17:29 | |
| mrmartin | clarkb, and if you check the install_modules.sh you can see, if module upgrade failed, it tries to install a new one, but fails, because 4.4.0 already deployed. | 17:29 |
| clarkb | mrmartin: but only because we are saying we want 4.3.2 | 17:29 |
| mrmartin | yes | 17:29 |
| clarkb | ok that one is probalby the safest one to update to 4.4.0 | 17:30 |
| mrmartin | https://github.com/openstack-infra/system-config/blob/master/install_modules.sh#L80 | 17:30 |
| clarkb | since they work hard to make stdlib compat | 17:30 |
| jeblair | mtreinish: 122311 aprvd | 17:30 |
| mrmartin | yeap it can be a solution, but need to check the dependency tree again | 17:30 |
| mrmartin | if you want, I can check it now | 17:30 |
| *** _shaps__ has joined #openstack-infra | 17:30 | |
| *** hdd has joined #openstack-infra | 17:31 | |
| *** Masahiro has quit IRC | 17:31 | |
| clarkb | mrmartin: ya lets bump stdlib to 4.4.0 explicitly. It looks like postgresql is another one that gets installed at the wrong version but that may be related to how we build those test slaves \o/ | 17:32 |
| *** _nadya_ has quit IRC | 17:32 | |
| *** ala_ has quit IRC | 17:32 | |
| *** _shaps_ has quit IRC | 17:33 | |
| clarkb | mrmartin: ya lets bump stdlib to 4.4.0 explicitly. It looks like postgresql is another one that gets installed at the wrong version but that may be related to how we build those test slaves \o/ | 17:33 |
| clarkb | er | 17:33 |
| *** alexpilotti has joined #openstack-infra | 17:33 | |
| clarkb | mrmartin: also we should rm -rf /etc/puppet/modules as the first step in that job if we don't already | 17:34 |
| clarkb | mrmartin: to prevent leakage from our image builds | 17:34 |
| *** alexpilotti has quit IRC | 17:34 | |
| clarkb | mrmartin: do you want to just add that to your change or put that under your change as a dependency? | 17:34 |
| *** alexpilotti has joined #openstack-infra | 17:34 | |
| *** johnthetubaguy is now known as zz_johnthetubagu | 17:35 | |
| mrmartin | clarkb: give me a sec, I want to test first with the newer stdlib module | 17:35 |
| clarkb | mrmartin: sure | 17:35 |
| mrmartin | nibalizer, what was the reason you nut suggested the rm -rf /etc/puppet/modules in install_modules.sh ? | 17:35 |
| mrmartin | nut / not | 17:35 |
| clarkb | mrmartin: you don't want it in install_modules.sh | 17:35 |
| clarkb | mrmartin: we should put it in the test script that runs that test | 17:36 |
| mrmartin | oh right | 17:36 |
| zaro | morning | 17:36 |
| clarkb | mrmartin: tools/apply-test.sh | 17:36 |
| *** yamamoto_ has quit IRC | 17:36 | |
| *** penguinRaider__ has joined #openstack-infra | 17:36 | |
| *** pc_m has quit IRC | 17:36 | |
| clarkb | I think not rm'ing in the test is what allowed us to wedge ourselves too | 17:37 |
| *** baoli has quit IRC | 17:37 | |
| clarkb | the initial run worked because it had older modules in /etc/puppet/modules then we update /etc/puppetmodules and updated images and now the tests don't pass | 17:37 |
| clarkb | fungi: ^ for when you are not afking | 17:37 |
| clarkb | ok time for morning things back in a bit | 17:38 |
| *** vigneshvar has joined #openstack-infra | 17:38 | |
| *** tonytan4ever has quit IRC | 17:39 | |
| *** rmcall has joined #openstack-infra | 17:39 | |
| *** penguinRaider has quit IRC | 17:40 | |
| *** Hal_ has quit IRC | 17:42 | |
| *** tnovacik has joined #openstack-infra | 17:43 | |
| *** baoli has joined #openstack-infra | 17:44 | |
| mrmartin | clarkb: paste.openstack.org throws an internal error when I'm trying to paste the output, but upgrade stdlibs to 4.4.0 won't broke the dependency | 17:44 |
| jeblair | mrmartin: if you try paste again it might work | 17:44 |
| *** eharney has joined #openstack-infra | 17:45 | |
| mrmartin | http://paste.openstack.org/show/149481/ | 17:45 |
| *** ayoung is now known as ayoung-afk | 17:46 | |
| *** amitgandhinz has quit IRC | 17:47 | |
| openstackgerrit | Marton Kiss proposed openstack-infra/system-config: Fix puppet module dependency order https://review.openstack.org/140440 | 17:48 |
| *** palar has joined #openstack-infra | 17:48 | |
| *** mika has quit IRC | 17:48 | |
| jeblair | clarkb: 128707 still relevant? want to aprv? | 17:49 |
| *** amitgandhinz has joined #openstack-infra | 17:49 | |
| *** mika has joined #openstack-infra | 17:49 | |
| *** bauzas has left #openstack-infra | 17:50 | |
| *** shayneburgess has joined #openstack-infra | 17:50 | |
| *** shashankhegde has quit IRC | 17:51 | |
| *** mpaolino has quit IRC | 17:52 | |
| clarkb | mrmartin: I think that templatedir warning is unrelated | 17:54 |
| *** emagana has quit IRC | 17:54 | |
| fungi | okay, i am now here | 17:54 |
| mrmartin | yeap, it is just a packaging bug in ubuntu | 17:54 |
| fungi | i may just need to declare bankruptcy on scrollback | 17:54 |
| *** e0ne has quit IRC | 17:54 | |
| fungi | 500+ lines since last night. my time may be better spent jumping right into whatever's broken | 17:55 |
| clarkb | jeblair: supposedly they are working on fixing the leak or have fixed it but I think being conservative on max passenger requests is a good thing. I will approve | 17:55 |
| *** asselin has joined #openstack-infra | 17:55 | |
| jeblair | fungi: check with mrmartin and clarkb on centos6 + puppet issue | 17:55 |
| *** achanda has quit IRC | 17:55 | |
| *** gyee has quit IRC | 17:55 | |
| fungi | jeblair: thanks. skimming now | 17:56 |
| *** achanda has joined #openstack-infra | 17:56 | |
| clarkb | and approved | 17:56 |
| *** bknudson has quit IRC | 17:56 | |
| mtreinish | jeblair: awesome thanks | 17:56 |
| *** ivar-lazzaro has joined #openstack-infra | 17:56 | |
| clarkb | I cross checked with the vhost on the master and those settings are still in place | 17:56 |
| mtreinish | I'm assuming the centos6 apply tests are still having trouble | 17:56 |
| clarkb | mrmartin: so it looks like 4.4.0 does make that better | 17:57 |
| fungi | the centos6 failures i observed are definitely nondeterministic, since rechecks sometimes caused them to succeed | 17:57 |
| *** sabari_ is now known as zz_sabari_ | 17:57 | |
| mtreinish | fungi: ok, then I'll try a recheck :) | 17:57 |
| clarkb | fungi: yup I think its a combo of not removing old /etc/puppet/modules and install order | 17:57 |
| fungi | ahh | 17:57 |
| clarkb | /etc/puppet/modules not being cleaned is what allowed us to wedge | 17:57 |
| clarkb | or made it easier to wedge | 17:57 |
| *** r-daneel has joined #openstack-infra | 17:57 | |
| mtreinish | fungi, clarkb, jeblair: if you guys get a sec can you create the "public" db user with limited priv. on the subunit2sql db? | 17:58 |
| fungi | mtreinish: i can do that now | 17:58 |
| *** zz_sabari_ is now known as sabari_ | 17:59 | |
| clarkb | I think the next steps are to look at stdlib release nodes to make sure 4.4.0 isn't crazy (it shouldn't be stdlib is good about that) then do our best to make sure puppetdb will continue to work with the new versions of firewall and puppetdb. inifile is already 1.1.3 on our master | 17:59 |
| fungi | mtreinish: did that change merge? | 17:59 |
| mtreinish | fungi: that's what bounced off the centos apply job | 17:59 |
| mtreinish | I just rechecked it | 18:00 |
| fungi | heh | 18:00 |
| *** mmaglana has joined #openstack-infra | 18:00 | |
| clarkb | hrm rm -rf /etc/puppet/modules makes it fail in new and exciting ways | 18:00 |
| *** jistr has quit IRC | 18:00 | |
| clarkb | oh I had a crazy idea just now. What if we made a puppet env for each certname | 18:00 |
| *** achanda has quit IRC | 18:00 | |
| clarkb | then ansible would run puppet agent --test --environment $hostname | 18:00 |
| clarkb | then we could have different puppet things for all the things and change their deps without needing to worry about the whole system at once | 18:01 |
| jeblair | nibalizer: replied on 129748 | 18:01 |
| jeblair | by popular demand I'm going to self-approve that | 18:01 |
| openstackgerrit | Merged openstack-infra/devstack-gate: Add neutron advanced services to PROJECTS list https://review.openstack.org/141077 | 18:02 |
| *** harlowja_away is now known as harlowja | 18:03 | |
| clarkb | also why does this only fail on centos6 | 18:03 |
| fungi | jeblair: thanks--it was popular with me anyway | 18:03 |
| clarkb | precise is still working as a job right? | 18:03 |
| *** jlibosva has quit IRC | 18:03 | |
| *** thedodd has quit IRC | 18:04 | |
| *** luqas has quit IRC | 18:04 | |
| *** kmartin has joined #openstack-infra | 18:04 | |
| *** penguinRaider has joined #openstack-infra | 18:04 | |
| *** bswartz has quit IRC | 18:04 | |
| *** penguinRaider__ has quit IRC | 18:04 | |
| *** MaxV has quit IRC | 18:06 | |
| *** BobBall is now known as BobBall_AWOL | 18:06 | |
| mtreinish | clarkb: that's what it looks like | 18:06 |
| * mrmartin going to eat something... | 18:06 | |
| *** derekh has quit IRC | 18:07 | |
| *** kmartin has quit IRC | 18:08 | |
| *** KurtMartin has quit IRC | 18:08 | |
| *** tonytan4ever has joined #openstack-infra | 18:09 | |
| *** luqas has joined #openstack-infra | 18:09 | |
| mtreinish | clarkb: err, looks like recheck failed on centos apply again. Should I just hold of for now? | 18:09 |
| mtreinish | s/of/off | 18:10 |
| *** bswartz has joined #openstack-infra | 18:10 | |
| clarkb | mtreinish: ya I think we need to solve this dep thing first | 18:10 |
| mtreinish | ok, then I'll take that as a cue to go find food :) | 18:11 |
| *** _nadya_ has joined #openstack-infra | 18:12 | |
| mordred | clarkb: amazingly enough, I kinda like your env-per-certname idea | 18:12 |
| clarkb | mordred: I think it will need more thought but I am sort of beginning to like it too | 18:12 |
| clarkb | mordred: it should be completely automatable | 18:12 |
| *** arosen has joined #openstack-infra | 18:12 | |
| fungi | input from crinkle and nibalizer would be good too | 18:13 |
| *** shashankhegde has joined #openstack-infra | 18:13 | |
| *** dangers is now known as dangers_away | 18:13 | |
| mordred | clarkb: yah. I kinda don't like that we're dealing with a dependency hell in the system taht's supposed to make dealing with dependencies better | 18:14 |
| clarkb | fungi: any idea why mrmartins latest patchset fails so hard? mrmartin did add the rm -rf like I suggested and I think that caused it but I don't understand why | 18:14 |
| *** achanda has joined #openstack-infra | 18:14 | |
| fungi | i was just pulling that back up to see how it was implemented | 18:14 |
| clarkb | its almost like puppet module install thinks the package is there when it isn't | 18:14 |
| *** avozza is now known as zz_avozza | 18:14 | |
| *** _nadya_ has quit IRC | 18:15 | |
| * crinkle scrolls back | 18:15 | |
| *** pelix has quit IRC | 18:16 | |
| fungi | crinkle: we're debating solutions to the problem described in https://review.openstack.org/140440 | 18:16 |
| clarkb | precise fails differently than centos6 too | 18:16 |
| fungi | this is slightly maddening | 18:16 |
| *** pelix has joined #openstack-infra | 18:16 | |
| *** tonytan4ever has quit IRC | 18:17 | |
| fungi | luckily i failed my sanity check a long, long time ago | 18:17 |
| fungi | Error: Could not find class pip | 18:18 |
| fungi | in the precise test | 18:18 |
| *** dimtruck is now known as zz_dimtruck | 18:18 | |
| clarkb | fungi: ya that may be a legit fail due to the new module state after we start from scratch | 18:18 |
| fungi | i think it wants the modules to be there and osn't finding them now | 18:18 |
| clarkb | fungi: I think the centos6 failure is the funky one | 18:18 |
| *** penguinRaider__ has joined #openstack-infra | 18:19 | |
| fungi | also, we should probably legitimately have a trusty apply test at this stage | 18:19 |
| clarkb | ++ | 18:19 |
| fungi | especially with nodepool running on it in production as of yesterday | 18:19 |
| fungi | the centos6 error looks the same to me, just a different module name (presumably the ordering is platform dependent?) | 18:20 |
| *** Ryan_Lane has joined #openstack-infra | 18:21 | |
| clarkb | sr | 18:21 |
| *** gyee has joined #openstack-infra | 18:22 | |
| clarkb | derp | 18:22 |
| *** pelix has quit IRC | 18:22 | |
| crinkle | so if you're using `puppet module install` and the metadata has conflicting dependencies, you're going to get install conflicts no matter what order you go in | 18:22 |
| clarkb | crinkle: thats not true | 18:22 |
| *** penguinRaider has quit IRC | 18:22 | |
| crinkle | yes it is | 18:22 |
| crinkle | if you don't want to deal with dependencies, install from git instead of the forge | 18:22 |
| clarkb | not if the conflicts are special | 18:22 |
| clarkb | you can have conflicts that arrise out of order. pip has the same problem | 18:23 |
| clarkb | the greater issue is deps are not globally evaluated | 18:23 |
