| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add skipped CRD tests https://review.openstack.org/531887 | 00:08 |
|---|---|---|
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: WIP: Support cross-source dependencies https://review.openstack.org/530806 | 00:08 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: WIP: Add cross-source tests https://review.openstack.org/532699 | 00:08 |
| corvus | the first test in that passes ^ (a Gerrit change that depends on a Github change in a shared queue in a gate pipeline). that's a milestone. :) | 00:09 |
| mordred | corvus: \o/ | 00:10 |
| SpamapS | daaaaaaaaaamn | 00:29 |
| SpamapS | that's hot | 00:29 |
| *** JasonCL has joined #zuul | 00:55 | |
| *** threestrands has quit IRC | 01:03 | |
| *** JasonCL has quit IRC | 01:35 | |
| *** JasonCL has joined #zuul | 01:41 | |
| pabelanger | just a heads up, https://review.openstack.org/532575/ was applied to ze04.o.o tonight, and seems server was rebooted due to live migration and went live. so we are starting ze04 for now until we can land rest of patches | 02:14 |
| Shrews | https://review.openstack.org/532709 and https://review.openstack.org/532594 are the rest of the patches, fwiw | 02:16 |
| *** threestrands has joined #zuul | 02:47 | |
| *** threestrands has quit IRC | 02:47 | |
| *** threestrands has joined #zuul | 02:47 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul feature/zuulv3: scheduler: better handle format status error https://review.openstack.org/532718 | 03:10 |
| dmsimard | tristanC: added a comment on ^ | 03:18 |
| dmsimard | good night | 03:19 |
| tristanC | dmsimard: thanks, good catch | 03:20 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul feature/zuulv3: scheduler: better handle format status error https://review.openstack.org/532718 | 03:21 |
| *** harlowja_ has quit IRC | 03:21 | |
| *** rlandy|bbl is now known as rlandy | 03:50 | |
| *** bhavik1 has joined #zuul | 04:35 | |
| *** rlandy has quit IRC | 04:53 | |
| SpamapS | hrm | 05:21 |
| SpamapS | have we considered the dangers of letting untested patches add irrelevant-files stanzas? | 05:21 |
| SpamapS | or files for that matter | 05:22 |
| *** harlowja has joined #zuul | 05:29 | |
| tobiash | SpamapS: what dangers do you have in mind? | 05:37 |
| SpamapS | tobiash: https://www.regular-expressions.info/catastrophic.html | 05:39 |
| * SpamapS is poking at his stage zuul now | 05:39 | |
| SpamapS | and actually.. afk'ing | 05:39 |
| SpamapS | first test did not cause much issue | 05:40 |
| *** bhavik1 has quit IRC | 05:42 | |
| *** xinliang has quit IRC | 05:56 | |
| *** xinliang has joined #zuul | 06:08 | |
| *** xinliang has quit IRC | 06:08 | |
| *** xinliang has joined #zuul | 06:08 | |
| *** harlowja has quit IRC | 06:47 | |
| *** threestrands has quit IRC | 07:00 | |
| *** threestrands has joined #zuul | 07:01 | |
| *** threestrands has quit IRC | 07:01 | |
| *** threestrands has joined #zuul | 07:01 | |
| *** threestrands has quit IRC | 07:02 | |
| *** threestrands has joined #zuul | 07:03 | |
| *** threestrands has quit IRC | 07:03 | |
| *** threestrands has joined #zuul | 07:03 | |
| *** threestrands has quit IRC | 07:04 | |
| *** threestrands has joined #zuul | 07:04 | |
| *** threestrands has quit IRC | 08:08 | |
| *** jpena|off is now known as jpena | 08:34 | |
| *** saop has joined #zuul | 08:56 | |
| *** ankkumar has joined #zuul | 08:57 | |
| saop | hello | 08:57 |
| saop | I am trying to build openstack/ironic CI using zuul v3 | 08:57 |
| *** hashar has joined #zuul | 08:57 | |
| saop | Can anyone tell me, how would it would read my zuul.yaml customized file, because openstack/ironic repo has there one? | 08:58 |
| saop | How would my gate will run, I mean using which zuul yaml files? | 08:58 |
| *** JasonCL has quit IRC | 11:11 | |
| *** jkilpatr has joined #zuul | 12:04 | |
| *** jpena is now known as jpena|lunch | 12:04 | |
| SpamapS | saop: Have you read this page? https://docs.openstack.org/infra/zuul/feature/zuulv3/user/config.html | 12:07 |
| *** JasonCL has joined #zuul | 12:25 | |
| *** JasonCL has quit IRC | 12:30 | |
| *** saop has quit IRC | 12:48 | |
| *** haint has joined #zuul | 12:48 | |
| *** ankkumar has quit IRC | 12:59 | |
| *** jpena|lunch is now known as jpena | 13:00 | |
