| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Make zuul ansible dir self-contained https://review.openstack.org/468208 | 00:11 |
|---|---|---|
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add untrusted-projects ansible test https://review.openstack.org/461881 | 00:12 |
| jeblair | SpamapS: it looked like "from . import paths" wasn't working out (i don't think "." in that context means what we hoped). so i pushed a new patchset to change 468208 with a variant on that approach based on an idea mordred had. i managed to get it to pass the tests in the following change locally. [cc pabelanger] | 00:14 |
| jeblair | jlk: i'm past eod and zonked (see above brain teaser), i'll have to look at that tomorrow, sorry | 00:15 |
| tobiash | morning | 04:21 |
| tobiash | do some of you also encounter this one on executor restart? | 04:21 |
| tobiash | http://paste.openstack.org/show/611156/ | 04:21 |
| tobiash | do you consider this as a bug in my service file (missing delete as pre-start) or the executor (should handle pre-existing ansible pycache)? | 04:32 |
| *** hashar has joined #zuul | 07:57 | |
| *** hashar has quit IRC | 08:05 | |
| *** hashar has joined #zuul | 08:05 | |
| *** hashar has quit IRC | 08:20 | |
| *** hashar has joined #zuul | 08:45 | |
| *** smyers has quit IRC | 09:04 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Implement a static driver for Nodepool https://review.openstack.org/468624 | 09:14 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Extend Nodepool configuration syntax to support multiple drivers https://review.openstack.org/468751 | 09:14 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Collect request handling implementation in an OpenStack driver https://review.openstack.org/468750 | 09:14 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Abstract Nodepool provider management code https://review.openstack.org/468749 | 09:14 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Abstract Nodepool request handling code https://review.openstack.org/468748 | 09:14 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Add support for custom ssh port https://review.openstack.org/468752 | 09:14 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Implement an OpenContainer driver https://review.openstack.org/468753 | 09:14 |
| *** smyers has joined #zuul | 09:14 | |
| mordred | tobiash: oh wow fun | 10:14 |
| tobiash | mordred: so should I fix the executor or my systemd service file? ;) | 10:23 |
| mordred | tobiash: I don't know - but I think the executor - I'm confused as to why copytree can't handle it | 10:25 |
| mordred | tobiash: oh! https://docs.python.org/3/library/shutil.html#shutil.copytree ... ignore= can take a callable that takes as an argument the list of things in the dir copytree is visiting and returns ones to ignore | 10:27 |
| mordred | tobiash: so def _ignore(dir, files): if "__pycache__" in files: return ["__pycache__"] else: return [] and copytree(ignore=_ignore) should probably do the trick, yeah? | 10:28 |
| mordred | since we don't really want to copy __pycache__ files | 10:29 |
| tobiash | mordred: thanks, will try that out | 10:32 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Make zuul ansible dir self-contained https://review.openstack.org/468208 | 10:38 |
| *** jkilpatr has quit IRC | 10:45 | |
| *** jkilpatr has joined #zuul | 11:02 | |
| pabelanger | jeblair: SpamapS: thanks, going to debug bwrap failure this morning. Not sure why job is currently failing | 12:02 |
| Shrews | i was hitting that pycache thing yesterday during streamer testing | 13:01 |
| *** dkranz has joined #zuul | 13:17 | |
| *** jkilpatr has quit IRC | 13:53 | |
| *** jkilpatr has joined #zuul | 13:55 | |
| SpamapS | pabelanger: sweet thanks | 14:16 |
| * SpamapS is just catching up with backscroll from yesterday | 14:16 | |
| SpamapS | jeblair: ACK, I saw the patches you and mordred made just now. I'm not sure I understand where the pythonpatch change comes in, but I'm glad we're making progress. | 14:17 |
| mordred | SpamapS: basically - copy not just each of the plugin dirs, but copy the whole zuul/ansible tree, and put it in a zuul dir with a __init__.py - then put the dir containing that zuul dir into PYTHONPATH so that the path lookups work - without needing to copy any more data than we're copying now | 14:29 |