| clarkb | typically this is true if an older version would satisfy two project's deps, but you install one that allows a nwere version first | 18:23 |
| fungi | can the modules be installed in a single call instead if iteratively one-by-one? | 18:24 |
| clarkb | that installs the newer version of dep, then second package install says I can't wrk with that and you can't downgrade -> fail | 18:24 |
| crinkle | fungi: I think nibalizer is working on a tool to do that | 18:24 |
| fungi | er, instead of | 18:24 |
| crinkle | it's not really possible right now without making a sort of monolithic module with all the other modules listed | 18:25 |
| fungi | ahh, a metamodule | 18:26 |
| crinkle | my honest recommentation would be to install from git using shell or r10k | 18:26 |
| crinkle | recommendation* | 18:26 |
| clarkb | if A depends on C>1.0,<2.0 and B depends on C>1.0 and C has version 1.0, 1.1, and 2.0 available order matters | 18:26 |
| crinkle | okay fair enough | 18:27 |
| *** doug-fish has left #openstack-infra | 18:28 | |
| *** vipul has quit IRC | 18:28 | |
| *** hyakuhei has quit IRC | 18:29 | |
| reed | I can't find explicit documentation to create a project on stackforge. Is this the best documentation we have for it? http://docs.openstack.org/infra/manual/creators.html?highlight=stackforge | 18:29 |
| *** achanda has quit IRC | 18:30 | |
| *** markmcclain has joined #openstack-infra | 18:30 | |
| fungi | reed: yes, if there are things missing we should improve that, but it's meant to cover creating any kind of project (stackforge, openstack, whatever) | 18:30 |
| *** achanda has joined #openstack-infra | 18:30 | |
| annegentle | reed: yep, that's the best I know of | 18:30 |
| openstackgerrit | Kyle Rockman proposed openstack-infra/jenkins-job-builder: hipchat bug fix https://review.openstack.org/133997 | 18:31 |
| jeblair | reed: the old docs were here http://ci.openstack.org/stackforge.html now point to your link | 18:31 |
| fungi | reed: previously we had one which was focused exclusively on stackforge, and then pointed anyone creating official projects to that and warned them about all sorts of caveats | 18:31 |
| *** HeOS has quit IRC | 18:31 | |
| fungi | so hopefully this new document is an improvement all around | 18:31 |
| *** jcoufal has joined #openstack-infra | 18:32 | |
| reed | discoverability of these documents all scattered on minisites is awful, btw | 18:32 |
| reed | thanks guys... | 18:32 |
| fungi | reed: hopefully docs.openstack.org is not as much of a minisite as ci.openstack.org | 18:32 |
| reed | i searched for stackforge on the wiki, got no useful result ... went to docs.o.o got no useful result either | 18:32 |
| * crinkle -> no laptop zone | 18:32 | |
| fungi | crinkle: good luck, take a torch | 18:33 |
| *** achanda has quit IRC | 18:33 | |
| *** achanda has joined #openstack-infra | 18:33 | |
| reed | fungi, hopefully there will be a unified search on docs.o.o | 18:33 |
| fungi | clarkb: so that latest patchset both bumped stdlib and cleared the modules directory. i'm guessing the job failures are due to the latter | 18:34 |
| mrmartin | clarkb, fungi: have you seen the logs of puppet module dependency patch: Error: Could not find class pip for bare-precise-1418061334.template.openstack.org on node bare-precise-1418061334.template.openstack.org ? | 18:34 |
| reed | I received a question about the CLA in stackforge, don't see that topic covered in the manual | 18:34 |
| fungi | mrmartin: yes, that's the one i'm looking at right now | 18:34 |
| reed | diverted the question to legal-discuss, will file a bug or fix the docs once the discussion is over | 18:35 |
| mrmartin | fungi: basically this must work, two thing changed only the rm -rf puppet modules and stdlib version number | 18:35 |
| *** tonytan4ever has joined #openstack-infra | 18:35 | |
| openstackgerrit | Kyle Rockman proposed openstack-infra/jenkins-job-builder: hipchat bug fix https://review.openstack.org/133997 | 18:35 |
| *** jgallard_ has quit IRC | 18:35 | |
| *** vipul has joined #openstack-infra | 18:35 | |
| *** tkelsey has quit IRC | 18:36 | |
| annegentle | reed: ah, I bet the current CSE doesn't include /developer | 18:36 |
| annegentle | reed: it was implemented before /developer existed | 18:36 |
| openstackgerrit | Doug Wiegley proposed openstack-infra/project-config: Add neutron jobs to service repos, until they get their own tests https://review.openstack.org/141122 | 18:36 |
| *** yamamoto has joined #openstack-infra | 18:37 | |
| *** atiwari has quit IRC | 18:37 | |
| *** penguinRaider__ has quit IRC | 18:38 | |
| fungi | reed: awesome--thanks! stackforge+icla is a muddy topic. some projects working their way toward official status seem to want to start with icla enforcement to avoid annoying contact problems getting approval from original contributors later, others add it because they're cargo cultists | 18:38 |
| *** penguinRaider has joined #openstack-infra | 18:38 | |
| *** hyakuhei has joined #openstack-infra | 18:38 | |
| fungi | but we've never recommended/required it for them | 18:39 |
| *** doude has quit IRC | 18:39 | |
| *** markmcclain has quit IRC | 18:40 | |
| *** pblaho has joined #openstack-infra | 18:41 | |
| * nibalizer erads scrollback | 18:42 | |
| *** yamamoto has quit IRC | 18:42 | |
| *** palar has left #openstack-infra | 18:42 | |
| *** yjiang5 is now known as yjiang5_away | 18:43 | |
| fungi | so much erading | 18:44 |
| nibalizer | fungi: i created this https://github.com/nibalizer/puppet-bundler last night | 18:44 |
| nibalizer | to solve the problem that puppet doesn't have a 'pip install -r requirements.txt' for whatever reason | 18:44 |
| fungi | nibalizer: you're quick. i guess that's what clarkb was talking about | 18:45 |
| nibalizer | i hope we get one in the official puppet module tool but for now we have a derpy gem | 18:45 |
| *** sarob has joined #openstack-infra | 18:45 | |
| jeblair | gem install --derpy puppet-bundler | 18:45 |
| openstackgerrit | Kyle Rockman proposed openstack-infra/jenkins-job-builder: hipchat bug fix https://review.openstack.org/133997 | 18:45 |
| *** achanda has quit IRC | 18:45 | |
| *** e0ne has joined #openstack-infra | 18:46 | |
| *** achanda has joined #openstack-infra | 18:46 | |
| *** zz_dimtruck is now known as dimtruck | 18:46 | |
| *** ZZelle has joined #openstack-infra | 18:46 | |
| fungi | i guess having it in ruby is understandable as that's the turing-complete language of choice in puppetdom | 18:47 |
| Mithrandir | Dz0gbmdW | 18:48 |
| Mithrandir | gah | 18:48 |
| greghaynes | That looks ominous ;) | 18:49 |
| Mithrandir | well, changed. :-P | 18:49 |
| *** arosen has quit IRC | 18:49 | |
| *** e0ne has quit IRC | 18:50 | |
| fungi | pwgen -s 8 1 | 18:51 |
| *** achanda has quit IRC | 18:51 | |
| *** e0ne has joined #openstack-infra | 18:51 | |
| jeblair | nibalizer: can you look at 134835? | 18:51 |
| Mithrandir | nah, an old one that came from pwgen. New ones come from word lists. | 18:51 |
| fungi | all teh better | 18:51 |
| mrmartin | nibalizer: can this bundler handle to upgrade / downgrade too? | 18:52 |
| *** doug-fish has joined #openstack-infra | 18:52 | |
| *** luqas has quit IRC | 18:53 | |
| *** luqas has joined #openstack-infra | 18:53 | |
| fungi | nibalizer: do you have a proposed patch yet to try gem installing that in the install_modules.sh and then using it there? | 18:53 |
| fungi | curious to compare the log if so | 18:53 |
| *** bknudson has joined #openstack-infra | 18:55 | |
| nibalizer | sure | 18:55 |
| *** koolhead17 has quit IRC | 18:56 | |
| *** achanda has joined #openstack-infra | 18:56 | |
| *** luqas has quit IRC | 18:57 | |
| *** luqas__ has joined #openstack-infra | 18:57 | |
| *** penguinRaider__ has joined #openstack-infra | 18:59 | |
| fungi | hrm, need to use the mysql cli to do specific user grants in trove | 19:00 |
| fungi | or if troveclient supports it, rackspace's current api doesn't | 19:00 |
| *** harlowja has quit IRC | 19:00 | |
| *** doude has joined #openstack-infra | 19:00 | |
| *** patrickeast has joined #openstack-infra | 19:01 | |
| mordred | I'd prefer mysql cli to duplicating all of the grant syntax in trove | 19:01 |
| *** sabari_ is now known as zz_sabari_ | 19:01 | |
| fungi | agreed | 19:01 |
| clarkb | I think that is correct. troves management of mysql its is minimal | 19:01 |
| SlickNik | mordred: ++ | 19:01 |
| SlickNik | It is, and it's an extension that's not part of the core API | 19:01 |
| SlickNik | so deployers can choose to deploy it or not | 19:02 |
| clarkb | also on abus right now. apparently if you want pho you must arrive early | 19:02 |
| fungi | it's more that rackspace's webui allows you to create full-access accounts but not more restricted ones, and just points you to mysql examples for creating more restricted accounts | 19:02 |
| fungi | which works fine for me | 19:02 |
| *** markmcclain has joined #openstack-infra | 19:02 | |
| *** thedodd has joined #openstack-infra | 19:02 | |
| *** penguinRaider has quit IRC | 19:02 | |
| fungi | just have to trove root-enable the instance first | 19:02 |
| *** zz_sabari_ is now known as sabari_ | 19:03 | |
| nibalizer | jeblair: responded | 19:04 |
| nibalizer | fungi: my intention was never to use puppet bundler as a deployment tool | 19:04 |
| nibalizer | it was to help developers quickly grab the dependencies they need to hack on a module | 19:04 |
| fungi | nibalizer: ahh, fair enough | 19:04 |
| nibalizer | not that we can't try to pivot there | 19:05 |
| clarkb | what if we just install all from git? | 19:05 |
| nibalizer | if we're set on declaring 'let the puppet module tool handle dependencies' a failed experiment, can we just install everything from git | 19:05 |
| clarkb | crinkle suggested that | 19:05 |
| nibalizer | clarkb: ++ yes lets do taht | 19:05 |
| fungi | nibalizer: so something else to create a meta-module with the needed deps is still another possible solution i guess? | 19:05 |
| nibalizer | lets write down, in a file, the modules and versions we want | 19:05 |
| clarkb | someone should push a test patch for that | 19:05 |
| nibalizer | yea, something advantageous about using a high level tool is being able to say 4.x so you get latest 4 | 19:06 |
| nibalizer | updating a bunch of hashes in a file would get tedious | 19:06 |
| clarkb | tags | 19:06 |
| clarkb | its not that different from what we do now | 19:06 |
| nibalizer | clarkb: i think youll find most community puppet modules don't tag regularly | 19:06 |
| clarkb | but we install from tag points via forge | 19:07 |
| fungi | grr... ERROR: publicURL endpoint for database service not found | 19:07 |
| *** otter768 has joined #openstack-infra | 19:07 | |
| clarkb | so thats no worse righy? | 19:07 |
| * fungi shakes fist at rackspace's keystone-light | 19:07 | |
| clarkb | fungi :) | 19:07 |
| * morganfainberg faceplams | 19:07 | |
| morganfainberg | I *really* hope that get that addressed in the move to keystone... | 19:07 |
| morganfainberg | real keystone | 19:07 |
| morganfainberg | that kind of stuff* | 19:08 |
| jeblair | nibalizer: what are you doing with the gerrit api? | 19:08 |
| fungi | morganfainberg: like putting api endpoints for your services in the catalog? | 19:08 |
| morganfainberg | lol right? | 19:08 |
| morganfainberg | i think we require that kind of stuff or more closely require it | 19:08 |
| * fungi looks up hackish envvar workaround | 19:08 | |
| morganfainberg | in the real implementation | 19:08 |
| clarkb | fungi it may be like swift and use different tenant id | 19:09 |
| nibalizer | jeblair: i just hacked a cute little 'gerrit wip' function into my bashrc | 19:09 |
| clarkb | do a full listing of catalog and check | 19:09 |
| fungi | clarkb: entirely possible. i'm hoping it's just the url though | 19:09 |
| jeblair | nibalizer: ah cool | 19:09 |
| nibalizer | so i can wip things right after i git-review them, without leaving the command line | 19:09 |
| nibalizer | and all other gerrit subcommands work from 'gerrit *' on my command line, /me is nerd | 19:10 |
| fungi | clarkb: yep, endpoint is there so tenantid must be different? | 19:10 |
| * fungi boggles | 19:11 | |
| clarkb | thats the problem with swift iylt shouldshow you the value of tenantid too | 19:11 |
| fungi | clarkb: oh, actually the service type is rax:database rather than database | 19:12 |
| clarkb | lol | 19:12 |
| *** otter768 has quit IRC | 19:12 | |
| fungi | that was apparently only part of the problem though. gives me an empty list | 19:12 |
| fungi | aha, no that second issue was between my chair and keyboard | 19:13 |
| *** Ryan_Lane has quit IRC | 19:13 | |
| jeblair | fungi: are groups ssl cert data in hiera now? | 19:14 |
