| openstackgerrit | Tristan Cacqueray proposed zuul/zuul master: Revert "Revert "web: upgrade react and react-scripts to ^2.0.0"" https://review.opendev.org/659991 | 00:59 |
|---|---|---|
| tristanC | mordred: i couldn't find the issue with the react bump, here is a vanilla revert of the revert ^ | 00:59 |
| mordred | tristanC: cool. hopefully it will show up in the build output | 00:59 |
| mordred | tristanC: I think it might have something to do with the data coming from the openstack status endpoint - when it was failing, it was only failing on openstack, not on opendev | 01:00 |
| tristanC | mordred: let's see, running it locally with start:openstack doesn't show issue atm | 01:03 |
| *** threestrands has joined #zuul | 02:34 | |
| *** bhavikdbavishi has joined #zuul | 03:35 | |
| *** swest has joined #zuul | 05:22 | |
| *** chkumar|pto is now known as chandankumar | 05:58 | |
| openstackgerrit | Merged zuul/zuul master: Get playbook data from extra_vars https://review.opendev.org/659802 | 06:20 |
| *** zxiiro has quit IRC | 06:47 | |
| *** jbryce has quit IRC | 06:47 | |
| *** tima has quit IRC | 06:47 | |
| *** Shrews has quit IRC | 06:47 | |
| *** tima has joined #zuul | 06:47 | |
| *** Shrews has joined #zuul | 06:47 | |
| *** jbryce has joined #zuul | 06:47 | |
| *** clayg has quit IRC | 06:47 | |
| *** clayg has joined #zuul | 06:47 | |
| *** jtanner has quit IRC | 06:48 | |
| *** johnsom has quit IRC | 06:48 | |
| *** mnasiadka has quit IRC | 06:48 | |
| *** jtanner has joined #zuul | 06:48 | |
| *** mnasiadka has joined #zuul | 06:48 | |
| *** johnsom has joined #zuul | 06:48 | |
| *** zxiiro has joined #zuul | 06:48 | |
| *** saneax has joined #zuul | 07:08 | |
| *** themroc has joined #zuul | 07:11 | |
| *** yolanda_ has quit IRC | 07:13 | |
| *** yolanda_ has joined #zuul | 07:20 | |
| *** threestrands has quit IRC | 07:23 | |
| *** pcaruana has joined #zuul | 07:23 | |
| *** bhavikdbavishi has quit IRC | 07:30 | |
| *** hashar has joined #zuul | 07:42 | |
| *** jpena|off is now known as jpena | 07:52 | |
| *** gtema has joined #zuul | 08:04 | |
| *** jangutter has joined #zuul | 08:18 | |
| *** bhavikdbavishi has joined #zuul | 08:21 | |
| *** gtema has quit IRC | 08:24 | |
| *** markwork has joined #zuul | 08:27 | |
| *** gtema has joined #zuul | 08:27 | |
| *** panda|rover|off is now known as panda|rover | 08:41 | |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: A reporter for Elasticsearch with the capability to index build and buildset results in an index. https://review.opendev.org/644927 | 09:32 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: A reporter for Elasticsearch https://review.opendev.org/644927 | 09:38 |
| openstackgerrit | Merged zuul/zuul-website master: Add slides for Spring 2019 Board Update https://review.opendev.org/656182 | 09:42 |
| *** gtema has quit IRC | 10:38 | |
| *** flaper87 has quit IRC | 10:42 | |
| *** panda|rover is now known as panda|lunch | 10:52 | |
| *** panda|lunch is now known as panda|rover|eat | 10:53 | |
| *** flaper87 has joined #zuul | 11:01 | |
| *** hashar has quit IRC | 11:03 | |
| *** yolanda_ has quit IRC | 11:04 | |
| *** hashar has joined #zuul | 11:07 | |
| *** yolanda_ has joined #zuul | 11:08 | |
| *** jpena is now known as jpena|lunch | 11:25 | |
| corvus | it looks like the image build jobs are broken again? | 11:44 |
| corvus | [0m[91mCould not open requirements file: [Errno 2] No such file or directory: '/output/zuul_base/requirements.txt' | 11:45 |
| *** gtema has joined #zuul | 11:49 | |
| openstackgerrit | Markus Hosch proposed zuul/zuul master: Evaluate CODEOWNERS settings during canMerge check https://review.opendev.org/644557 | 11:51 |
| AJaeger | zuul-jobs maintainer, what do you think about https://review.opendev.org/648658 ? | 11:56 |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Keep task stdout/stderr separate in result object https://review.opendev.org/650276 | 11:57 |
| *** panda|rover|eat is now known as panda|rover | 11:57 | |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Keep task stdout/stderr separate in result object https://review.opendev.org/650276 | 11:59 |