| mordred | SpamapS: since there isn't anything in zuul/ansible other than dirs we copy anyway - its the same amount of copying - but we don't want to copy the _entire_ zuul tree cause that _is_ more things to copy | 14:31 |
| *** openstackgerrit has quit IRC | 14:34 | |
| SpamapS | mordred: so the thing I think I was missing was the PYTHONPATH change | 14:39 |
| mordred | yah | 14:56 |
| mordred | although my fix of the pep8 failure sure did not go well :) | 14:56 |
| mordred | oh - that's ust the setuptools/six issue | 14:57 |
| *** jkilpatr has quit IRC | 14:59 | |
| *** jkilpatr has joined #zuul | 15:09 | |
| *** toabctl has quit IRC | 15:22 | |
| *** toabctl has joined #zuul | 15:23 | |
| *** hashar has quit IRC | 15:28 | |
| jeblair | mordred, SpamapS: ah, i botched the rework of the zuul.ansible change anyway. fixing. | 16:11 |
| *** openstackgerrit has joined #zuul | 16:14 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Make zuul ansible dir self-contained https://review.openstack.org/468208 | 16:14 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add untrusted-projects ansible test https://review.openstack.org/461881 | 16:14 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Remove source from reporter https://review.openstack.org/462362 | 16:23 |
| jeblair | mordred: want to +3 https://review.openstack.org/461981 which you previously reviewed? | 16:24 |
| jeblair | SpamapS, mordred, pabelanger: the bwrap changes all pass tests now! | 16:25 |
| jeblair | i'm just going to fix my -1 on 453851 since SpamapS is conferencing | 16:25 |
| pabelanger | jeblair: thanks for looking into the test | 16:26 |
| jlk | o/ | 16:26 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Make zuul ansible dir self-contained https://review.openstack.org/468208 | 16:27 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add untrusted-projects ansible test https://review.openstack.org/461881 | 16:27 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add support for bwrap https://review.openstack.org/453851 | 16:27 |
| mordred | jeblair: done | 16:27 |
| mordred | jeblair: do you think it's worth checking to see if a PYTHONPATH exists and appending to it if it does? | 16:27 |
| jeblair | mordred: hrm, that does seem like the polite thing to do | 16:28 |
| jeblair | mordred: adding that now | 16:28 |
| mordred | jeblair: env_copy['PYTHONPATH'] = os.path.sep.join(['self.executor_server.ansible_dir'] + [env_copy.get('PYTHONPATH')]) shouldn't be _terrible_ | 16:28 |
| jlk | jeblair: looking at https://review.openstack.org/#/c/461981/7/zuul/driver/gerrit/gerritreporter.py and from your advice yesterday, couldn't this just check change.project.connection_name and self.connection.connection_name ? | 16:31 |
| jeblair | mordred: it's separated by ':' right? | 16:31 |
| mordred | jeblair: yah. well - or os.path.pathsep possibly | 16:32 |
| mordred | jeblair: I'm not sure it it's separated by somethign else on non-linux -but we liely have other issues there anyway | 16:33 |
| mordred | so I doubt os.path.pathsep will be the thing that kills us :) | 16:33 |
| jeblair | mordred: ah yeah, so os.path.pathsep. | 16:34 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Make zuul ansible dir self-contained https://review.openstack.org/468208 | 16:34 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add untrusted-projects ansible test https://review.openstack.org/461881 | 16:34 |
| jeblair | mordred: i had already typed ':' | 16:35 |
| jeblair | meh, i'll change it | 16:35 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Make zuul ansible dir self-contained https://review.openstack.org/468208 | 16:36 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add untrusted-projects ansible test https://review.openstack.org/461881 | 16:36 |
| *** nt has joined #zuul | 16:36 | |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Only report to gerrit if the action is from gerrit https://review.openstack.org/461981 | 16:37 |
| jeblair | jlk: yeah, i've been trying to wrap my head around that -- i think the canonical name is desired over the connection name there because in the use case tobiash was referencing in earlier comments, the user actually wants it to report to a connection other than the source connection for the change. | 16:38 |