| *** amuller has quit IRC | 19:14 | |
| *** bswartz has quit IRC | 19:14 | |
| clarkb | I am on a bridge with two trains going over a river one atop the other | 19:14 |
| clarkb | (I thought that was cool) | 19:14 |
| mrmartin | clarkb: where are u? | 19:14 |
| fungi | jeblair: yes, i bought and added them ti hiera a couple weeks ago | 19:14 |
| fungi | s/ti/to/ | 19:14 |
| clarkb | mrmartin portland oregon | 19:15 |
| *** Masahiro has joined #openstack-infra | 19:15 | |
| jeblair | i am in a chair in my apartment and it is not moving. for the moment. never know when it'll start moving around here. | 19:15 |
| mrmartin | fungi: have you added the chain variable too? | 19:15 |
| fungi | mrmartin: yes, i added it at the same time | 19:15 |
| jeblair | fungi: safe to aprv? | 19:15 |
| mrmartin | clarkb: yep, they have a nice bridge-building industry | 19:15 |
| clarkb | should be it works on -dev | 19:15 |
| fungi | jeblair: absolutely | 19:15 |
| *** bswartz has joined #openstack-infra | 19:15 | |
| mrmartin | jeblair: approve that, worst case we'll fix it | 19:16 |
| clarkb | sorry I didnt get around to groups proper | 19:16 |
| jeblair | (i'm assuming well do a query for V:-2 once the centos6 thing is fixed) | 19:16 |
| *** arosen has joined #openstack-infra | 19:16 | |
| *** SotK has quit IRC | 19:16 | |
| fungi | advisable, yes | 19:16 |
| jeblair | since i'm approving all kinds of changes that are just bouncing off zuul | 19:16 |
| fungi | gerritball | 19:16 |
| clarkb | so I think we do just give git for all modules a go | 19:17 |
| fungi | clarkb: yeah, so apparently the only special trick with rackspace and troveclient is that they decided to use a nonstandard service name in the catalog | 19:17 |
| clarkb | after lunch I can get installed versions and translate to git | 19:17 |
| arosen | Hi, I'm facing this transient failure a lot of times in my 3rd party ci | 19:17 |
| arosen | [ERROR] /opt/stack/new/devstack/lib/glance:328 g-api did not start | 19:17 |
| arosen | http://logs2.aaronorosen.com/82/140782/1/check/dsvm-tempest-full-congress-nodepool/ed3b157/logs/devstacklog.txt.gz#_2014-12-11_08_14_46_987 | 19:17 |
| fungi | clarkb: and that's just a matter of figuring out the tagnames for the versions we're using, or...? | 19:18 |
| clarkb | checking metadata.json of each module for the actual version | 19:18 |
| arosen | I'm try to search in logstash.openstack.org to see if this occurs with the upstream gate. Should be this the right search string to enter? | 19:18 |
| clarkb | and checking repos for those tags | 19:18 |
| arosen | message: "/opt/stack/new/devstack/lib/glance:328 g-api did not start" | 19:18 |
| clarkb | if tag is missing it gets hard | 19:18 |
| clarkb | and git log -p will be used | 19:19 |
| *** Masahiro has quit IRC | 19:19 | |
| arosen | i was also wondering if anyone else is experiencing this. Or if there is a work around. It looks like devstack-gate doesn't have a way to pass SERVICE_TIMEOUT to devstack to increase the timeout for checking if a service is alive. | 19:19 |
| openstackgerrit | James Carey proposed openstack-dev/hacking: Add optional hacking checks https://review.openstack.org/134052 | 19:20 |
| *** spzala has quit IRC | 19:20 | |
| fungi | arosen: it's possible the fine folks in #openstack-qa have more of a handle on the current state of glance problems in devstack | 19:21 |
| *** ayoung-afk is now known as ayoung | 19:21 | |
| arosen | fungi: out of curiosity is there a way to set this: https://github.com/openstack-dev/devstack/blob/master/lib/glance#L327 SERVICE_TIMEOUT through devstack-gate? | 19:22 |
| arosen | Looks like I would need to extend devstack-gate to know about that variable | 19:22 |
| *** ivar-laz_ has joined #openstack-infra | 19:22 | |
| *** penguinRaider has joined #openstack-infra | 19:22 | |
| *** penguinRaider__ has quit IRC | 19:22 | |
| *** ivar-laz_ has quit IRC | 19:23 | |
| *** ivar-laz_ has joined #openstack-infra | 19:23 | |
| *** xyang1 has quit IRC | 19:25 | |
| *** thedodd has quit IRC | 19:25 | |
| *** ivar-lazzaro has quit IRC | 19:26 | |
| fungi | arosen: yeah, i don't see that as a pass-through in devstack-gate | 19:28 |
| fungi | arosen: unless you can tack it onto a stub localrc and tell devstack-gate to append to the existing localrc rather than replacing it | 19:28 |
| mordred | fungi: https://github.com/stackforge/os-client-config/blob/master/os_client_config/vendors.py#L23 | 19:29 |
| mordred | fungi: that's the rax trove service name | 19:29 |
| fungi | mordred: yeah, i got that. i take it rax trove is as opposed to normal trove? | 19:30 |
| *** xyang1 has joined #openstack-infra | 19:30 | |
| mordred | fungi: normal troveclient works | 19:30 |
| mordred | fungi: but they had it deployed before trove was drove | 19:30 |
| mordred | trove | 19:30 |
| fungi | mordred: yep, just not discoverable | 19:30 |
| mordred | yah | 19:30 |
| fungi | i suspect "vendors.py" is an abbreviation for "embarrassing-workarounds.py" | 19:31 |
| arosen | fungi: is there already a way to do that in devstack-vm-gate.sh? | 19:31 |
| clarkb | they could add a second entry | 19:31 |
| mordred | "things you have to know that you can't discover" | 19:31 |
| clarkb | and be compat wurh both | 19:31 |
| fungi | arosen: there's a switch to tell it to keep the localrc | 19:31 |
| *** changbl has joined #openstack-infra | 19:32 | |
| arosen | fungi: ah okay yea i found that one. | 19:32 |
| arosen | thanks for the info. | 19:32 |
| fungi | arosen: export KEEP_LOCALRC=1 | 19:32 |
| *** dims has quit IRC | 19:34 | |
| *** dims has joined #openstack-infra | 19:35 | |
| *** Apsu has joined #openstack-infra | 19:37 | |
| *** cpowell has left #openstack-infra | 19:38 | |
| *** cdent has quit IRC | 19:39 | |
| *** dims has quit IRC | 19:39 | |
| jogo | zaro: got a gerrit question | 19:42 |
| *** dims has joined #openstack-infra | 19:42 | |
| *** btran has quit IRC | 19:43 | |
| jeblair | clarkb, fungi, krotscheck: added bup-storyboard and approved related change | 19:43 |
| krotscheck | bup? | 19:44 |
| jeblair | clarkb, fungi: review.o.o has the needed key, but did not have the backup server's fingerprint in known hosts; i have corrected that, but we should probabl engineer a way to prevent that from happening | 19:44 |
| jeblair | krotscheck: backup user | 19:44 |
| jeblair | krotscheck: 138447 | 19:44 |
| krotscheck | jeblair: Gotcha, thanks | 19:45 |
| jogo | zaro: the command to look for +1s by a specific group doesn't seem to work | 19:45 |
| jogo | label:Code-Review=+1,group=ldap/linux.workflow | 19:45 |
| *** penguinRaider__ has joined #openstack-infra | 19:46 | |
| *** hdd has quit IRC | 19:46 | |
| *** e0ne has quit IRC | 19:48 | |
| *** penguinRaider has quit IRC | 19:50 | |
| *** carl_baldwin has joined #openstack-infra | 19:51 | |
| fungi | jeblair: i'm starting to wonder if we should just globally puppet /etc/ssh/known_hosts | 19:52 |
| jeblair | clarkb, fungi: so let's try to remember to do a restore test tomorrow or next week to check on everything | 19:52 |
| fungi | solves a number of related problems there | 19:52 |
| jeblair | fungi: probably so | 19:52 |
| fungi | problems that can also be solved via sshfp, but only if we have working dnssec everywhere | 19:52 |
| clarkb | jeblair +1 | 19:53 |
| jeblair | fungi: the launch script should propose a change whenever it runs :) | 19:53 |
| *** eharney has quit IRC | 19:53 | |
| *** baoli has quit IRC | 19:53 | |
| fungi | jeblair: ooh! not a bad idea at all | 19:53 |
| *** pc_m has joined #openstack-infra | 19:53 | |
| *** pc_m has quit IRC | 19:54 | |
| fungi | mtreinish: whenever the proxy is up and running, there is a user named "query" with a password of "query which is granted select on subunit2sql.* and nothing else. tested it to make sure it sees no other databases and is refused update/insert | 19:55 |
| *** pc_m has joined #openstack-infra | 19:55 | |
| fungi | mordred: it also seems to be able to run show commands, to your earlier comment, not sure how to cripple that | 19:55 |
| fungi | though maybe the proxy blocks show already? | 19:55 |
| jeblair | fungi: seriously, i read your sentence 3 times before i figured out that the password was not actually "query which is granted select on subunit2sql.* and nothing else." | 19:56 |
| mtreinish | fungi: ok, awesome thanks | 19:57 |
| fungi | jeblair: heh. unterminated string! | 19:57 |
| jeblair | i mean, that's one way of communicating that fact... | 19:57 |
| mtreinish | that's quite the password :) | 19:57 |
| jeblair | fungi is just pen-testing us | 19:57 |
| fungi | and my constant typos are actually a fuzz test, yes | 19:58 |
| fungi | username: query password: query | 19:58 |
| *** kmartin has joined #openstack-infra | 19:59 | |
| mtreinish | hahaha oh wait jeblair I did the same thing you did :) | 19:59 |
| mtreinish | even after you said it | 19:59 |
| jeblair | roflol | 19:59 |
| fungi | clearly my pentest was successful | 19:59 |
| *** e0ne has joined #openstack-infra | 19:59 | |
| *** _nadya_ has joined #openstack-infra | 20:02 | |
| openstackgerrit | Matthew Treinish proposed openstack-infra/project-config: Add nodepool script to preseed testrepository from subunit2sql https://review.openstack.org/136234 | 20:02 |
| zaro | jogo: known issue https://code.google.com/p/gerrit/issues/detail?id=3018 | 20:02 |
| mtreinish | fungi: ok, updated ^^^ with those creds | 20:02 |
| *** HeOS has joined #openstack-infra | 20:02 | |
| *** Ryan_Lane has joined #openstack-infra | 20:03 | |
| zaro | jogo: haven't gotten around to investigating it. | 20:03 |
| *** dprince has quit IRC | 20:03 | |
| *** e0ne has quit IRC | 20:04 | |
| *** spzala has joined #openstack-infra | 20:04 | |
| *** doug-fish has quit IRC | 20:05 | |
| *** doug-fish has joined #openstack-infra | 20:06 | |
| pleia2 | dhellmann: happen to be about? could use your wisdom re: https://review.openstack.org/#/c/140041/ | 20:06 |
| *** cody-somerville has quit IRC | 20:06 | |
| pleia2 | (and really, anyone who cares about XML feeds for specs) | 20:07 |
| *** KurtMartin has joined #openstack-infra | 20:07 | |
| *** cnesa has quit IRC | 20:07 | |
| jogo | zaro: cool, thanks | 20:09 |
| *** kmartin has quit IRC | 20:10 | |
| dhellmann | pleia2: looking | 20:11 |
| pleia2 | thank you | 20:12 |
| fungi | mtreinish: any gut feeling for how long that would take to run? wondering where we would benefit from applying it in our automation | 20:12 |
| dhellmann | pleia2: we can generate an opml file from the rss feeds listed in the index.html file in the directory where that static opml is going, no? | 20:12 |
| dhellmann | we shouldn't need to change anything in the projects | 20:12 |
| dhellmann | the opml file should only list the rss feed urls themselves, not their contents | 20:12 |
| pleia2 | dhellmann: ah good, can you comment with some thoughts? | 20:13 |
| mtreinish | fungi: hopefully it should be pretty quick most of the operations should be almost instant. The only thing that'll take more time is the sql2subunit calls, which I'm not sure about speed on a db with >24M rows | 20:14 |
| dhellmann | pleia2: sure. Can we agree in principle that we'll put a static file in place and then write something to generate it? | 20:14 |
| mtreinish | when the proxy comes up I can test it | 20:14 |
| dhellmann | pleia2: the current contributor isn't a dev, and I don't have time to do it this week, but I will come back and do it | 20:14 |
| arosen | is there a way to trigger zuul to run all the patch sets on review rather then posting recheck xxxx on each one manually? | 20:14 |
| fungi | mtreinish: luckily, we used a larger trove instance for this in anticipation | 20:14 |
| arosen | I know you can script that with ssh and gerrit but just curious if there is a way to do it on the backend. | 20:14 |
| fungi | mtreinish: we have 8gb ram and 100gb disk | 20:14 |
| *** penguinRaider__ has quit IRC | 20:15 | |
| dhellmann | pleia2: actually, belay that, let me spend a few minutes tinkering | 20:15 |
| pleia2 | dhellmann: wfm, I just don't want humans editing an .opml file long term (tbh, it's hard enough to get them to write the index.html file properly on their first go)) | 20:15 |
| fungi | arosen: zuul enqueue | 20:15 |
| fungi | arosen: http://ci.openstack.org/zuul/client.html#usage | 20:15 |
| dhellmann | pleia2: we could generate the index.html, too, if you want | 20:15 |
| fungi | arosen: assuming you're talking about a zuul you manage | 20:15 |
| arosen | fungi: yup, awesome thanks! | 20:15 |
| *** penguinRaider has joined #openstack-infra | 20:15 | |
| mtreinish | fungi: once https://storyboard.openstack.org/#!/story/2000023 is fixed I can implement an average option for sql2subunit which will switch to a single call in that script | 20:15 |