| corvus | oh, that's not a problem with the build job -- that's actually an error in that change: //review.opendev.org/659255 | 12:02 |
| corvus | er... sorry about the copy/paste but you get the idea | 12:02 |
| corvus | mordred: ^ i left a -1 comment on that | 12:06 |
| openstackgerrit | Christian Fetzer proposed zuul/zuul master: pass-to-parent: Fix passing multiple secrets to parent https://review.opendev.org/660105 | 12:16 |
| *** bhavikdbavishi has quit IRC | 12:20 | |
| *** jpena|lunch is now known as jpena | 12:22 | |
| *** bhavikdbavishi has joined #zuul | 12:23 | |
| *** rlandy has joined #zuul | 12:27 | |
| *** ianychoi has quit IRC | 12:28 | |
| *** ianychoi has joined #zuul | 12:29 | |
| *** markwork has quit IRC | 12:34 | |
| Shrews | so odd to see corvus active before, well, anyone | 12:35 |
| corvus | thanks for not "accidentally" hitting [enter] after the word 'active' :) | 12:42 |
| AJaeger | tristanC: did you see my comment on https://review.opendev.org/532083 ? Shouldn't we cap ansible-lint version like we do in requirements today? | 12:48 |
| AJaeger | corvus: did you see my question on the zuul list regarding version for ansible to run in linters? Right now we use 2.5 - should we update to 2.7? Do we have a policy for this already? | 12:49 |
| AJaeger | tristanC: could you look at https://review.opendev.org/#/c/632620/ ? There's an open comment waiting for your followup... | 12:51 |
| corvus | AJaeger: no, i'm not caught up on email; however, i'm concerned that ansible-lint is getting far too opinionated -- we discussed maybe we need to rethink that. (cc pabelanger) | 12:54 |
| AJaeger | corvus: ok, keep my WIP on the changes a bit longer until I hear feedback. | 12:55 |
| AJaeger | regarding ansnible-lint: yes, the recent changes make me wonder whether an uncapped version is the right thing... | 12:56 |
| fungi | well, if an uncapped version is not the right thing, then we should probably also reconsider our use of it overall | 12:56 |
| fungi | because we won't want to leave it on last week's version forever | 12:57 |
| fungi | capping and raising the cap are fine if we expect to want to follow the new rules they're adding and just need time to nondisruptively adjust all the code to conform to them | 12:57 |
| fungi | if we think they're going in a different (more opinionated) direction than is compatible with our projects, we should probably just replace those jobs with something like yamllint or even just a parser | 12:59 |
| AJaeger | good point | 12:59 |
| fungi | i wish more linters would follow the same model as lintian does, where rules are ranked by importance and you don't get the really opinionated stuff without passing --pedantic | 13:01 |
| fungi | i think bandit's going in the right direction, allowing you to configure it to only warn about violations of rules at or above a certain severity | 13:02 |
| *** bhavikdbavishi has quit IRC | 13:13 | |
| *** bhavikdbavishi has joined #zuul | 13:14 | |
| tristanC | AJaeger: thank you for the feedback, i'll update/comment the review | 13:19 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: add-build-sshkey: remove previously authorized build-sshkey https://review.opendev.org/632620 | 13:22 |
| *** bhavikdbavishi has quit IRC | 13:32 | |
| *** pcaruana has quit IRC | 13:33 | |
| *** gtema has quit IRC | 13:34 | |
| openstackgerrit | Christian Fetzer proposed zuul/zuul master: pass-to-parent: Fix passing multiple secrets to parent https://review.opendev.org/660105 | 13:34 |
| pabelanger | q: does zuul dashboard support regex for builds tab? | 13:37 |
| corvus | pabelanger: i don't think so, though i believe it shuold support multiple values | 13:41 |
| pabelanger | corvus: ack, thanks | 13:42 |
| corvus | (so you should be able to add two projects and see all the builds for both) | 13:43 |
| openstackgerrit | Flavio Percoco proposed zuul/zuul master: Fix typo in log message https://review.opendev.org/660118 | 13:43 |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Keep task stdout/stderr separate in result object https://review.opendev.org/650276 | 13:46 |