| jeblair | jlk: the use case is basically to report to gerrit as a different user than you might normally use | 16:38 |
| jlk | oh | 16:38 |
| jeblair | jlk: so it's using the fact that multiple connections might have the same canonical name to do that. tbh, i'm not sure that's entirely going to hold up. | 16:39 |
| jlk | yeah, that makes it weird if I've got multiple tenants connected to public github | 16:39 |
| jlk | they're all going to share the same canonical URL, but I don't necessarily want them cross-talking | 16:40 |
| jeblair | jlk: i expect the tenants will all share the same connection though | 16:40 |
| jlk | maybe tenant was the wrong word there | 16:40 |
| jlk | If I have multiple connections to public github | 16:40 |
| jlk | maybe I wouldn't? | 16:40 |
| jlk | But I might if they need to have different trigger / report / require / reject | 16:41 |
| jlk | maybe I'm overthinking this | 16:41 |
| jeblair | jlk: all of those should be able to happen with one connection. normally a zuul install would have one connection to each server (one gerrit, one github, etc). only reason to have more than one for a given server is to use a new set of credentials (ie, if you wanted your zuul to present as two separate github apps). | 16:43 |
| jlk | yeah, or one as an app, and one as a distinct user | 16:44 |
| jeblair | oh you know what, i think that check may actually just want to check the *hostname* not the *canonical_hostname*. | 16:44 |
| jeblair | what we're looking for in that change is "if i report to review.openstack.org, is it going to understand this change" | 16:44 |
| jeblair | lemme etherpad this | 16:45 |
| tobiash | jeblair: yes, this is a quite nice summary of the intent of that change | 16:51 |
| tobiash | so my personal use case was to separate different gerrits from each other | 16:52 |
| tobiash | for this the connection name would be sufficient | 16:52 |
| tobiash | but in the tests there is also the use case to have the source in one connection and reporting to a different connection to the same gerrit | 16:52 |
| tobiash | I personally don't have the latter use case but as there is a test case for that I thought comparing the host names would enable my use case and also the already existing test case | 16:53 |
| jeblair | tobiash: yeah, and i think the actual hostname (vs canonical hostname, which may be different) should work for your case as well, right? | 16:55 |
| tobiash | jeblair: I think so | 16:55 |
| tobiash | isn't the canonical hostname the fqdn? | 16:55 |
| tobiash | so my intent was to identify the same gerrit by using the real dns name of the gerrit (aka review.openstack.org) | 16:56 |
| mordred | tobiash: canoncal_hostname is the fdqn of the place from which git repos are known to be located | 16:57 |
| jeblair | tobiash: that's the "hostname". the "canonical_hostname" might be that, or something completely different. canonical_hostname is there so that we know what to call the repos when we put them on disk. | 16:57 |
| mordred | tobiash: so, for openstack, it's "git.openstack.org" | 16:58 |
| jeblair | right, so that people writing golang import lines can say "git.openstack.org/openstack/something" | 16:58 |
| mordred | yup | 16:58 |
| mordred | I agree, I think hostname should work for tobiash use case | 16:59 |
| jeblair | but now that i've dug into this further, i think either one will work just as well. maybe we should leave it as canonical_hostname for now and see if there are problems. :) | 16:59 |
| tobiash | seems that I misunderstood the canonical hostname... | 16:59 |
| jeblair | tobiash: it defaults to the fqdn if you don't set it to anything else | 17:00 |
| tobiash | ok, rtfm helps... | 17:01 |
| tobiash | that comes from the connection config | 17:01 |
| jeblair | i think the thing that's been confusing me is that we haven't really explored whether or why someone might have 2 connections to the same server and set either the same or different canonical_hostnames for them. | 17:01 |
| tobiash | so I really meant the hostname | 17:01 |
| jeblair | jlk: the approach in 469297 lgtm | 17:05 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Case sensitive label matching in canMerge https://review.openstack.org/469946 | 17:06 |