| mtreinish | which would make things much faster | 20:16 |
| pleia2 | dhellmann: that'd be nice too :) | 20:16 |
| mtreinish | fungi: cool, yeah hopefully that'll be big enough :) | 20:16 |
| fungi | mtreinish: yeah, hopefully retrieving the 10 most recent records from that table will go quickly | 20:16 |
| arosen | fungi: If i want to do it for all the patches on review i still need a script that talks to gerrit to get that list right? | 20:16 |
| grue_pm | mtreinish: morning mac t ;-p | 20:16 |
| arosen | this just lets me do it for individual patches on the backend. | 20:17 |
| *** aysyd has quit IRC | 20:17 | |
| fungi | arosen: right, you'd need to query gerrit for change,patchset numbers | 20:17 |
| mtreinish | fungi: it's pulling the ~1500 rows from the >24M for the sql2subunit to generate the stream which I think might be a bit slower | 20:17 |
| arosen | fungi: awesome this is what i was looking for. Thanks! | 20:17 |
| mtreinish | grue_pm: morning :) | 20:17 |
| mtreinish | fungi: and it does that 10x serially which I probably should change | 20:18 |
| *** bhunter71 has quit IRC | 20:18 | |
| mtreinish | but I still need to get a fix for that bug before I move forward with anything on that script... | 20:18 |
| fungi | mtreinish: ahh, yep. i wonder if we would benefit from running it during nodepool image creation so that it's cached locally on each worker for the jobs we care about, and thus doesn | 20:18 |
| fungi | 't add potential instability to each job run | 20:19 |
| mtreinish | fungi: oh I assumed that's how the scripts in project-config/nodepool/scripts were run. That was my intent with it | 20:19 |
| fungi | yep, perfect then. i just worry that every new network-connected dependency we add within a job it makes the runs that much more fragile | 20:20 |
| fungi | but we already accept that risk hundreds of ways during image creation | 20:20 |
| mtreinish | yeah I didn't want to integrate it into the runs because of that | 20:20 |
| fungi | so as long as day-plus freshness is acceptable for this data, should work out | 20:20 |
| mtreinish | which is why even when I get around to finishing the sql repository for testrepository I wasn't going to use it for the gate | 20:21 |
| *** thedodd has joined #openstack-infra | 20:21 | |
| *** cody-somerville has joined #openstack-infra | 20:21 | |
| mtreinish | fungi: yeah that's fine the sample size is so small compared to the overall dataset it won't make a difference either way | 20:21 |
| mtreinish | the scheduler optimization also probably won't make much of a difference either, I just want to shake up the test execution order a bit | 20:22 |
| mtreinish | we should expose some new bugs by doing that | 20:22 |
| fungi | mtreinish: and i only just paid attention to the file path after you said that. we'll presumably want something in nodepool/elements to trigger this for dib too | 20:22 |
| *** ayoung has quit IRC | 20:22 | |
| *** sdake has quit IRC | 20:22 | |
| mtreinish | fungi: ok, I can work on that. I'll put it on my queue for this stuff | 20:23 |
| *** dprince has joined #openstack-infra | 20:23 | |
| *** e0ne has joined #openstack-infra | 20:23 | |
| *** eharney has joined #openstack-infra | 20:23 | |
| *** eharney has joined #openstack-infra | 20:23 | |
| mtreinish | it's all being held up by the centos thing anyway... | 20:23 |
| *** Sukhdev has joined #openstack-infra | 20:23 | |
| mtreinish | clarkb, fungi: is there anything I can do to help with that? | 20:24 |
| *** cnesa has joined #openstack-infra | 20:24 | |
| *** markmcclain has quit IRC | 20:25 | |
| fungi | mtreinish: clarkb was working on figuring out the refs we need to do module installation from git next i think | 20:25 |
| arosen | fungi: I wonder if zuul.rpcclient and the client stuff should be pulled out of the zuul repo to something like: python-zuulclient? This way one doesn't need to install all of zuul to just get the client code | 20:26 |
| pleia2 | so, nova is looking for an infra home for http://54.201.139.117/nova-bugs.html (code is on github, currently hosted on AWS), thoughts about project-specific things on status.o.o? | 20:26 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack-infra/devstack-gate: Add tooz to devstack-vm-gate-wrap.sh https://review.openstack.org/141151 | 20:26 |
| fungi | arosen: perhaps eventually (zuul-cloner would fall in that category too) but zuul itself isn't especially heavyweight... just don't start any of its services | 20:27 |
| jeblair | pleia2: is there anything nova-specific about that? | 20:27 |
| arosen | fungi: yup i understand, not a big deal | 20:27 |
| pleia2 | jeblair: with a little work I think it could work with other projects using launchpad for bugs, but as it is now the script takes 30 minutes to generate the report | 20:27 |
| fungi | pleia2: looks sort of like reviewday | 20:27 |
| fungi | pleia2: reviewday mk ii? | 20:28 |
| pleia2 | fungi: yeah, but takes much longer to run | 20:28 |
| fungi | we have machines for that | 20:28 |
| jeblair | cpu time is cheaper than people time :) | 20:28 |
| pleia2 | indeed | 20:29 |
| *** tonytan4ever has quit IRC | 20:29 | |
| pleia2 | I pull in dprince and see if maybe we can reviewday mk ii this | 20:29 |
| *** kmartin has joined #openstack-infra | 20:29 | |
| jeblair | pleia2: who runs / wrote that? | 20:30 |
| fungi | i see features here which have been wishlist for reviewday for a while | 20:30 |
| pleia2 | Tracy Jones of VMware https://github.com/tracyajones/nova_bugs | 20:30 |
| pleia2 | she was at our infra sprint in NYC | 20:30 |
| jeblair | cool, so she already knows how to bootstrap it in infra ;) | 20:30 |
| pleia2 | :) | 20:31 |
| fungi | and how to start hanging out in irc! | 20:31 |
| *** dimtruck is now known as zz_dimtruck | 20:31 | |
| *** fifieldt_ has quit IRC | 20:32 | |
| *** zz_dimtruck is now known as dimtruck | 20:32 | |
| jeblair | next time we have a bootstrap thing, we should have 1 hour of practicing hanging out in irc. :) | 20:32 |
| jeblair | no talking! | 20:32 |
| pleia2 | haha ++ | 20:32 |
| fungi | looks like she's in #-dev, our channel's just not cool enough | 20:32 |
| jeblair | i am _certain_ we told everyone which one was the cool channel | 20:32 |
| fungi | i remember seeing it on paper and on a projector | 20:33 |
| mtreinish | jeblair: #openstack-qa right... | 20:33 |
| jeblair | mtreinish: close. you got most of the letters right. | 20:33 |
| fungi | q is just an abbrev for infr | 20:33 |
| *** KurtMartin has quit IRC | 20:34 | |
| jeblair | infruite so. | 20:34 |
| fungi | most recent gearman timeout on nodepool-dev was 2014-12-11 10:18:36 | 20:35 |
| fungi | seems to not be especially broken | 20:35 |
| *** pblaho has quit IRC | 20:35 | |
| fungi | looks like that was right around the most recent large node spike on the graph | 20:35 |
| mordred | mtreinish: patch https://review.openstack.org/#/c/140522/1 ... how did you test that? because nothing about that patch seems like lines that would affect what the comit message says they would | 20:36 |
| *** penguinRaider__ has joined #openstack-infra | 20:36 | |
| * mordred wants to be able to reproduce the problem - because it may be a pbr bug | 20:36 | |
| *** tonytan4ever has joined #openstack-infra | 20:36 | |
| jeblair | fungi, clarkb: ooh! zuul-gear disconnect at 17:17 ! | 20:37 |
| fungi | huh | 20:37 |
| fungi | seems like it survived that remarkably well | 20:37 |
| jeblair | 714007 packets captured | 20:37 |
| jeblair | 1428014 packets received by filter | 20:37 |
| jeblair | 0 packets dropped by kernel | 20:37 |
| jeblair | looks like a good tcpdump. i will review after lunch. | 20:37 |
| mtreinish | mordred: I did pip install . on the repo checked the install dir for the alembic.ini file and saw it wasn't there | 20:38 |
| mtreinish | I also got errors matching that gertty couldn't find the ini file | 20:38 |
| mtreinish | then I uninstalled it, added the manifest file, reinstalled it confirmed the file was there and gerrty ran without the errors | 20:38 |
| mtreinish | mordred: I had the same issue with subunit2sql at one point: http://git.openstack.org/cgit/openstack-infra/subunit2sql/commit/?id=fe24b4a39684e2792aa49fec214a7f97fb62ec71 | 20:39 |
| mordred | brain boggles | 20:39 |
| *** penguinRaider has quit IRC | 20:39 | |
| mordred | MANIFEST.in should not have anything to do with which files get installed | 20:39 |
| dstufft | wellll | 20:40 |
| mordred | oh. does it? | 20:40 |
| dstufft | not exactly | 20:40 |
| *** unicell has quit IRC | 20:40 | |
| dstufft | but if controls what is put into the tarball, and you can't install something that isn't in the tarball | 20:40 |
| dstufft | it controls* | 20:40 |
| mordred | dstufft: does "pip install ." make a tarbal first? | 20:41 |
| dstufft | um | 20:41 |
| dstufft | no I don't think so | 20:41 |
| *** tjones1 has joined #openstack-infra | 20:41 | |
| * mordred has a hunch ... | 20:43 | |
| *** wenlock has joined #openstack-infra | 20:44 | |
| *** thedodd has quit IRC | 20:44 | |
| dhellmann | pleia2: ok, I have a script to generate both files from a yaml input file. I need some help figuring out which job I tie into to make the publishing work. | 20:44 |
| *** ociuhandu has quit IRC | 20:45 | |
| fungi | http://lists.openstack.org/pipermail/openstack-dev/2014-December/052757.html wants to spark a discussion on job filtering and pipeline proliferation | 20:45 |
| mtreinish | mordred: IIRC, I might have looked at the pbr code to find it was only copying .py files or something | 20:46 |
| arosen | is there an easy way to tell jenkins to not delete a node after a test run so one can login to debug why something is failing? | 20:46 |
| mordred | mtreinish: well, it sholdn't be | 20:46 |
| mtreinish | which could have been why I added a manifest, but don't quote me on any of that :) | 20:46 |
| * mordred is poking | 20:46 | |
| fungi | arosen: 'nodepool hold <id>' | 20:46 |
| jeblair | fungi: yeah, clarkb mentioned that earlier. clarkb did not understand what problem was attempting to be solved, and i did not understand the conclusion of the message which sort of just trails off. | 20:47 |
| fungi | arosen: http://ci.openstack.org/nodepool/operation.html#usage | 20:47 |
| *** e0ne has quit IRC | 20:47 | |
| arosen | thanks | 20:47 |
| arosen | hold | 20:47 |
| mordred | ok. I have verified that regardless of MANIFEST.in, gertty/alembic/ and gertty/alembic.ini are included in the tarball produced by setup.py sdist as expected | 20:47 |
| mordred | I have also verified that VERY UNEXPECTEDLY, they are not installed even though install_package_data should be being set | 20:48 |
| mordred | dstufft: has anything changed recently with install_package_data ? | 20:48 |
| dstufft | mordred: um, I do't think so, but it might have been | 20:48 |
| *** alexpilotti has quit IRC | 20:49 | |
| *** penguinRaider__ has quit IRC | 20:49 | |
| *** fifieldt_ has joined #openstack-infra | 20:49 | |
| *** penguinRaider has joined #openstack-infra | 20:49 | |
| pleia2 | dhellmann: great, I'm not as familiar with the publishing jobs though, so hopefully someone else can help | 20:49 |
| fungi | so this is a difference between what sdist generates and what pip install . will install without a tarball? | 20:49 |
| fungi | interesting and surprising behavior difference | 20:50 |
| mordred | well, not quite | 20:50 |
| mtreinish | mordred: oh, I just rechecked the source code and I think that I was just jumping to conclusions. I think I was confusing the autodoc find module stuff | 20:50 |
| mordred | except yes | 20:50 |
| *** wenlock has quit IRC | 20:51 | |
| mordred | dstufft: I can confirm that adding an entry to MANIFEST.in DOES affect what pip is installing | 20:51 |
| mordred | dstufft: althuogh the sdist output is the same in each case | 20:51 |
| dhellmann | fungi, clarkb, jeblair, mordred: can one of you point me to the job that handles publishing the specs repo home page from project-config/specs/index.html to wherever? we want to generate the file from data instead of editing html by hand, but I don't know what I need to change in the job structure. | 20:51 |
| fungi | dhellmann: it's not a job, just a puppeted file | 20:52 |
| *** sdake has joined #openstack-infra | 20:52 | |
| *** sdake has quit IRC | 20:52 | |
| *** sdake has joined #openstack-infra | 20:52 | |
| mordred | oh. nope. still not clean enough test case | 20:52 |
| mordred | one sec | 20:52 |
| *** reed has quit IRC | 20:52 | |
| dhellmann | fungi: is that still in project-config somewhere? | 20:52 |
| mordred | rm -rf foo gertty.egg-info/ dist/ ; virtualenv foo ; foo/bin/pip install . ; find foo | grep alembic.ini | 20:53 |
| mordred | that works and alembic.ini is installed as expected | 20:53 |