| corvus | Shrews: when you have a sec, can you look at https://review.opendev.org/658889 wrt the implications for nodepool handling that info? | 13:54 |
| Shrews | corvus: sure | 13:58 |
| corvus | Shrews, tobiash: i like that feature, but maybe we should think about the best way to pass that info to nodepool so it's not too zuul-specific (since i don't think it needs to be -- an external identifier for a request makes sense generically) | 14:00 |
| tobiash | corvus: you mean something more like meta.event_id? instead of meta.zuul_event_id? | 14:02 |
| tobiash | or something completely different? | 14:02 |
| Shrews | non-zuul specific name is a good idea. any reason the event ID is within a 'meta' dict and not a direct attribute? | 14:04 |
| Shrews | are you expecting more metadata? | 14:05 |
| tobiash | I was thinking that we might want to add more in the future | 14:05 |
| tobiash | but I also can easily remove the meta if you like | 14:06 |
| tobiash | meta just sounded good for stuff that's not needed for request processing and nodepool might or might not care about | 14:07 |
| Shrews | i mean, it's either relevant for the request or isn't. we have metadata on node objects b/c of the various drivers and whatnot, so it makes sense there. it seems like a level of abstraction not needed for a request. | 14:08 |
| Shrews | if nodepool doesn't care about it, why send it? | 14:08 |
| Shrews | nothing else is going to process it | 14:09 |
| Shrews | maybe i'm alone in that line of thinking though | 14:09 |
| tobiash | Shrews: the idea is to send it and if it exists nodepool can annotate the logs with it | 14:11 |
| tobiash | but the idea is to keep that optional | 14:11 |
| Shrews | right. that doesn't explain why it can't be its own attribute | 14:11 |
| tobiash | sure it can, I'm completely fine with both options | 14:12 |
| Shrews | nodepool can just as easily say request.get('event_id') or request.meta.get('event_id), so i don't follow the reasoning | 14:12 |
| *** hashar has quit IRC | 14:13 | |
| tobiash | I just looked at it as some kind of metadata | 14:13 |
| tobiash | but I'm not opinionated on this point | 14:13 |
| tobiash | so I'm completely fine making it its own attribute | 14:14 |
| Shrews | well, i'm the only dissenting voice, so go with what the majority likes :) | 14:16 |
| Shrews | i'll leave my comment on the review and let someone else make the final call | 14:17 |
| tobiash | Shrews: right now, you're the majority as I'm split on that topic ;) | 14:17 |
| tobiash | let's hear how corvus thinks about it | 14:17 |
| Shrews | commented | 14:21 |
| tobiash | Shrews: thanks, will update | 14:22 |
| *** jamesmcarthur has joined #zuul | 14:22 | |
| Shrews | tobiash: i do like that we will have that for problem tracing... finally! | 14:22 |
| tobiash | yes :) | 14:22 |
| corvus | tobiash, Shrews: i think Shrews proposal makes sense, so if there's no technical reason to avoid it, dropping "meta.zuul" makes sense | 14:23 |
| *** armstrongs has joined #zuul | 14:24 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Annotate node request processing with event id https://review.opendev.org/658889 | 14:25 |
| armstrongs | hey guys why you start the zuul-executor using systemd then when its running execute zuul-executor verbose is it expected that the live streamed logs when a job is in progress are run with ansible verbosity -vvv? It doesn't seem to be having any impact ... | 14:25 |
| armstrongs | when* | 14:25 |
| armstrongs | or is the verbosity only applied to the final saved log? | 14:26 |
| corvus | armstrongs: it should apply to the ansible output saved in the executor's own logs -- i don't think it should affect the log output saved by the job | 14:27 |
| armstrongs | so the web uis streamed ansible logs won't be -vvv? | 14:28 |
| fungi | e.g. /var/log/zuul/executor-debug.log | 14:29 |
| armstrongs | is there any way to make them more verbose? | 14:29 |
| Shrews | tobiash: you need to replace all the request.meta.get('zuul_event_id') portions | 14:29 |
| armstrongs | at job level or project level | 14:29 |
| *** jamesmcarthur has quit IRC | 14:29 | |
| tobiash | Shrews: oops, fixing | 14:30 |