| tobiash | I don't know when you are planning the gerrit update to 2.13. We did in march and it broke our gate. This is my proposed fix ^^^ | 17:07 |
| mordred | jeblair: I agree with your confusion | 17:08 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Case sensitive label matching in canMerge https://review.openstack.org/469946 | 17:08 |
| jlk | jeblair: re 469297 okay. I'm chasing down an issue now where the zuul trigger is instantiated wrong, and the "connection" attribute is being set to a list (the config) rather than the real connection object. | 17:09 |
| pabelanger | I've +3'd https://review.openstack.org/#/c/468532/ to zuulv3-dev.o.o to fix our apache index.html. | 17:14 |
| pabelanger | I've also applied the change manually since puppet doesn't run on the server | 17:15 |
| *** dmsimard is now known as dmsimard|off | 17:16 | |
| pabelanger | would it be possible to append a build ID to our logging? Trying to follow along which playbook is running for which job is a little confusing in executor-debug.log | 17:18 |
| pabelanger | DEBUG zuul.AnsibleJob currently, hoping we could DEBUG zuul.AnsibleJob.<job id> | 17:19 |
| jlk | oooooh poop. | 17:20 |
| jeblair | pabelanger: you may be interested in reviewing https://review.openstack.org/468554 | 17:21 |
| jlk | jeblair: looking at https://git.openstack.org/cgit/openstack-infra/zuul/tree/tests/fixtures/config/zuultrigger/parent-change-enqueued/git/common-config/zuul.yaml?h=feature/zuulv3 | 17:21 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul feature/zuulv3: Do not merge https://review.openstack.org/469953 | 17:21 |
| pabelanger | jeblair: ty looking | 17:22 |
| jlk | jeblair: It looks like a zuul trigger will happen for a gerrit change. I think my change will block that from going through. But still digging. | 17:22 |
| jlk | 2017-06-01 17:22:37,881 zuul.IndependentPipelineManager DEBUG Filter <ZuulEventFilter types: parent-change-enqueued pipelines: gate> skipped for change <Change 0x7ff594139b50 1,1> due to mismatched connections | 17:23 |
| jlk | yeah damn. | 17:23 |
| jeblair | jlk: there's a zuul requirement filter? | 17:24 |
| jlk | Yeah it's a zuul event filter | 17:24 |
| jeblair | jlk: oh! you did trigger filters too, i didn't notice that | 17:24 |
| jlk | was that too far? | 17:25 |
| jlk | but this isn't a trigger filter, it's an event filter | 17:25 |
| jlk | a pipeline requirement | 17:25 |
| jeblair | that looks like a trigger filter, which is an event filter | 17:26 |
| pabelanger | jeblair: looks exactly what I was hoping for. +2 but left a question | 17:26 |
| jeblair | (vs a requirement filter, which is a ref filter) | 17:26 |
| jlk | I'm so confused :) | 17:26 |
| jlk | I thought a trigger filter was something like 'require-approval:' | 17:27 |
| jeblair | we're making new things! :) | 17:27 |
| jeblair | jlk: i think the difference is positive vs negative matching. with EventFilters, which are created by triggers, if any one matches, the event matches and we enqueue the change. | 17:27 |
| jlk | my original intent was to prevent a github connection change from being blocked by a gerrit connection pipeline requirement. | 17:28 |
| jeblair | jlk: with RefFilters, which are created by require/reject pipeline requirements, we want them all to match. | 17:28 |
| jlk | I may have gone too far and touched on the trigger filters | 17:28 |
| tobiash | gate failed here with some unrelated error: https://review.openstack.org/#/c/465046/ | 17:28 |
| tobiash | http://paste.openstack.org/show/611233/ | 17:28 |
| tobiash | shall I recheck? | 17:29 |
| jeblair | jlk: right, i think you only need the change on line 293, not line 127 of zuul/manager/__init__.py | 17:29 |
| jlk | alrighty | 17:29 |
| jeblair | tobiash: yeah, that's showing up *a lot*. i was just starting to look into it. Shrews, have you gotten anywhere with that? | 17:30 |
| jlk | for my own brain meat, what's the right term for a requirement expressed at the pipeline level vs a requirement expressed at the trigger level, vs just a regular old trigger criteria? | 17:30 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Fix bad text wrap in status page https://review.openstack.org/468532 | 17:30 |
| tobiash | jeblair: ok | 17:31 |