| *** rlandy has joined #zuul | 13:36 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Improve logging around ZooKeeper suspension https://review.openstack.org/532823 | 13:48 |
| *** dkranz has joined #zuul | 13:55 | |
| dmsimard | mordred: so the role path thing ? I looked and I'm probably even more confused now | 13:57 |
| mordred | dmsimard: right? | 14:35 |
| dmsimard | mordred: I kinda want to run with --keep so I can understand what is going on in your review | 14:38 |
| dmsimard | but I'm not super familiar with --keep and how it works .. I guess we can't decide to keep a particular job/review/build, it's an all or nothing deal | 14:38 |
| *** jkilpatr has quit IRC | 14:39 | |
| mordred | dmsimard: yah - that's right. although it might be a nice feature to consider being able to run with keep for a particular job/review/build - kind of like how hold works for nodes | 14:43 |
| dmsimard | mordred: yeah, that'd be useful | 14:43 |
| dmsimard | because now we need to hit --keep on all 10 executors | 14:43 |
| mordred | dmsimard: or maybe tie it together - whenever you hold a node for a job it sets keep for the ansible dirs too | 14:43 |
| dmsimard | doesn't make a whole lot of sense | 14:43 |
| mordred | that way you have the all the pieces of the puzzle when you need to debug a thing | 14:44 |
| pabelanger | seems like a good idea | 14:44 |
| *** jkilpatr has joined #zuul | 14:45 | |
| *** JasonCL has joined #zuul | 14:51 | |
| dmsimard | How would that work ? I don't think we want that to be a zuul job parameter (since it could lead to executors running out of space) -- maybe autohold could autokeep the builds ? (and then delete the builds once we delete the hold) | 15:11 |
| *** jkilpatr has quit IRC | 15:13 | |
| *** jkilpatr has joined #zuul | 15:13 | |
| *** jkilpatr has quit IRC | 15:13 | |
| *** jkilpatr has joined #zuul | 15:13 | |
| *** hashar is now known as hasharAway | 15:19 | |
| *** sshnaidm|afk is now known as sshnaidm | 15:21 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Log request ID on request failure https://review.openstack.org/532857 | 15:23 |
| *** JasonCL has quit IRC | 16:07 | |
| *** JasonCL has joined #zuul | 16:55 | |
| *** electrofelix has left #zuul | 17:16 | |
| *** electrofelix has joined #zuul | 17:26 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Delete the pool thread when a provider is removed https://review.openstack.org/532931 | 18:13 |
| *** weshay is now known as weshay_interview | 18:14 | |
| Shrews | clarkb: i think ^^ might be the cause of the log thing you were seeing earlier | 18:15 |
| Shrews | or at least, one possible cause | 18:16 |
| Shrews | this rabbit hole goes deep | 18:16 |
| *** jpena is now known as jpena|off | 18:16 | |
| corvus | tobiash, SpamapS, jlk: how important is cachecontrol? can we replace it? | 18:20 |
| jlk | in what manner? | 18:20 |
| corvus | it depends on msgpack which seems to be having a lot of problems with releases lately, and updating it causes python segfaults | 18:20 |
| jlk | We definitely want to honor the etags that we get from github API, to limit the API hit count | 18:21 |
| jlk | that's... ugly :( | 18:21 |
| corvus | jlk: as in, is there an alternative to using that library | 18:21 |
| tobiash | corvus: cachecontrol saves a lot of requests regarding the request limit | 18:21 |
| corvus | jlk: yeah, the msgpack thing is kind of a tire fire at the moment that i'd like to stay away from until they figure it out :| | 18:21 |
| jlk | The specific use of cachecontrol I don't think is critical. Having something that makes use of etags to save API count IS critical | 18:21 |
| jlk | so alternatives are totally fine | 18:21 |
| corvus | see https://pypi.python.org/pypi/msgpack/0.5.1 for an example of what's going on | 18:21 |
| corvus | and https://github.com/msgpack/msgpack-python/issues/268 and https://github.com/msgpack/msgpack-python/issues/266 | 18:22 |
| corvus | oh fungi points out that it's the latest release that uses it | 18:22 |
| corvus | 18:21 < fungi> introduced in 0.12.0, while 0.11.2 and earlier don't use it | 18:23 |
| corvus | would it be possible to pin to 11.2? | 18:23 |
| corvus | that's very old though -- 2015-03-19 | 18:24 |
| fungi | yjere | 18:24 |
| fungi | there's also this pr: https://github.com/ionrock/cachecontrol/pull/161 | 18:24 |
| fungi | to make msgpack optional | 18:24 |