| tobiash | thanks | 14:30 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Annotate node request processing with event id https://review.opendev.org/658889 | 14:32 |
| tobiash | armstrongs: no, buildlog.txt stays as is, but you can look into buildlog.json which has much more information | 14:32 |
| armstrongs | ok so just clarify you currently can't up the verbosity on the streamed logs to the zuul interface. buildlog.txt is the completed version of that, for more verbose ansible information i should look in buildlog.json | 14:35 |
| armstrongs | correct? | 14:35 |
| fungi | that's correct as far as i know | 14:36 |
| armstrongs | thanks guys | 14:36 |
| fungi | if we did add something to make it configurable, i'd expect that would be exposed as a settable attribute along the lines of job.ansible-verbosity (but with the detailed json output i'm not sure it's warranted) | 14:37 |
| *** jamesmcarthur has joined #zuul | 14:37 | |
| armstrongs | that would be great | 14:38 |
| armstrongs | as our users are asking for more verbosity | 14:39 |
| armstrongs | as for long running tests they are used to watching it in some instances | 14:39 |
| clarkb | I think our users rely on ara for that extra ansible insight | 14:40 |
| armstrongs | any pointers on setting it up alongside zuul? | 14:40 |
| armstrongs | we are doing a poc so may be worth a look | 14:41 |
| fungi | well, ara's postprocessed so doesn't help the users who like to spend their time sitting and staring at the screen while the job runs | 14:41 |
| * fungi feels like that's akin to watching paint dry, but each to their own | 14:42 | |
| armstrongs | we have test qa engineers | 14:42 |
| armstrongs | so no comment | 14:42 |
| armstrongs | :) | 14:42 |
| fungi | heh | 14:42 |
| armstrongs | they are used to jenkins output | 14:42 |
| armstrongs | and looked different | 14:42 |
| armstrongs | sigh | 14:43 |
| *** jamesmcarthur_ has joined #zuul | 14:43 | |
| fungi | if you were interested in trying to implement it, you could try to plumb in an attribute using https://zuul-ci.org/docs/zuul/user/config.html#attr-job.ansible-version as an example, and maybe just have it as an integer count of the number of -v's to pass to the command | 14:44 |
| clarkb | armstrongs: there is an ara output role in zuul-jobs you can add to your jobs | 14:44 |
| clarkb | fungi: it is sublter than that. Console logs are generated within ansible by a shim we ship | 14:44 |
| clarkb | they arent prt of the normal -vvvv processing in ansible iirc | 14:45 |
| fungi | ohh... so it wouldn't matter how ansible itself was invoked by the executor anyway? | 14:45 |
| clarkb | correct | 14:46 |
| fungi | armstrongs: an example of the ara integration you get by adding the ara output role from zuul-jobs: http://logs.openstack.org/46/657046/5/promote/swift-promote-image/ffb657d/ara-report/ | 14:46 |
| fungi | again, that's only after the job completes though, not available while it's in progress | 14:46 |
| *** jamesmcarthur has quit IRC | 14:47 | |
| armstrongs | thanks for the info everyone | 14:47 |
| corvus | armstrongs: zuul does have streaming console logs for watching tests as they run -- so anything that goes to stdout of a command task should stream in real time | 14:47 |
| fungi | one thing we've talked about wanting for in-progress jobs, and the streaming protocol choice explicitly allows for this if the necessary glue is written, is the ability to stream additional logs from the job by specifying a particular logfile you want to see | 14:47 |
| corvus | fungi: oh yeah, i still think that will be great to add :) | 14:48 |
| fungi | so that would be something possibly worth working on if you have qa engineers who like to watch logs | 14:48 |
| clarkb | thr way to think of the console log is it shows you what your job us doing not what ansible is doing | 14:49 |
| armstrongs | ah ok so if i shell out to tests they will be captured. Its just teh ansible bits that are surpressed | 14:49 |
| clarkb | so you can invrease the verbosityof your commands run by ansible | 14:49 |
| armstrongs | the ara reports look cool | 14:49 |