| jeblair | jlk: i think 'trigger requirement' and 'trigger criteria' are really the same thing. that may be the point of confusion. when you're adding an approval requirement to a trigger, you're not implicitly adding a pipeline requirement, you're just saying, for this trigger to fire, this has to be met. that naturally already works with cross-connection changes since a gerrit change will never fire the github trigger. and that's okay, because it ... | 17:33 |
| jeblair | ... might fire the gerrit trigger that immediately follows and still end up in the pipeline. | 17:33 |
| Shrews | jeblair: no. Failed to reproduce on all my local runs. SpamapS says it's similar to the FD issue he's been chasing. | 17:33 |
| jeblair | Shrews: i managed to repro locally in a tight loop of single foreground runs with: while ttrun -epy27 tests.unit.test_log_streamer; do echo ok; don | 17:34 |
| jeblair | Shrews: i managed to repro locally in a tight loop of single foreground runs with: while ttrun -epy27 tests.unit.test_log_streamer; do echo ok; done | 17:34 |
| jeblair | Shrews: it took a little while to hit though | 17:34 |
| jlk | okay. It's just that trigger requirements are a bit "special" because some of them require deeper knowledge of the change in question, knowledge that isn't necessarily available in the raw event content. | 17:34 |
| Shrews | jeblair: hmm, I did something similar | 17:34 |
| Shrews | Will keep poking | 17:35 |
| jeblair | jlk: yep. i think it's okay for a github trigger event filter to say "i can't evaluate this, return False" | 17:35 |
| jlk | like the event content could just be a "recheck" comment, but to trigger, the change requires that there be a certain approval level on the change. | 17:35 |
| jlk | wtf... | 17:43 |
| jlk | I've got a test, where a job is clearly executing, but self.builds and self.history are both empty lists. | 17:43 |
| jlk | aaaand I've just found another bug :/ | 17:47 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Whitelist pydevd debug threads https://review.openstack.org/465046 | 17:47 |
| jlk | jeblair: interesting question, how do you report on a ref-updated type event, given that there is no change associated with it? How would you report success/failure? | 17:52 |
| SpamapS | jlk: email? | 17:53 |
| SpamapS | like, statically configured address, SMTP Reporter? | 17:53 |
| jlk | whose? | 17:54 |
| SpamapS | You'd define it in the job | 17:54 |
| jlk | a pipeline runs many projects through, you'd think the email target would be project or job specific. | 17:54 |
| jlk | oh can you define that in the job? | 17:54 |
| SpamapS | Good point | 17:55 |
| SpamapS | I think that's in pipeline config. | 17:55 |
| SpamapS | seems like it might be useful to be able to vary reporter config in jobs | 17:56 |
| SpamapS | This kind of goes to the same problem of being able to communicate job results to zuul and/or next jobs | 17:57 |
| jlk | ah solved first issue. Don't use noop job if you want something to actually hit self.builds or self.history | 17:59 |
| *** jkilpatr_ has joined #zuul | 18:15 | |
| SpamapS | 18:16 | |
| SpamapS | derp | 18:16 |
| jlk | 18:16 | |
| *** jkilpatr has quit IRC | 18:18 | |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Handle change related reqs on push like events https://review.openstack.org/469297 | 18:35 |
| jlk | jeblair: ^^ is now ready for real reviews. | 18:36 |
| Shrews | jeblair: i'm well over 1000 runs of that test now, still haven't hit it | 18:59 |
| Shrews | heh, figures. failure at run 1284 | 19:03 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Remove pipeline argument from various report fncts https://review.openstack.org/464252 | 19:22 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Case sensitive label matching https://review.openstack.org/469946 | 19:41 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix tests for "Case sensitive label matching" https://review.openstack.org/469985 | 19:41 |
| tobiash | Due to a behavior change in gerrit 2.13 review and submit does not work if the casing of the label is not the same as configured in gerrit | 19:42 |