| fungi | dhellmann: project-config | 20:53 |
| *** harlowja has joined #openstack-infra | 20:53 | |
| *** penguinRaider__ has joined #openstack-infra | 20:53 | |
| fungi | dhellmann: http://git.openstack.org/cgit/openstack-infra/project-config/tree/specs/index.html | 20:53 |
| *** mriedem has quit IRC | 20:54 | |
| fungi | dhellmann: it's just continuously deployed onto the webserver from there for now | 20:54 |
| mordred | but I did see the behavior that mtreinish was reporting | 20:54 |
| mordred | so the question is - what's the actual problem and what was the actual solution | 20:54 |
| dhellmann | fungi: I want to replace that static file with one we generate. I have a script to generate a file. I'm trying to figure out what I would have to change to run that script. | 20:54 |
| pleia2 | thanks fungi | 20:54 |
| fungi | dhellmann: we figured that was a fine starting point, until someone got annoyed enough to expend the effort generate it dynamically somehow | 20:54 |
| dhellmann | fungi: I am annoyed enough to be doing that *as we speak* :-) | 20:55 |
| fungi | dhellmann: you're talking about the top-level index, right? | 20:55 |
| dhellmann | yep | 20:55 |
| dhellmann | and I will be adding an OPML file at the same time | 20:55 |
| fungi | okay, just confirming. sublevel indices are handled by the specs projects' jobs | 20:55 |
| dhellmann | right, these files will both be top-level files | 20:55 |
| *** penguinRaider has quit IRC | 20:57 | |
| mtreinish | jogo: I think this is why we stopped using the round robin mirror before: http://logs.openstack.org/36/141136/1/gate/gate-tempest-dsvm-full/dec39bb/logs/devstacklog.txt.gz#_2014-12-11_20_44_08_459 | 20:58 |
| *** tjones1 has quit IRC | 20:59 | |
| fungi | mtreinish: was that an experimental change? | 21:00 |
| fungi | mtreinish: or are we actually retrieving fedora cloud images over the internet in some of our jobs now? | 21:00 |
| mtreinish | fungi: no, a regular tempest-full gate job | 21:00 |
| *** bswartz has quit IRC | 21:00 | |
| mtreinish | fungi: we shouldn't be | 21:01 |
| *** tnovacik has quit IRC | 21:01 | |
| fungi | mtreinish: i could have sworn we cached those on the workers' filesystems | 21:01 |
| mtreinish | fungi: that's what I thought too | 21:01 |
| *** carl_baldwin has quit IRC | 21:01 | |
| *** ChuckC has quit IRC | 21:02 | |
| *** flwang has joined #openstack-infra | 21:02 | |
| *** tjones1 has joined #openstack-infra | 21:02 | |
| *** carl_baldwin has joined #openstack-infra | 21:02 | |
| flwang | hi there, anybody know how the infra update the tempest.conf to get the correct image id and network id? Thanks a lot | 21:02 |
| *** mrmartin has quit IRC | 21:03 | |
| fungi | mtreinish: 20141203 looks very recent. maybe this has changed in devstack in the past day or two and our images are behind? | 21:03 |
| * fungi checks things | 21:03 | |
| mtreinish | fungi: could be, let me check the devstack git log | 21:03 |
| *** penguinRaider__ has quit IRC | 21:03 | |
| *** penguinRaider has joined #openstack-infra | 21:03 | |
| *** Masahiro has joined #openstack-infra | 21:03 | |
| mtreinish | fungi: oh, I think I remember a devstack patch switching titled switching to fedora 21 fly by at one point | 21:04 |
| fungi | mtreinish: yep | 21:04 |
| fungi | mtreinish: merged a few hours ago | 21:04 |
| mtreinish | fungi: http://git.openstack.org/cgit/openstack-dev/devstack/commit/?id=21dbe993348b794a1b77c4f9db0081d1cc32138c | 21:04 |
| *** bswartz has joined #openstack-infra | 21:04 | |
| fungi | mtreinish: so hopefully that clears up as soon as our next image updates happen | 21:05 |
| mtreinish | I guess we'll get a bunch of random errors until then | 21:05 |
| mtreinish | I guess I should write an e-r query for that... | 21:05 |
| *** andreykurilin_ has quit IRC | 21:06 | |
| *** baoli has joined #openstack-infra | 21:07 | |
| *** pc_m has quit IRC | 21:07 | |
| notmyname | any chance that current or future gerrit will have a built-in link shortener? most shortener services bork on the long URLs (beyond just being a separate service) | 21:07 |
| fungi | there ought to be a way to stage these in the future so you can merge the bits that cause us to pre-cache on our images and then a day or two later merge the bits that make devstack actively try to retrieve | 21:08 |
| *** otter768 has joined #openstack-infra | 21:08 | |
| *** Masahiro has quit IRC | 21:08 | |
| pleia2 | gate-infra-puppet-apply-centos6 is reliably failing at the moment (in case no one is looking at it already), looking at logs now | 21:08 |
| fungi | zaro: see notmyname's question. you're more in touch with the gerrit roadmap than most of us | 21:08 |
| fungi | pleia2: yeah, it's been under discussion | 21:09 |
| pleia2 | fungi: ah good | 21:09 |
| mordred | notmyname: "most shortener services bork on the long URLs" ... are you telling me that shortener services are not able to shorten the URLs? | 21:09 |
| fungi | pleia2: current next step seems to be drafting a change to install modules from git and see if that works | 21:09 |
| *** craigbr has quit IRC | 21:09 | |
| pleia2 | fungi: ok, thanks | 21:09 |
| mattoliverau | Morning | 21:09 |
| pleia2 | clearly my do reviews, read email and keep up with IRC superpowers aren't engaged today ;) | 21:10 |
| fungi | pleia2: it seems that puppet module installation is subject to dependency hell, and one of our recent module additions has tipped the scale into nearly-intractible or something | 21:10 |
| *** penguinRaider__ has joined #openstack-infra | 21:10 | |
| pleia2 | puppet is fun | 21:10 |
| fungi | "puppet for fun and madness" | 21:11 |
| *** ihrachyshka has quit IRC | 21:11 | |
| notmyname | mordred: seems to be inconsistent. sometimes it works. sometimes it doesn't | 21:11 |
| *** otter768 has quit IRC | 21:12 | |
| *** carl_baldwin has quit IRC | 21:12 | |
| *** penguinRaider has quit IRC | 21:14 | |
| *** dprince has quit IRC | 21:14 | |
| clarkb | ok pho was good. line was long | 21:15 |
| clarkb | fungi: should I go ahead and start working on gitifying all of the module installs? | 21:15 |
| *** luqas has joined #openstack-infra | 21:15 | |
| fungi | clarkb: presumably. i have been trying to catch up still so didn't dig further into it yet | 21:15 |
| *** luqas__ has quit IRC | 21:16 | |
| clarkb | ok let me give that a go | 21:16 |
| *** ihrachyshka has joined #openstack-infra | 21:16 | |
| *** Longgeek has quit IRC | 21:17 | |
| *** baoli has quit IRC | 21:18 | |
| *** baoli has joined #openstack-infra | 21:19 | |
| *** penguinRaider has joined #openstack-infra | 21:19 | |
| *** penguinRaider__ has quit IRC | 21:19 | |
| *** _nadya_ has quit IRC | 21:20 | |
| *** luqas has quit IRC | 21:21 | |
| *** luqas has joined #openstack-infra | 21:21 | |
| *** penguinRaider__ has joined #openstack-infra | 21:21 | |
| *** luqas has quit IRC | 21:24 | |
| *** luqas has joined #openstack-infra | 21:24 | |
| *** penguinRaider has quit IRC | 21:25 | |
| *** ChuckC has joined #openstack-infra | 21:25 | |
| *** erikmwilson has quit IRC | 21:25 | |
| mtreinish | fungi: yeah, that's what I was thinking, I guess we could do it 2 part and have one which manually added it to the precache list | 21:26 |
| pleia2 | fungi jeblair - dprince has reminded us that our limiting factor on reviewday/nova bugs is not our side for compute power, it's epic slow launchpad lib, but we're discussing merge plans anyway, so thanks for the input | 21:26 |
| mtreinish | and then wait 24hrs to merge using it (and removing the manual add) | 21:26 |
| pleia2 | long live storyboard | 21:26 |
| pleia2 | (until storyboard has an api, I'm all unicorns and rainbows about it <3) | 21:27 |
| *** ihrachyshka has quit IRC | 21:27 | |
| mtreinish | fungi: I'll add something about caching images on the qa meeting agenda for in ~30min | 21:29 |
| *** luqas__ has joined #openstack-infra | 21:29 | |
| armax | I have a question to a permission failure I see while handling logs during a job build | 21:29 |
| armax | I can see here: http://logs.openstack.org/51/140851/15/check/check-neutron-dsvm-functional/c97b26f/console.html#_2014-12-11_19_23_12_650 | 21:30 |
| krotscheck | pleia2: It has an API. | 21:30 |
| *** luqas has quit IRC | 21:30 | |
| krotscheck | pleia2: In fact, it _is_ an API. | 21:30 |
| armax | is there any kind soul that could give me an input as to why this happens? | 21:30 |
| pleia2 | krotscheck: oh :) | 21:30 |
| *** MarkAtwood has joined #openstack-infra | 21:31 | |
| krotscheck | Urm, I’m getting a cert invalid from storyboard.o.o | 21:31 |
| pleia2 | krotscheck: what were we wait on, python stuff? | 21:31 |
| krotscheck | In chrome. | 21:31 |
| pleia2 | hm, my chrome thinks it's ok | 21:32 |
| krotscheck | pleia2: The API client, and the ability to issue tokens for api users. | 21:32 |
| *** mbacchi has quit IRC | 21:32 | |
| krotscheck | Odd. | 21:32 |
| pleia2 | krotscheck: ah, gotcha | 21:32 |
| * krotscheck is on crappy internets | 21:32 | |
| krotscheck | pleia2: THe API client is busy being built. The issue token bit is under review. | 21:32 |
| pleia2 | krotscheck: thanks, I'll have a look | 21:33 |
| fungi | armax: my first instinct is that the file probably doesn't have execute permissions set on it in the filesystem | 21:33 |
| *** penguinRaider has joined #openstack-infra | 21:33 | |
| *** penguinRaider__ has quit IRC | 21:33 | |
| openstackgerrit | Doug Hellmann proposed openstack-infra/project-config: Generate specs.o.o root page from template https://review.openstack.org/141173 | 21:33 |
| dhellmann | pleia2, fungi : ^^ is going to need some work to unhook the puppet stuff, I expect | 21:33 |
| fungi | armax: chmod a+x the file or explicitly invoke the interpreter you want to execute it | 21:33 |
| *** markvan has quit IRC | 21:34 | |
| armax | fungi: let me check the permissions, I can’t recall for sure | 21:34 |
| fungi | dhellmann: the puppet unhook would involve removing the file resource for it in the system-config repo (openstack_project::static class) | 21:34 |
| armax | fungi: thanks for the tip | 21:34 |
| *** tonytan4ever has quit IRC | 21:34 | |
| fungi | armax: oh, wait | 21:34 |
| fungi | armax: i misread the error | 21:35 |
| *** signed8bit has quit IRC | 21:35 | |
| fungi | armax: so it's saying that line 11 of the post_test_hook.sh script is trying to do something to .testrepository/0 and _that's_ what's denied | 21:35 |
| *** signed8bit has joined #openstack-infra | 21:35 | |
| armax | fungi: ya | 21:35 |
| fungi | armax: just a sec i'll look at what that operation is | 21:35 |
| armax | fungi: hang on | 21:36 |
| armax | fungi: I’ll get it for you | 21:36 |
| fungi | armax: is it from the change being tested, or already in the neutron repo? | 21:36 |
| fungi | k | 21:36 |
| armax | fungi: it fails here: .tox/$venv/bin/subunit-1to2 < .testrepository/0 > ./subunit_log.txt | 21:36 |
| *** markvan has joined #openstack-infra | 21:36 | |
| armax | fungi: I dumped some info about the env here: http://logs.openstack.org/51/140851/14/check/check-neutron-dsvm-functional/52d76f8/console.html#_2014-12-11_17_53_09_356 | 21:36 |
| armax | fungi: I can’t spot anything obvious | 21:37 |
| *** shashankhegde has quit IRC | 21:37 | |
| fungi | armax: okay, so that looks like a read permission error | 21:37 |
| mtreinish | armax: well besides the fact you shouldn't give a binary a .txt extension that should work assuming the user has read perms | 21:37 |
| pleia2 | dhellmann: thanks | 21:38 |
| armax | mtreinish: I copied this from run-tox.sh from the scripts project-infra :) | 21:38 |
| openstackgerrit | Monty Taylor proposed openstack-infra/project-config: Add elements for Infra servers https://review.openstack.org/140840 | 21:38 |
| *** baoli has quit IRC | 21:38 | |
| armax | mtreinish: I am still experimenting though, I am not sure I really want to maintain my own script | 21:38 |
| *** bhunter71 has joined #openstack-infra | 21:38 | |
| armax | mtreinish, fungi: ultimately the problem I am trying to solve is to get a nice html output for the functional neutron job | 21:39 |
| *** baoli has joined #openstack-infra | 21:39 | |
| fungi | armax: it's a shame your ls doesn't add the -a flag there, so i can't see the ownership on .testrepository | 21:39 |
| mtreinish | armax: oh, well that's done indpendently of copying the subunit out | 21:39 |
| openstackgerrit | Doug Hellmann proposed openstack-infra/system-config: Generate specs.o.o root page from template https://review.openstack.org/141175 | 21:39 |
| *** EmilienM is now known as EmilienM|pto | 21:39 | |
| dhellmann | fungi, pleia2 : ^^ | 21:39 |
| fungi | armax: oh, nevermind, there it is further down | 21:39 |
| fungi | armax: aha, only readable by the stack user | 21:40 |
| fungi | armax: and that post-hook script is i think running as jenkins not stack | 21:40 |