| corvus | also if there's something in particular that shows up if you run ansible from the command line but is missing in the streaming/saved logs, we might be able to correct that (the logging plugin is custom zuul code which is similar to the standard ansible log, but not 100% the same) | 14:49 |
| *** hashar has joined #zuul | 14:50 | |
| corvus | armstrongs: yeah, we did the command streaming specifically to preserve the behavior for people who liked to watch "test..OK" a lot :) | 14:50 |
| fungi | armstrongs: yeah, as an example, in our deployment we have some fairly complex integration test suites implemented as massive shell script spaghetti, and use `set -x` to turn on echoing of every shell command run by the script | 14:50 |
| *** pcaruana has joined #zuul | 14:53 | |
| Shrews | tobiash: lgtm. is there a nodepool equivalent change yet? i couldn't find it | 14:54 |
| tobiash | Shrews: not yet | 14:57 |
| tobiash | I'm working iteratively through the whole stack | 14:57 |
| Shrews | k | 14:57 |
| corvus | everything up to that point is +3d (again!) | 15:08 |
| tobiash | and the first change failed in the gate :( | 15:09 |
| tobiash | I've rechecked only the first three for now | 15:09 |
| tobiash | maybe it's easier to get them in slowly ;) | 15:09 |
| corvus | i'm heading out, probably away again until wednesday (though i'll try to review on the plane again) | 15:09 |
| corvus | tobiash: it must get nervious with all those changes behind them | 15:10 |
| *** themroc has quit IRC | 15:22 | |
| openstackgerrit | Merged zuul/zuul master: Fix typo in log message https://review.opendev.org/660118 | 15:29 |
| *** hashar has quit IRC | 16:00 | |
| *** chandankumar is now known as raukadah | 16:01 | |
| *** mattw4 has joined #zuul | 16:06 | |
| *** mattw4 has quit IRC | 16:20 | |
| *** jangutter has quit IRC | 16:50 | |
| *** sshnaidm is now known as sshnaidm|afk | 17:00 | |
| *** jpena is now known as jpena|off | 17:07 | |
| *** mattw4 has joined #zuul | 17:14 | |
| fungi | pabelanger: i'm about to update https://review.opendev.org/659674 but not sure if you've seen my replies to your comments | 17:16 |
| mordred | tobiash: did we ever land your statsd_exporter mapping file? | 17:19 |
| tobiash | mordred: I don't think so | 17:20 |
| mordred | tobiash: good, that explains why I can't find it - and not just that I'm dumb :) | 17:20 |
| tobiash | mordred: I'm not sure if it was a change or just somewhere on paste.openstack.org | 17:21 |
| mordred | tobiash: yeah - I don't see anything in gerrit, so I'm guessing it was just a paste - and then I think tristanC and I went somewhat down the path of seeing if we could support both sanely, and then didn't finish that | 17:22 |
| tobiash | mordred: http://paste.openstack.org/show/729542/ | 17:23 |
| tobiash | and http://paste.openstack.org/show/729541/ | 17:23 |
| tobiash | mordred: I think that's why I didn't create a change | 17:24 |
| tobiash | those are probably a bit outdated | 17:24 |
| tobiash | I think I'll create a change with my recent mappings tomorrow | 17:24 |
| mordred | cool - that'll be a good thing to point people to who are doing k8s deploys | 17:30 |
| *** jamesmcarthur has joined #zuul | 17:31 | |
| *** jamesmcarthur_ has quit IRC | 17:34 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Annotate node request processing with event id https://review.opendev.org/658889 | 17:40 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: WIP: Annotate builds with event id https://review.opendev.org/658895 | 17:40 |
| openstackgerrit | Jeremy Stanley proposed zuul/zuul master: Install latest git-review from PyPI in quickstart https://review.opendev.org/659674 | 17:48 |
| *** bhavikdbavishi has joined #zuul | 17:50 | |
| *** armstrongs has quit IRC | 18:04 | |
| openstackgerrit | Merged zuul/zuul master: Annotate github logs with the event id https://review.opendev.org/658645 | 18:15 |
| tobiash | \o/ | 18:20 |
| openstackgerrit | David Shrewsbury proposed zuul/nodepool master: WIP: Convert to using py3 pathlib https://review.opendev.org/660191 | 18:31 |
| openstackgerrit | Merged zuul/zuul master: Annotate gerrit event logs https://review.opendev.org/658646 | 18:39 |