| fungi | i'm mostly looking to see if anyone has pointed out to cachecontrol upstream yet that their msgpack-python dep needs pinning or updating | 18:24 |
| fungi | cachecontrol seems pretty slow-moving though. no merged commits since august | 18:25 |
| clarkb | fungi: if they depend on msgpack-python instead of msgpack won't things just work? | 18:27 |
| clarkb | or is there some magic that causes it to be rewriten to a msgpack install? | 18:27 |
| fungi | other way around i think? | 18:27 |
| fungi | msgpack seems to be the new one | 18:27 |
| clarkb | msgpack-python is the old name | 18:28 |
| clarkb | msgpack is the new one | 18:28 |
| clarkb | so if they stick with the old one wouldn't it just continue working? | 18:28 |
| fungi | stick with the old name? | 18:28 |
| clarkb | yes then there is no transition that breaks users | 18:28 |
| fungi | or stick with (pin to) an old release | 18:28 |
| clarkb | (and upstream seems to be willing to publish under the old name) | 18:28 |
| fungi | they uploaded msgpack-python 0.5.0 as a transitional package to drag in the new msgpack package | 18:29 |
| fungi | which seems to be when all of this began | 18:29 |
| clarkb | oh gotcha | 18:29 |
| fungi | oh, you mean if msgpack upstream switched back to using msgpack-python as the proper name of their package | 18:29 |
| clarkb | I didn't realize msgpack-python 0.5.0 was just a redirect | 18:30 |
| clarkb | I thought cachecontrol had updated to msgpack | 18:30 |
| fungi | yeah, cachecontrol depended on msgpack-python which is now apparently a (buggy) transition to msgpack | 18:30 |
| fungi | so unless msgpack-python reverts to a proper package again, cachecontrol probably needs a new release changing their dep to msgpack | 18:31 |
| pabelanger | fungi: optional msgpack seems like a good thing | 18:32 |
| corvus | msgpack 0.5.1 appears to be a real package | 18:36 |
| corvus | grr | 18:36 |
| corvus | msgpack-python 0.5.1 appears to be a real package | 18:36 |
| corvus | the last release of msgpack-python before 0.5.0 was 0.4.8 in july. that's before we went into production, so if this happens on every upgrade, not just crazy transitional upgrades, we won't have noticed. | 18:40 |
| corvus | i can reproduce locally | 18:44 |
| corvus | i had 0.4.8 installed | 18:44 |
| corvus | then i ran >>> import msgpack | 18:44 |
| corvus | >>> msgpack.packb([1,2,3]) | 18:44 |
| corvus | then i installed 0.5.0 and ran the second line again and it segfaulted | 18:45 |
| corvus | trying more combos now | 18:45 |
| pabelanger | ah, great. I've been trying to get github.com setup myself | 18:45 |
| SpamapS | so can we just pin msgpack? | 18:50 |
| * SpamapS missed the whole story I think | 18:50 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Capture and report errors in sibling installation https://review.openstack.org/532216 | 18:51 |
| clarkb | we could use constraints for that | 18:51 |
| clarkb | not sure its a good long term solution but for short term is probably doable | 18:51 |
| jlk | yeah I'd agree with the short term strategy | 18:54 |
| jlk | long term, maybe an alternative to cachecontrol in general, or start throwing more commits at that upstream too... | 18:54 |
| corvus | okay this is looking promising: http://paste.openstack.org/show/642996/ | 18:58 |
| corvus | i think the main issue is the crazy transitional package | 18:58 |
| SpamapS | so can we do a "!0.5.0" ? | 18:58 |
| corvus | because msgpack-python 0.5.0 was an empty transitional package. so the transition from 0.4.8 -> 0.5.0 caused a segfault. and the transition from 0.5.0 to 0.5.1 (which was, once again, a real package) caused a segfault. but a transition from 0.4.8 to 0.5.1 does not. | 18:59 |
| corvus | SpamapS: it's not a direct dep for us -- does that work or does that require constraints as well? | 18:59 |
| corvus | i'm going to try a few more upgrades, like 0.4.7 -> 0.4.8 | 19:00 |
| *** jkilpatr_ has joined #zuul | 19:00 | |
| *** jkilpatr has quit IRC | 19:01 | |
| SpamapS | Probably requires constraints. | 19:01 |
| SpamapS | But IIRC, you can direct-dep on it before the transitive dep with the ! and that works. | 19:02 |
| SpamapS | Since that will cause it to pull in 0.5.1 or 0.4.8, and then the dep is satisfied when it gets to cachecontrol. | 19:02 |