| armax | but then I do: + sudo -H -u stack .tox/dsvm-functional/bin/subunit-1to2 | 21:40 |
| armax | fungi: shouldn’t that do it? | 21:40 |
| armax | can the ‘jenkins’ user impersonate the ‘stack’ user? | 21:41 |
| mtreinish | armax: but it reads the file from stdin which is done as the script user | 21:41 |
| fungi | armax: you're running subunit-1to2 as stack, but the error looks like it's coming from something running after that | 21:41 |
| armax | fungi: ah, that must be it then | 21:41 |
| fungi | mtreinish: depends on how sudo privs are for this particular environment | 21:41 |
| fungi | armax: is the neutron/tests/functional/contrib/post_test_hook.sh coming from master? | 21:42 |
| *** pc_m has joined #openstack-infra | 21:42 | |
| armax | fungi: ya, but I posted a patch to geneate the testr_results.html | 21:42 |
| *** pc_m has quit IRC | 21:42 | |
| armax | fungi: https://review.openstack.org/#/c/140851/ | 21:42 |
| mtreinish | fungi: no I mean its 'sudo subunit-1to2 < file' I think only the subunit-1to2 runs as stack in that case | 21:42 |
| fungi | armax: yep, found that | 21:43 |
| *** pc_m has joined #openstack-infra | 21:43 | |
| armax | fungi: I had a few brainless iterations :) | 21:43 |
| *** doug-fish has quit IRC | 21:43 | |
| *** erikmwilson has joined #openstack-infra | 21:44 | |
| mordred | oops | 21:44 |
| nibalizer | clarkb: so i think you want to incorporate the version bumps from here https://review.openstack.org/#/c/140902/ | 21:44 |
| fungi | mtreinish: armax: yep, https://review.openstack.org/#/c/140851/15/neutron/tests/functional/contrib/post_test_hook.sh line 11 | 21:44 |
| dhellmann | fungi, pleia2 : I don't understand the failure of gate-infra-puppet-apply-centos6 for the change to project-config :-/ #noob | 21:45 |
| mordred | mtreinish: now that the debs->apts change landed in devstack, there should be no instances of apts dirs anywhere anymore, right? | 21:45 |
| clarkb | nibalizer: I want to start with matching exactly what we have on the master now | 21:45 |
| clarkb | also this is tedious checking all the things | 21:45 |
| mtreinish | mordred: probably not, but I would check with dtroyer to be sure | 21:45 |
| mordred | dtroyer: ^^ ? | 21:45 |
| fungi | armax: so as mtreinish points out you're using sudo to call subunit-1to2 but then your stdin redirect from .testrepository/0 is from the calling environment, not sudo | 21:45 |
| pleia2 | dhellmann: it's broken for everyone right now | 21:45 |
| nibalizer | clarkb: mk | 21:46 |
| fungi | armax: i'll leave a suggested workaround in an inline comment | 21:46 |
| armax | fungi: I’ll look into that, thanks! | 21:46 |
| dhellmann | pleia2: whew, I thought I was going to have to level up again :-) | 21:46 |
| mtreinish | armax: I would just do sudo cat file | subunit-1to2 > testrepository.subunit | 21:46 |
| *** hashar has joined #openstack-infra | 21:46 | |
| nibalizer | one of the inocmpatibilites is that we ask for puppetdb 3.0.1 which requires postgres <4 bu then we install postgres 4 | 21:46 |
| jhesketh | Morning | 21:46 |
| armax | mtreinish: yeah, that’s a good point | 21:46 |
| nibalizer | but we can solve that problem after your patch | 21:46 |
| openstackgerrit | Monty Taylor proposed openstack-infra/project-config: Support name change in devstack to debs https://review.openstack.org/141177 | 21:46 |
| mordred | clarkb, fungi, jhesketh, mtreinish ^^ | 21:47 |
| mtreinish | mordred: I just checked the repo, apts is symlinked to debs in devstack/files | 21:47 |
| mordred | mtreinish: oh right - i remember that now. So we don't need to rush/worry | 21:47 |
| fungi | armax: comment inline on that change | 21:47 |
| armax | mtreinish, fungi or I could just do chmod o+r to the .testrepository dir | 21:47 |
| armax | fungi: thanks! | 21:48 |
| *** dimtruck is now known as zz_dimtruck | 21:48 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/project-config: Add elements for Infra servers https://review.openstack.org/140840 | 21:48 |
| openstackgerrit | Monty Taylor proposed openstack-infra/project-config: Move to using package-installs.yaml https://review.openstack.org/140841 | 21:48 |
| armax | fungi: I figured ultimately that would have fixed it, but I wanted to get more input from you on the right way to go about it | 21:48 |
| *** tjones1 has quit IRC | 21:48 | |
| clarkb | mordred: we should make a rule that all software ever tag all releases | 21:49 |
| fungi | armax: yes, making .testrepository and its contents world-readable is another option, however you may end up with write permission problems since you're still redirecting stdout to a file with > and that's also handled bye the calling shell rather than elevated with sudo | 21:49 |
| fungi | s/bye/by/ | 21:49 |
| armax | fungi: good point | 21:50 |
| fungi | armax: so if you go that route you'll also want to make sure wherever you're writing has write allowed for that calling user (or use the syntax in added in my comment) | 21:50 |
| *** penguinRaider__ has joined #openstack-infra | 21:50 | |
| *** andreaf has joined #openstack-infra | 21:50 | |
| armax | fungi: I’ll adopt your suggestion, it seems the safest | 21:50 |
| fungi | armax: and if the calling user can read the file and write to the destination you want, then there's no need to use sudo there anyway | 21:50 |
| *** tjones1 has joined #openstack-infra | 21:50 | |
| mtreinish | armax: iirc d-g just copies the file before doing any operations on it, maybe that would be easier | 21:50 |
| armax | fungi: the calling user is jenkins as far as I can tell | 21:51 |
| fungi | armax: yeah, it would be at that point in the job | 21:51 |
| armax | mtreinish: it doesn’t seem to work for the functional jobs | 21:51 |
| armax | I was unable to roocause why that doesn’t do it, I looked at the d-g config for neutron-functional in the project-config | 21:51 |
| *** penguinRaider__ has quit IRC | 21:52 | |
| *** Rockyg has joined #openstack-infra | 21:52 | |
| armax | but I am still digesting all the various magics tricks of it | 21:52 |
| *** penguinRaider__ has joined #openstack-infra | 21:52 | |
| *** dustins has quit IRC | 21:52 | |
| *** signed8b_ has joined #openstack-infra | 21:52 | |
| *** penguinRaider has quit IRC | 21:54 | |
| armax | fungi, mtreinish: thanks for your input, much appreciated | 21:54 |
| *** shashankhegde has joined #openstack-infra | 21:54 | |
| *** signed8bit has quit IRC | 21:55 | |
| fungi | armax: any time! | 21:56 |
| *** oomichi has joined #openstack-infra | 21:56 | |
| *** amitgandhinz has quit IRC | 21:57 | |
| *** amitgandhinz has joined #openstack-infra | 21:58 | |
| *** signed8b_ has quit IRC | 21:58 | |
| mtreinish | armax: sure np, 1 more thing you should still rename subunit.txt to something else | 21:58 |
| mtreinish | because when you download in either apache, or the http clients like to try and encode it as if it were text | 21:59 |
| mtreinish | which breaks parsing it | 21:59 |
| armax | mtreinish: doh, I think you’re right, otherwise the publisher won’t find it, will it? | 21:59 |
| *** MaxV has joined #openstack-infra | 21:59 | |
| mtreinish | armax: d-g uses testrepository.subunit which works fine | 21:59 |
| armax | mtreinish: ok, let me dig into this some more..it’s been a useful learning experience | 22:00 |
| mtreinish | armax: oh, I didn't even think of that, yeah it won't get picked up if it's subunit.txt :) | 22:00 |
| *** dims has quit IRC | 22:00 | |
| *** dims has joined #openstack-infra | 22:01 | |
| *** radez is now known as radez_g0n3 | 22:02 | |
| *** yjiang5_away is now known as yjiang5 | 22:02 | |
| *** markmcclain has joined #openstack-infra | 22:02 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config: Install all puppet modules from source https://review.openstack.org/141182 | 22:02 |
| *** amitgandhinz has quit IRC | 22:03 | |
| clarkb | fungi: ^ that probably needs the rm -rf that mrmartin's change does too | 22:03 |
| clarkb | and we should think long and hard about going this route | 22:03 |
| fungi | clarkb: checking it out--thanks | 22:03 |
| *** dims has quit IRC | 22:05 | |
| fungi | clarkb: cosmetic issue marked inline. i'm sure this will need further iteration based on current test results, so feel free to correct later | 22:06 |
| *** bswartz has quit IRC | 22:06 | |
| clarkb | ya I will be pushing a patch to do the rm in the test | 22:06 |
| clarkb | so comment away | 22:06 |
| openstackgerrit | Doug Hellmann proposed openstack/requirements: Add developer script to build wheels https://review.openstack.org/141184 | 22:07 |
| fungi | instafailed on centos6 | 22:07 |
| *** shashankhegde has quit IRC | 22:07 | |
| clarkb | huh | 22:07 |
| * nibalizer looks | 22:08 | |
| fungi | the loglink at the end of the console log is also no good | 22:08 |
| fungi | jhesketh: ^ https://jenkins05.openstack.org/job/gate-infra-puppet-apply-centos6/417/console | 22:08 |
| *** signed8bit has joined #openstack-infra | 22:08 | |
| fungi | jhesketh: has a link to http://logs.openstack.org/82/141182/1/check/gate-infra-puppet-apply-centos6/300a184/index.html which is a 404 for me | 22:08 |
| clarkb | hrm I thought the from source stuff would rm an exisint module and replace it from source for me | 22:09 |
| clarkb | doesn't look like it (which is why I didn't start with the rm things) | 22:09 |
| *** dannywilson has quit IRC | 22:09 | |
| *** dannywilson has joined #openstack-infra | 22:10 | |
| *** dannywilson has quit IRC | 22:11 | |
| *** dannywil_ has joined #openstack-infra | 22:11 | |
| *** dkliban is now known as dkliban_afk | 22:11 | |
| *** patrickeast_ has joined #openstack-infra | 22:12 | |
| *** andreaf has quit IRC | 22:13 | |
| *** asettle is now known as asettle-BRB | 22:13 | |
| *** asettle-BRB has quit IRC | 22:13 | |
| *** reed has joined #openstack-infra | 22:14 | |
| fungi | jeblair: clarkb: any feel for how long we should wait before we rename nodepool-dev in dns and decom the old precise nodepool? | 22:14 |
| fungi | i guess we can undo the loosened iptables rules at the same time if we want | 22:14 |
| *** patrickeast has quit IRC | 22:14 | |
| *** patrickeast_ is now known as patrickeast | 22:14 | |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config: Install all puppet modules from source https://review.openstack.org/141182 | 22:15 |
| clarkb | let us see how that does | 22:15 |
| clarkb | fungi: I was thinking tuesday just because the next couple work days are terrible for me. And today felt too soon | 22:15 |
| clarkb | fungi: but if you want to do it tomorrow or monday I say go for it. | 22:15 |
| clarkb | fungi: I think I would change DNS now though | 22:15 |
| clarkb | then clean up alien images tomorrow/monday | 22:16 |
| fungi | clarkb: well, waiting until there are more people around is fine. it's not like it's causing major trouble | 22:16 |
| jeblair | yeah. i think we're stable enough that we can stick with what we've got | 22:16 |
| jeblair | i'm digging into the zuul timout now | 22:17 |
| jeblair | but i'm also cool waiting until next week; no rush. | 22:17 |
| *** mattfarina has quit IRC | 22:17 | |
| *** ivar-laz_ has quit IRC | 22:18 | |
| fungi | clarkb: good point that puppet module list may not have been finding everything installed at that point in the script | 22:19 |
| clarkb | fungi: well it clones now | 22:19 |
| clarkb | fungi: but the apply still fails | 22:19 |
| fungi | Error: Invalid parameter provider on Vcsrepo | 22:20 |
| fungi | again | 22:20 |
| fungi | are we ending up with the wrong version installed? | 22:21 |
| fungi | should the job also spit out the puppet module list at the end? | 22:21 |
| *** ryanpetrello has quit IRC | 22:22 | |
| *** ivar-lazzaro has joined #openstack-infra | 22:22 | |
| nibalizer | fungi: thats a pretty good idea | 22:22 |
| clarkb | fungi: I can add that to my change | 22:22 |
| clarkb | that said I think we are getting the correct version installed now | 22:23 |
| clarkb | but let us see | 22:23 |
| fungi | sort of like our python jobs do with pip freeze | 22:23 |
| fungi | a helpful sanity check (we can use all the sanity we can get) | 22:23 |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config: Install all puppet modules from source https://review.openstack.org/141182 | 22:23 |
| *** Sukhdev has quit IRC | 22:23 | |
| *** penguinRaider__ has quit IRC | 22:24 | |
| *** penguinRaider has joined #openstack-infra | 22:24 | |
| clarkb | you know | 22:25 |
| clarkb | this might be fighting the zuul cloner too | 22:25 |
| fungi | oho | 22:25 |
| clarkb | it shouldn't because if its already git and not listed in integration modules it should be left alone | 22:25 |
| nibalizer | it shouldn't be though right | 22:25 |
| clarkb | ya it shouldnt | 22:25 |