| mordred | tobiash: woot! | 18:44 |
| tobiash | Next one failed :( | 18:45 |
| *** bhavikdbavishi has quit IRC | 19:00 | |
| *** jamesmcarthur has quit IRC | 19:05 | |
| *** jamesmcarthur has joined #zuul | 19:19 | |
| mordred | booo | 19:35 |
| *** toabctl has quit IRC | 19:43 | |
| *** toabctl has joined #zuul | 19:46 | |
| *** dmsimard5 has joined #zuul | 19:50 | |
| *** dmsimard has quit IRC | 19:51 | |
| openstackgerrit | David Shrewsbury proposed zuul/nodepool master: Use py3 pathlib in DibImageFile https://review.opendev.org/660191 | 20:07 |
| openstackgerrit | Merged zuul/zuul master: Attach event to queue item https://review.opendev.org/658647 | 20:17 |
| mordred | Shrews: your python3 is too modern for my old brainhole | 20:28 |
| Shrews | mordred: i find it more palatable than all the os.foo.bar calls | 20:29 |
| mordred | Shrews: yah. looks non-insane | 20:30 |
| Shrews | plus... objects and stuff | 20:30 |
| openstackgerrit | Merged zuul/zuul master: Annotate some logs in the scheduler with event id https://review.opendev.org/658648 | 20:30 |
| Shrews | i think my -2 was unnecessary. still convincing myself | 20:30 |
| Shrews | mordred: fwiw, i blame https://datawhatnow.com/things-you-are-probably-not-using-in-python-3-but-should/ | 20:31 |
| fungi | wow, i should read this | 20:34 |
| Shrews | i can't wait to enum all the things | 20:34 |
| fungi | pathlib overloaded the / operator as concatenation? intriguing | 20:34 |
| mordred | Shrews: ooh. dataclasses look nice | 20:35 |
| mordred | Shrews, fungi: yes - that was a helpful post | 20:35 |
| fungi | yep, looking forward to being able to use data classes once 3.7 is better entrenched | 20:39 |
| fungi | i have to say though, while the items on that list were usefully obscure, it was disappointingly short | 20:39 |
| fungi | there's a ton of great stuff in py3k which folks fail to take advantage of if they're too used to 2.x | 20:40 |
| Shrews | the problem with all the great new things in py3 is the greatness being distributed across multiple versions, so it makes discovering them rather difficult unless you really go digging, so most people (myself included) don't make a whole-hearted attempt. | 20:43 |
| fungi | and i'll be the first to admit i still write c-style printf format strings | 20:43 |
| Shrews | blogs that point them out (like that one) are super useful | 20:43 |
| Shrews | i like the f-strings. i'd like to move our min version for that alone | 20:45 |
| fungi | yeah, i just need to remember about them long enough to actually start using them | 20:46 |
| fungi | "foo %s bar %s baz" % (xyzzy, plugh) is just too ingrained in my fingers | 20:47 |
| fungi | need to remember i can f"foo {xyzzy} bar {plugh} baz" and save a ton of typing | 20:47 |
| openstackgerrit | Merged zuul/zuul master: Annotate logs in the zuul driver with event ids https://review.opendev.org/658649 | 20:48 |
| fungi | granted, that example save 6 characters, of which 3 were optional ;) | 20:49 |
| mordred | fungi: :) | 21:02 |
| mordred | Shrews: I concur, I am looking forward to f strings and am excited to move our min | 21:03 |
| *** tjgresha has joined #zuul | 21:03 | |
| *** mattw4 has quit IRC | 21:04 | |
| *** tjgresha has quit IRC | 21:05 | |
| openstackgerrit | Merged zuul/zuul master: Add event id to timer events https://review.opendev.org/658650 | 21:08 |
| clarkb | Shrews: left a couple notes on the pathlib change. Its possible the second one is realted to the failing tests, but of that I am not certain | 21:23 |
| *** mattw4 has joined #zuul | 21:24 | |
| Shrews | i am looking into the fail already. it's related to file ext | 21:24 |
| clarkb | ya my second comment is related to that as well | 21:24 |
| clarkb | basically there is ab ehavior change between the old and new code | 21:24 |
| Shrews | clarkb: i think your rsplit comment is wrong | 21:26 |
| Shrews | >>> s = "foo.qcow2.md5" | 21:27 |
| Shrews | >>> s.rsplit('.', 1) | 21:27 |
| Shrews | ['foo.qcow2', 'md5'] | 21:27 |
| clarkb | oh bah I read is as lsplit for some reason but its rsplit | 21:27 |
| clarkb | so ya nevermind I think it is the same then | 21:27 |