| corvus | 0.4.7 -> 0.4.8 works, and 0.4.7 -> 0.5.1 works | 19:04 |
| corvus | so i think once we're past this transition back to a real package, we'll be okay | 19:04 |
| corvus | i think i'm inclined to leave off the direct dep for now, since it's probably too late to help anyone at this point, and the situation, as best as i understad it, is resolved by the current release | 19:06 |
| *** jkilpatr_ has quit IRC | 19:07 | |
| clarkb | if you list it at top level that should be honored | 19:09 |
| clarkb | but constraints is the wins all option | 19:09 |
| fungi | sorry, had to step away, but yeah i think we can just move on in that case. we could try to declare that 0.5.0 is broken but it's not a direct dep of ours and nobody's legitimately going to attempt to upgrade to 0.5.0 now that 0.5.1 is latest | 19:11 |
| fungi | so, basically, agreeing with corvus | 19:12 |
| *** sshnaidm is now known as sshnaidm|afk | 19:15 | |
| *** weshay_interview is now known as weshay | 19:16 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Fix races around deleting a provider https://review.openstack.org/532931 | 19:16 |
| Shrews | awesome. tox has stopped working for me locally | 19:16 |
| Shrews | something something dbm.gnu missing | 19:16 |
| *** jkilpatr_ has joined #zuul | 19:21 | |
| clarkb | Shrews: are you on linux? | 19:35 |
| Shrews | yep. fedora | 19:36 |
| clarkb | any chance you ran with python2 then python3? | 19:36 |
| clarkb | there are two semi related issues with testr around that | 19:36 |
| clarkb | the first is on some platforms like OSX python lacks some dbs for licensing reasons | 19:36 |
| clarkb | the other is python2 and python3 had different default db implementations so if you python2 before python3 it breaks | 19:37 |
| clarkb | but if you python3 then python2 it owrks | 19:37 |
| Shrews | clarkb: i don't think so? i mean, i just 'tox -epy36 -r some.unit.test' and that fails | 19:41 |
| clarkb | Shrews: huh. I would try a tox -re py36 just to make sure virtualenv is healthy and delete or move .testrepository | 19:43 |
| *** harlowja has joined #zuul | 19:43 | |
| Shrews | clarkb: thx. rm'ing the dir worked | 19:44 |
| *** dtruong has joined #zuul | 19:54 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Short-circuit request handling on disable provider https://review.openstack.org/532957 | 19:57 |
| *** hasharAway is now known as hashar | 20:08 | |
| *** dkranz has quit IRC | 20:14 | |
| *** JasonCL has quit IRC | 20:22 | |
| *** dkranz has joined #zuul | 20:30 | |
| *** JasonCL has joined #zuul | 20:36 | |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Fix races around deleting a provider https://review.openstack.org/532931 | 20:49 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Short-circuit request handling on disable provider https://review.openstack.org/532957 | 20:49 |
| *** JasonCL has quit IRC | 21:13 | |
| *** JasonCL has joined #zuul | 21:23 | |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Improve logging around ZooKeeper suspension https://review.openstack.org/532823 | 21:26 |
| *** jkilpatr_ has quit IRC | 21:27 | |
| *** JasonCL has quit IRC | 21:33 | |
| *** JasonCL has joined #zuul | 21:49 | |
| *** jkilpatr has joined #zuul | 21:52 | |
| *** threestrands has joined #zuul | 21:58 | |
| *** threestrands has quit IRC | 21:58 | |
| *** threestrands has joined #zuul | 21:58 | |
| *** threestrands has quit IRC | 21:59 | |
| *** threestrands has joined #zuul | 21:59 | |
| *** threestrands has quit IRC | 21:59 | |
| *** threestrands has joined #zuul | 21:59 | |
| *** threestrands has quit IRC | 22:00 | |
| *** threestrands has joined #zuul | 22:01 | |
| *** threestrands has quit IRC | 22:02 | |
| *** dkranz has quit IRC | 22:02 | |
| *** threestrands has joined #zuul | 22:02 | |
| *** hashar has quit IRC | 22:24 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: WIP: Support cross-source dependencies https://review.openstack.org/530806 | 22:35 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: WIP: Add cross-source tests https://review.openstack.org/532699 | 22:35 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Delete stale jobdirs at startup https://review.openstack.org/531510 | 22:44 |
| *** rlandy is now known as rlandy|bbl | 22:52 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!