| nibalizer | becuaes yea that ^ | 22:25 |
| fungi | shades of devstack-gate jobs and install from pip vs source | 22:25 |
| *** dims__ has joined #openstack-infra | 22:26 | |
| *** vigneshvar has quit IRC | 22:26 | |
| fungi | clarkb: set -e kills it before it gets to the module list | 22:27 |
| clarkb | gah | 22:27 |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config: Install all puppet modules from source https://review.openstack.org/141182 | 22:28 |
| openstackgerrit | Armando Migliaccio proposed openstack-infra/project-config: Check that Gerrit ACL files are normalized https://review.openstack.org/140821 | 22:28 |
| openstackgerrit | Armando Migliaccio proposed openstack-infra/project-config: Check that Gerrit ACL files are normalized https://review.openstack.org/140821 | 22:30 |
| armax | fungi: ^^^ I collapsed the ACL checks in a single file | 22:30 |
| fungi | armax: cool | 22:31 |
| clarkb | fungi: nibalizer https://jenkins02.openstack.org/job/gate-infra-puppet-apply-centos6/558/console | 22:31 |
| clarkb | so puppet module list is useless | 22:31 |
| nibalizer | ty ty | 22:31 |
| clarkb | but ls shows us things in /etc/puppet/modules | 22:31 |
| jcoufal | hey, can anybody help with gating problem? | 22:32 |
| *** andreaf has joined #openstack-infra | 22:32 | |
| nibalizer | i wonder if the modulepath got set to not /etc/puppet/modules | 22:32 |
| *** sabari_ is now known as zz_sabari_ | 22:32 | |
| jcoufal | in tuskar-ui we are having very small patch which should pass gate tests without problem | 22:32 |
| jcoufal | https://review.openstack.org/#/c/141119/ | 22:32 |
| jcoufal | but all tests are failing on installing dependencies | 22:32 |
| jcoufal | that setup.py is missing | 22:32 |
| *** ChuckC has quit IRC | 22:32 | |
| *** KurtMartin has joined #openstack-infra | 22:33 | |
| jcoufal | I think that problem might be in latest horizon master tarball | 22:33 |
| *** lhcheng has joined #openstack-infra | 22:33 | |
| jcoufal | it looks corrupted | 22:33 |
| jcoufal | lhcheng: not sure though | 22:33 |
| *** pc_m has quit IRC | 22:34 | |
| fungi | jcoufal: yeah, taking a look now, but ideally horizon should be installed via git clone/pull rather than from tarballs.o.o | 22:34 |
| jogo | mtreinish: 403 haha | 22:34 |
| *** oomichi is now known as oomichi_cooking_ | 22:35 | |
| mtreinish | jogo: yeah there are no guarantees with that round robin mirror | 22:35 |
| fungi | jcoufal: branch tarball replacement is not (yet) atomic, so you may at times read it partway through being uploaded | 22:35 |
| jogo | mtreinish: we should have been caching it | 22:35 |
| jogo | maybe the caching logic borked | 22:35 |
| jcoufal | fungi: I see | 22:35 |
| fungi | jcoufal: and there's a local clone of horizon already on all our workers | 22:35 |
| fungi | jcoufal: though i guess you want to have horizon pip installed from your project's requirements list even though it's not published to pypi | 22:36 |
| mtreinish | jogo: look at the scrollback | 22:36 |
| *** timcline_ has joined #openstack-infra | 22:36 | |
| mtreinish | we added a new image | 22:36 |
| *** kmartin has quit IRC | 22:36 | |
| jcoufal | fungi: it sounds about right | 22:36 |
| *** shashankhegde has joined #openstack-infra | 22:37 | |
| fungi | jcoufal: ideally this gets fixed with the horizon-lib split | 22:37 |
| jcoufal | fungi: the latest news were that it might not happen | 22:37 |
| jcoufal | since there is intention to rewrite all the templates into angular directives | 22:38 |
| fungi | jcoufal: gzip: horizon-master.tar.gz: invalid compressed data--crc error | 22:38 |
| fungi | jcoufal: your theory seems to be valid | 22:38 |
| jcoufal | fungi: how can I get the latest tarball rebuilt? | 22:38 |
| *** stevemar has joined #openstack-infra | 22:38 | |
| jogo | mtreinish: ahhh that makes sense | 22:38 |
| fungi | jcoufal: i can retrigger the most recent post job for it | 22:38 |
| fungi | just a few | 22:38 |
| jcoufal | fungi: thank you | 22:39 |
| *** timcline has quit IRC | 22:39 | |
| ZZelle | fungi, hi | 22:39 |
| ZZelle | fungi, one question about https://wiki.openstack.org/wiki/Vulnerability_Management#Downstream_stakeholders | 22:39 |
| *** zz_sabari_ is now known as sabari | 22:39 | |
| fungi | jcoufal: though if 136539,2 merges in ~17 minutes it's going to self-correct anyway. mind if i wait? | 22:39 |
| jcoufal | fungi: shouldn't be a problem | 22:39 |
| jcoufal | fungi: thanks for looking at this | 22:39 |
| fungi | ZZelle: ask away | 22:40 |
| openstackgerrit | Clark Boylan proposed openstack-infra/system-config: Install all puppet modules from source https://review.openstack.org/141182 | 22:40 |
| ZZelle | fungi, what should we provide as a "rationale" to become a member of VMT? | 22:40 |
| clarkb | I think that will fix some of the errors | 22:40 |
| *** timcline_ has quit IRC | 22:40 | |
| *** yamamoto has joined #openstack-infra | 22:40 | |
| *** penguinRaider__ has joined #openstack-infra | 22:41 | |
| *** zz_dimtruck is now known as dimtruck | 22:41 | |
| fungi | ZZelle: basically if your organization redistributes parts of openstack and could benefit from a few days advance warning/access to embargoed security bug fixes for testing and preparation purposes, that's really all we need to know | 22:42 |
| fungi | ZZelle: oh, to become a member of the vmt or to become a downstream stakeholder? | 22:42 |
| fungi | ZZelle: those are two very different things | 22:42 |
| ZZelle | fungi, sorry, i was ambiguous, a downstream stakeholder | 22:42 |
| fungi | ZZelle: then, yeah, what i said first | 22:43 |
| clarkb | so new hypothesis | 22:43 |
| clarkb | new puppet release has broken us | 22:44 |
| fungi | ZZelle: e-mail me and cc the other three listed at http://www.openstack.org/projects/openstack-security/ and we'll discuss | 22:44 |
| *** achanda has quit IRC | 22:44 | |
| *** penguinRaider has quit IRC | 22:45 | |
| *** yamamoto has quit IRC | 22:45 | |
| ZZelle | fungi, is it fine if we create an enterprise mailing list and requst to add it as a downstream stakeholder or each physical person should register? | 22:45 |
| *** gyee has quit IRC | 22:45 | |
| ZZelle | fungi, oh, fine for me | 22:45 |
| jhesketh | fungi: those log links that you refer to have always been like that, I haven't changed anything for swift | 22:45 |
| jhesketh | ironically those links will work once swift is turned on | 22:46 |
| fungi | ZZelle: we can go over the details in e-mail. we discourage broad distribution of the messages we send, but don't really have control over whether the address you have us include is an alias going to a couple of stakeholders in your organization or just you | 22:46 |
| fungi | jhesketh: oh, good by me then | 22:46 |
| ZZelle | fungi, ok, let continue through mailing :) | 22:46 |
| *** achanda_ has joined #openstack-infra | 22:47 | |
| clarkb | ok that fixed precise | 22:47 |
| clarkb | but centos is still hella broken | 22:47 |
| *** shashankhegde has quit IRC | 22:48 | |
| fungi | clarkb: would it help to stab at undoing the recent module additions and see what happens? like start by backing out puppet-httpd from the list (i think it was the most recent to be tacked on) | 22:48 |
| clarkb | fungi: no nibalizer thinks its vcsrepo now. I have no idea why | 22:48 |
| *** MarkAtwood has quit IRC | 22:48 | |
| fungi | oh, right, we did split that out fairly recently | 22:49 |
| *** dmsimard is now known as dmsimard_away | 22:49 | |
| fungi | wait, no | 22:49 |
| fungi | we've always had our fork of that separate | 22:49 |
| *** dims__ has quit IRC | 22:49 | |
| fungi | did we add something which is depending on $other vcsrepo? | 22:49 |
| *** eharney has quit IRC | 22:49 | |
| *** dims__ has joined #openstack-infra | 22:50 | |
| *** markmcclain has quit IRC | 22:50 | |
| *** oomichi_cooking_ has quit IRC | 22:51 | |
| *** dimtruck is now known as zz_dimtruck | 22:51 | |
| *** achanda_ has quit IRC | 22:51 | |
| *** andreykurilin_ has joined #openstack-infra | 22:51 | |
| *** achanda has joined #openstack-infra | 22:51 | |
| *** MarkAtwood has joined #openstack-infra | 22:52 | |
| *** Masahiro has joined #openstack-infra | 22:52 | |
| *** penguinRaider__ has quit IRC | 22:52 | |
| *** penguinRaider has joined #openstack-infra | 22:52 | |
| clarkb | fungi: I don't think so | 22:53 |
| *** baoli has quit IRC | 22:53 | |
| clarkb | I am going to hold a node | 22:54 |
| fungi | went over the log very carefully and i see no signs that it's doing anything out of the ordinary | 22:54 |
| *** baoli has joined #openstack-infra | 22:54 | |
| fungi | great idea | 22:54 |
| *** dims__ has quit IRC | 22:54 | |
| fungi | i don't see any reason why this would be different on centos | 22:54 |
| *** achanda has quit IRC | 22:56 | |
| *** shashankhegde has joined #openstack-infra | 22:56 | |
| *** Masahiro has quit IRC | 22:57 | |
| *** amitgandhinz has joined #openstack-infra | 22:58 | |
| *** jamespage_ has joined #openstack-infra | 22:58 | |
| *** stevemar has quit IRC | 23:00 | |
| *** jamespage_ has quit IRC | 23:01 | |
| *** jcoufal has quit IRC | 23:02 | |
| *** MarkAtwood has quit IRC | 23:03 | |
| *** MarkAtwood has joined #openstack-infra | 23:05 | |
| dougwig | hi infra, could someone help me sort out this failure? adding the same check/gate jobs as neutron to the new neutron-service repos. seems to be checking if haproxy is available for centos and bombing. https://review.openstack.org/#/c/141122/ | 23:06 |
| pleia2 | dougwig: centos job is broken at the moment, fix is in the works | 23:07 |
| fungi | dougwig: clarkb is hacking on a solution to that right now | 23:07 |
| clarkb | this is fun. I have a minimal reproduction case | 23:07 |
| fungi | oh? | 23:07 |
| *** dannywil_ has quit IRC | 23:08 | |
| *** patrickeast has quit IRC | 23:08 | |
| clarkb | vcsrepo { '/home/clarkb/derp': provider => git, source => 'https://git.openstack.org/openstack-infra/project-config' } | 23:08 |
| clarkb | puppet apply that and its breaks | 23:08 |
| clarkb | it says the provider parameter is invalid | 23:08 |
| dougwig | pleia2, fungi, clarkb - thank you | 23:08 |
| clarkb | also our modulepath is not set properly on centos6 either | 23:08 |
| clarkb | but thats fine because the test hard sets it | 23:08 |
| *** penguinRaider__ has joined #openstack-infra | 23:08 | |
| *** otter768 has joined #openstack-infra | 23:09 | |
| *** dannywilson has joined #openstack-infra | 23:09 | |
| *** dannywilson has quit IRC | 23:10 | |
| *** patrickeast has joined #openstack-infra | 23:10 | |
| fungi | logstash says this error was first seen yesterday in patchset 2 of https://review.openstack.org/140738 | 23:10 |
| *** dannywilson has joined #openstack-infra | 23:11 | |
| fungi | 2014-12-10 15:53:04 | 23:11 |
| mtreinish | fungi: did anything new land around then? | 23:11 |
| fungi | mtreinish: that's what i'm checking now | 23:11 |
| fungi | oh, you know what? new worker images were coming online at that time | 23:11 |
| *** penguinRaider has quit IRC | 23:12 | |
| fungi | that might have been the first puppet apply job to run on yesterday's centos6 images | 23:12 |
| *** thedodd has joined #openstack-infra | 23:12 | |
| *** otter768 has quit IRC | 23:13 | |
| clarkb | ya I think it may end up being puppet pr ruby updates on centos | 23:13 |
| mtreinish | ah, that could be it | 23:13 |
| clarkb | however building the new git nodes totally worked | 23:13 |
| *** oomichi_cooking_ has joined #openstack-infra | 23:13 | |
| fungi | definitely no suspect merges to system-config or project-config around that time | 23:14 |
| *** andreaf has quit IRC | 23:14 | |
| *** oomichi_cooking_ has quit IRC | 23:14 | |
| clarkb | also this doesn't break on precise | 23:14 |
| fungi | lending further support to that theory | 23:14 |
| *** ChuckC has joined #openstack-infra | 23:15 | |
| jeblair | clarkb, fungi: as best as i can tell from the tcpdump and logs: geard had a > 30 second backlog in processing. evidenced by the fact that work status packets were arriving in 2's and then 3's (there should be only 1 every 10 seconds for any given job) | 23:15 |
| jeblair | clarkb, fungi: this might be because it stopped processing for a time, or network traffic was interrupted for a time | 23:15 |
| *** pradk has quit IRC | 23:15 | |
| jeblair | clarkb, fungi: i'm having trouble causing it to be busy enough to generate that kind of backlog under normal circumstances | 23:16 |
| *** sabari is now known as zz_sabari | 23:16 | |
| *** Rockyg has quit IRC | 23:16 | |
| *** wznoinsk_ has joined #openstack-infra | 23:17 | |
| fungi | jcoufal has disappeared on us, but his horizon tarball problem is resolved | 23:18 |
| *** wznoinsk has quit IRC | 23:19 | |
| *** luqas__ has quit IRC | 23:19 | |