| openstackgerrit | Merged zuul/zuul master: Annotate pipeline processing with event id https://review.opendev.org/658651 | 21:27 |
| clarkb | I think because lsplit is what makes sense to me there | 21:27 |
| clarkb | but I guess filenames could haev .'s in them | 21:28 |
| Shrews | clarkb: well that's where the new code makes that easier to read :) i still have a problem with cleaning up the checksum files though. will have to figure that one out tomorrow | 21:38 |
| *** pcaruana has quit IRC | 21:40 | |
| Shrews | but yeah, i have problems remembering rsplit vs lsplit too. Path.suffix reads much better | 21:40 |
| clarkb | ya the only thing that reads weirder to me is the / concatenation (+ is used for that everywhere else in python so uh why not here) | 21:41 |
| *** jamesmcarthur has quit IRC | 21:45 | |
| openstackgerrit | Merged zuul/zuul master: Annotate merger logs with event id https://review.opendev.org/658652 | 21:45 |
| *** jamesmcarthur has joined #zuul | 21:46 | |
| fungi | clarkb: especially since / is only the unix file separator. other platforms use \ or : and python's path handling figures it out | 22:09 |
| fungi | so to choose one of them which happens to be an available operator and overload that is... an odd choice | 22:09 |
| openstackgerrit | Merged zuul/zuul master: Annotate job freezing logs with event id https://review.opendev.org/658888 | 22:18 |
| pabelanger | I cannot remember, does https://zuul-ci.org/docs/zuul/user/config.html#attr-job.files affect jobs in a periodic pipeline? | 22:29 |
| pabelanger | my gut says no | 22:29 |
| clarkb | pabelanger: iirc no because periodic timer triggers don't have files | 22:30 |
| clarkb | its possible the result is that you'll never get that job to run in periodic with a files matcher | 22:31 |
| pabelanger | yah, I'll test it out tomorrow for sure | 22:32 |
| *** dmsimard5 is now known as dmsimard | 22:44 | |
| SpamapS | pabelanger: IIRC you can use `files: []` in the attachment to the periodic pipeline. | 23:21 |
| SpamapS | but that may also just backfire and still match nothing | 23:22 |
| *** rlandy has quit IRC | 23:25 | |
| *** jamesmcarthur has quit IRC | 23:44 | |
| tristanC | mordred: Shrews: still no reproducer for the react-script bump ( https://review.opendev.org/659991 ). Do you remember on which page it failed? | 23:47 |
| clarkb | tristanC: it failed on zuul.openstack.org's status page | 23:47 |
| mordred | tristanC, clarkb: yeah. | 23:47 |
| mordred | although that seems to be working this time | 23:48 |
| clarkb | it also failed on zuul.opendev.org's openstack tenant status page | 23:48 |
| clarkb | so it was likely something to do with the openstack status json | 23:48 |
| mordred | hrm | 23:48 |
| mordred | clarkb: the old builds are also now working | 23:48 |
| tristanC | currently in the multi-tenant build, ansible status there are 2 changes from github that also seems to be working fine | 23:48 |
| clarkb | maybe the thing to do is fix the minimization breaks debugging problem, then we can reenable it and if it happens again hopefully we get the data necessary to debug | 23:48 |
| mordred | so it may have been something in the status.json at that moment that isn't there now | 23:49 |
| mordred | tristanC: yes - what clarkb said - the biggest issue was that the error message in the js debugger had no context for debugging at all | 23:50 |
| tristanC | that's odd, the console log should give a pointer at the map or source file | 23:50 |
| clarkb | it gave a link to a generic react debug site | 23:50 |
| mordred | yes - that would be my expectations - instead it gave a message with a link to ^^ | 23:50 |
| mordred | yeah, that | 23:50 |
| mordred | it was very strange | 23:51 |
| tristanC | let me introduce a bug then... | 23:52 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul master: DNM: this should cause a React exception https://review.opendev.org/660231 | 23:53 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul master: DNM: this should cause a React exception https://review.opendev.org/660231 | 23:54 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!