| *** luqas__ has joined #openstack-infra | 23:19 | |
| fungi | clouds? | 23:19 |
| *** lttrl has quit IRC | 23:20 | |
| fungi | (could the entire virtual machine pausing for a minute do this?) | 23:20 |
| *** mfink has quit IRC | 23:22 | |
| *** tjones1 has quit IRC | 23:22 | |
| jeblair | it could | 23:22 |
| *** rlandy has quit IRC | 23:22 | |
| fungi | i definitely don't think we can rule out arbitrary resources (including the processor itself) stopping for some extended period and then resuming | 23:23 |
| fungi | hrm. no "we helpfully live-migrated your instance to another compute host" tickets from rax | 23:24 |
| fungi | no issues reported on status.rackspace.com for a while either | 23:25 |
| clarkb | https://github.com/puppetlabs/puppet/pull/2963 is the problem | 23:25 |
| clarkb | I have confirmed on held node by applying that patch and running my test case | 23:25 |
| fungi | bam! | 23:25 |
| clarkb | I am a few blocks from puppetlabs and about to head over there | 23:25 |
| clarkb | because this is ridiculous | 23:26 |
| fungi | bwahahahaha | 23:26 |
| *** dkranz has quit IRC | 23:26 | |
| openstackgerrit | Steven Weston proposed openstack-infra/system-config: Adds puppet system-config module split script https://review.openstack.org/137991 | 23:26 |
| mordred | mikal: you could just turn your phone off | 23:26 |
| clarkb | mordred: jeblair ^ if you need ditch puppet fuel | 23:26 |
| clarkb | I think that is like nuclear bomb levels of fuel | 23:26 |
| sweston | ^ asselin get it while it's hot :-) | 23:27 |
| *** mfink has joined #openstack-infra | 23:27 | |
| fungi | The Puppet Labs packages are built against the "system" ruby for a specific distribution. On Debian/Ubuntu, vendor_ruby is the appropriate place to look for a packaged Puppet installation. | 23:27 |
| mikal | mordred: It wasn't just me | 23:27 |
| mikal | Only one call was me | 23:27 |
| * asselin looking | 23:27 | |
| mordred | clarkb: I'm hoping to have it all replaced in a few weeks | 23:27 |
| mordred | clarkb: :) | 23:27 |
| *** jedimike has quit IRC | 23:27 | |
| *** penguinRaider__ has quit IRC | 23:27 | |
| *** penguinRaider has joined #openstack-infra | 23:27 | |
| clarkb | the killer is it won't be fixed until puppet 4 | 23:28 |
| mtreinish | sigh... | 23:28 |
| sweston | it appears I have angered the white space gods once again ;-) | 23:28 |
| fungi | sweston: red mark of shame | 23:28 |
| sweston | fungi: ;-) | 23:29 |
| clarkb | so I propose we pin puppet on centos6 for now | 23:29 |
| * sweston puts on the shame hat and stands in the corner | 23:29 | |
| mordred | clarkb: s/on centos6// | 23:29 |
| mtreinish | clarkb: or just abandon it on centos... | 23:29 |
| clarkb | mordred: no its only a centos6 problem | 23:29 |
| fungi | sweston: join me in renouncing the whitespace gods and the depths of insanity they represent | 23:29 |
| clarkb | ok nibalizer says that may not fix it | 23:30 |
| jeblair | fungi, clarkb: so i have 1000 workers each sending status updates to my client every 10 seconds, then i have it submit 920 jobs that won't run to simulate a big batch of merger or set_description jobs. | 23:30 |
| sweston | fungi: +100000 | 23:30 |
| nibalizer | idunno exactly | 23:30 |
| nibalizer | more testing needed | 23:30 |
| mordred | clarkb: yah - but if we pin, we should just pin across the board | 23:30 |
| clarkb | mordred: oh ya ++ | 23:30 |
| mordred | clarkb: so that we're consistently using the same version | 23:30 |
| jeblair | fungi, clarkb: that makes it pretty busy, but it doesn't backlog. submit job time is still 0.003 seconds | 23:30 |
| clarkb | so I need to do more testing | 23:30 |
| asselin | fungi, what would happen if you silently changed gerrit to not show whitespace errors? | 23:30 |
| clarkb | jeblair: maybe if you mix in some admin statuses? | 23:31 |
| clarkb | jeblair: since nodepool would be doining that at a fairly constant rate | 23:31 |
| jeblair | clarkb: i did a few of those by hand too... at a greatly increased size (30,000 registered functions) | 23:31 |
| sweston | asselin: I think gerrit may come after him | 23:31 |
| mtreinish | asselin: if it's python code the pe8 job would still explode (or bashate on bash) | 23:31 |
| fungi | asselin: it could catalyze a revolt, or i might just be martyred... not sure i want to find out yet | 23:31 |
| asselin | :) | 23:32 |
| jeblair | clarkb: if i time it just right and do an admin status request during the submit job, it takes 0.3 seconds | 23:33 |
| mordred | jeblair: that's 100x slower | 23:34 |
| *** thedodd has quit IRC | 23:34 | |
| jeblair | mordred: yes it is. but that only happens every 10 seconds | 23:34 |
| openstackgerrit | Steven Weston proposed openstack-infra/system-config: Adds puppet system-config module split script https://review.openstack.org/137991 | 23:34 |
| jeblair | mordred: and it's 100x faster than the observed behavior we're troubleshooting | 23:35 |
| *** asettle has joined #openstack-infra | 23:35 | |
| jeblair | i'm going to break for a bit, but i'm starting to think that the right answer is 'raise the timeout'. | 23:36 |
| jeblair | and maybe revisit that if clarb's improvements end up giving us an order of magnitude or more performance increase | 23:37 |
| nibalizer | okay it looks like we can change the puppet.conf to disable dircetory environments | 23:37 |
| clarkb | ok nibalizer and I have a workaround | 23:37 |
| nibalizer | and that fixes it | 23:37 |
| mordred | can you explain "Disable directory environments" more? | 23:37 |
| fungi | so this doesn't involve dragging a lart over to the puppetlabs offices and getting some much-needed exercise? | 23:38 |
| mtreinish | mordred: did you ever push an explode release with that typo fix? | 23:38 |
| mordred | mtreinish: nope! lemme do that | 23:39 |
| clarkb | mordred: fungi: those things we have for puppet envs | 23:39 |
| clarkb | mordred: fungi: we stop doing that and go back to the old thing | 23:39 |
| clarkb | which actually makes me :( | 23:39 |
| *** kgiusti has quit IRC | 23:39 | |
| mordred | so... | 23:39 |
| clarkb | I do want to test older 3.7.x on centos6 though | 23:39 |
| mordred | I do not understand what the new thing is | 23:39 |
| mordred | and I didnt' understand it when it went into the config | 23:40 |
| clarkb | mordred: its puppets new way of doing environments | 23:40 |
| mordred | so if you're saying "we're just going to go back to a directory==environment world" then I'm fine with it | 23:40 |
| clarkb | mordred: old way to be removed in 4 iirc | 23:40 |
| clarkb | so this was sort of future proofing except the future is broken | 23:40 |
| mordred | clarkb: when is the last time that puppet's new way of doing something made our lives better? | 23:40 |
| clarkb | however this is only an issue where you compile the catalog | 23:40 |
| fungi | the future is _always_ broken | 23:41 |
| clarkb | so our production centos servers are fine because trusty master is like "lol it works" | 23:41 |
| clarkb | but puppet apply in our tests compiles on centos6 and fails | 23:41 |
| *** penguinRaider__ has joined #openstack-infra | 23:41 | |
| mordred | clarkb: I quite literally cannot even snark on that | 23:41 |
| mordred | clarkb: it's so stupid | 23:41 |
| *** sarob has quit IRC | 23:41 | |
| clarkb | mordred: so you don't understand | 23:41 |
| clarkb | I am literally a 5 minute walk from their offices | 23:41 |
| mordred | yah | 23:41 |
| clarkb | staying in my current seat is very difficult right now | 23:41 |
| mtreinish | clarkb: do you want to borrow my rifle? | 23:41 |
| clarkb | and I probably need to not try and fix this asap so that I can be reasonable about the solution | 23:42 |
| mordred | clarkb: listen, if you go over there, please tell them that I think they are poo-poo heads | 23:42 |
| nibalizer | we could have directory envornments on for the puppet master but off for the test nodes | 23:42 |
| clarkb | does anyone know how to make yum downgrade a package? | 23:42 |
| clarkb | nibalizer: but then we aren't testing things well | 23:42 |
| *** zz_sabari is now known as sabari | 23:42 | |
| nibalizer | because the test nodes run puppet apply | 23:42 |
| fungi | clarkb: i think you just yum install and provide the target version, no? | 23:43 |
| mtreinish | clarkb: there is a --downgrade or something like that | 23:43 |
| mtreinish | actually I think it's yum downgrade package | 23:43 |
| mordred | well, if I can ever get past the dib work, everything should be running puppet apply | 23:43 |
| jeblair | mordred: why does that depend on dib? | 23:43 |
| fungi | clarkb: aha... yum downgrade puppet-1.2.3-4.el5 | 23:43 |
| mtreinish | clarkb: err, nm it's changed in centos6: https://access.redhat.com/node/64069 | 23:43 |
| mordred | jeblair: dib is occupying the top slot on the stack | 23:44 |
| nibalizer | mordred: so how do you test mordredtestenviornment in a puppet apply world? | 23:44 |
| mordred | jeblair: but also, there is a piece of the launch-node thing that will be made WAY saner once we have sane base images | 23:44 |
| mordred | jeblair: so it's also a bit of a rabbithole | 23:44 |
| *** penguinRaider has quit IRC | 23:44 | |
| fungi | clarkb: or what mtreinish linked looks useful, since you can undo an upgrade rather than explicitly downgrade | 23:44 |
| nibalizer | so the problem is catalog compilation, which happens master side | 23:44 |
| *** MaxV has quit IRC | 23:45 | |
| jeblair | mordred: just offering you a rope up out of the rabbithole; up to you if you want to grab it ;) | 23:45 |
| mordred | mmm. rope | 23:45 |
| *** MaxV has joined #openstack-infra | 23:45 | |
| *** hashar has quit IRC | 23:45 | |
| *** sabeen1 has quit IRC | 23:45 | |
| mordred | jeblair: oh no! it got flooded away by the super-scary-storm which brought an inch of water | 23:45 |
| clarkb | downgrading does not fix the problem | 23:46 |
| clarkb | 3.7.x is all broken | 23:46 |
| mordred | WOOT! | 23:46 |
| mtreinish | clarkb: then what made it start failing? | 23:46 |
| jeblair | mordred: 2" in berkeley | 23:46 |
| mordred | jeblair: ok. I have made AND BOOTED both ubuntu and centos | 23:47 |
| clarkb | mtreinish: we switched to directory environments to fix puppet runs on our master | 23:47 |
| clarkb | it was a change nibalizer wrote | 23:47 |
| mtreinish | ah, ok | 23:47 |
| *** dims__ has joined #openstack-infra | 23:47 | |
| jeblair | mordred: 2" in pao too | 23:47 |
| fungi | clarkb: that was a while ago though, yeah? | 23:47 |
| nibalizer | directory envroenmetns came with puppet 3 actually | 23:47 |
| jeblair | mordred: it's literally twice as bad as you thought | 23:47 |
| mordred | jeblair: EEK. get the canoe! | 23:47 |
| clarkb | fungi: it was this week | 23:47 |
| fungi | clarkb: oh! | 23:48 |
| clarkb | but I need a break after derping on this | 23:48 |
| fungi | clarkb: and then maybe we didn't have updated images for a day or two after? | 23:48 |
| clarkb | too frustrated to computer | 23:48 |
| clarkb | fungi: ya | 23:48 |
| mordred | clarkb: I recommend beer | 23:48 |
| clarkb | yes | 23:48 |
| mordred | clarkb: also, go tell the folks at puppetlabs that I'm going to stop by and explain computers to them at some point | 23:48 |
| jeblair | mordred: that's what people used before phones, right? | 23:49 |
| fungi | computators | 23:49 |
| *** shakamunyi has quit IRC | 23:49 | |
| mordred | jeblair: did you see the new cloud-enabled typewriter? | 23:49 |
| jeblair | fungi: mm tasty | 23:49 |
| *** andreykurilin_ has quit IRC | 23:49 | |
| *** andreykurilin__ has joined #openstack-infra | 23:49 | |
| mordred | jeblair: http://hemingwrite.com/ | 23:50 |
| fungi | jeblair: computaters then if you prefer | 23:50 |
| *** MaxV has quit IRC | 23:50 | |
| fungi | built from computater chips | 23:50 |
| *** penguinRaider__ has quit IRC | 23:50 | |
| *** penguinRaider has joined #openstack-infra | 23:50 | |
| mtreinish | mordred: I would totally want one of those if it had a model m's buckling spring for the keys | 23:50 |
| jeblair | mordred: wow. i don't even. hopefully soon someone will hack it to run android apps and you can facebook from it | 23:51 |
| mtreinish | but, I'll just stick with my selectric II | 23:51 |
| fungi | mtreinish: can you hook your selectric ii up to the faceblocks and the tweeters? | 23:56 |
| mtreinish | fungi: heh, probably not. But if I had a few months I'm sure I could figure out something | 23:57 |
| mtreinish | unless they would let me us the postal service to submit things... :) | 23:57 |
| *** ddieterly has quit IRC | 23:58 | |
| fungi | nice. ip-over-postal-carrier | 23:58 |
| *** sarob has joined #openstack-infra | 23:58 | |
| fungi | a lot more laggy than ip-over-carrier-pigeon but (barely) more reliable | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!