| tobiash | the above patches remove case conversion for the labels in zuul so this issue can be solved | 19:43 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix tests for "Case sensitive label matching" https://review.openstack.org/469985 | 19:46 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix tests for "Case sensitive label matching" https://review.openstack.org/469985 | 19:47 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: Fix log streamer closing https://review.openstack.org/469989 | 19:57 |
| Shrews | jeblair: that ^^^ should cover the random fail. I'm still not sure exactly why the socket is sometimes already closed. | 19:57 |
| pabelanger | zuul question, why if a job is running and has a depends-on patch that is reuploaded, that job reports: This change depends on a change that failed to merge. | 20:13 |
| *** jkilpatr_ has quit IRC | 20:47 | |
| mordred | jeblair: we missed Zuul's birthday on Monday! | 20:59 |
| mordred | jeblair: Date: Tue May 29 14:49:32 2012 -0700 | 20:59 |
| mordred | Initial commit. | 20:59 |
| SpamapS | man, we're such a-holes | 21:01 |
| *** jkilpatr has joined #zuul | 21:06 | |
| mordred | SpamapS: ikr? | 21:12 |
| mordred | jlk, SpamapS: re: ref-updated events - there have been a few requests in openstack land from folks who want a specific job to emails results to a specific place, but they do not want to define a whole new pipeline | 21:13 |
| mordred | so maybe there's a thing we should think about | 21:14 |
| mordred | jlk, SpamapS: OTOH, I was also thinking that, at least with github, github allows you to leave arbitrary comments on files in a repo | 21:14 |
| mordred | so maybe the github plugin could have the ability to leave ref-updated comments on the commit - or even in-line ones if we got fancy | 21:15 |
| mordred | https://github.com/ansible/ansible/commit/eb83c6f4e772c6861e41b4cc17296e00747d83aa <-- at bottom, example of leaving a comment on a specific commit | 21:16 |
| SpamapS | that's interesting | 21:20 |
| jlk | yeah that's... interesting | 21:25 |
| jlk | another "extension" to the git standard | 21:25 |
| jlk | In Github you can also set a status on any commit object, it is not tied to a pull request | 21:35 |
| *** dkranz has quit IRC | 21:36 | |
| jeblair | mordred: were you planning on continuing review of 469595-468564? | 21:46 |
| jeblair | Shrews: re 469989 -- if we get an error from shutdown, do we not want to run close? | 21:47 |
| Shrews | jeblair: the error comes from shutdown when it's already closed | 21:48 |
| jeblair | Shrews: gotcha | 21:49 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Fix log streamer closing https://review.openstack.org/469989 | 21:58 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Fix fake gearman py3 bug and re-enable sched tests https://review.openstack.org/468583 | 22:13 |
| mordred | jeblair: yes! | 22:17 |
| mordred | jeblair: I kinda feel like we should have a tool that makes little javascript slideshows like the zuul demo given a test case | 22:23 |
| mordred | jeblair: I don't actually want to write that, mind you | 22:23 |
| jeblair | that would be nice :) | 22:23 |
| jeblair | if we do something for the status page, we could automatically derive it from the tests | 22:24 |
| mordred | jeblair: zomg. a view of the status page for a given change or pipeline that would be animated like the slideshow | 22:25 |
| mordred | would be fun to put on a really big monitor | 22:25 |
| mordred | little boxes floating down to little branch tips | 22:26 |
| jeblair | mordred: the idea that someone might do that someday is actually why the canvas animation has an object model that matches zuul's. :) | 22:32 |
| mordred | jeblair: :) | 22:33 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Extend job list when inheriting jobs https://review.openstack.org/465393 | 22:35 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_project_override https://review.openstack.org/470049 | 22:41 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Enable test_periodic https://review.openstack.org/470053 | 22:55 |
| SpamapS | you guys are making my presentation brain explode with desire | 23:01 |
| mordred | SpamapS: ikr? | 23:02 |
| SpamapS | FYI, Zookeeper CVE .. CVE-2017-5637 https://people.canonical.com/~ubuntu-security/cve/2017/CVE-2017-5637.html | 23:06 |
| SpamapS | "The wchp/wchc four letter words can be exploited in a DOS attack on the | 23:07 |
| SpamapS | ZK client port - typically 2181. | 23:07 |
| SpamapS | " | 23:07 |
| SpamapS | jeblair: so, 453851, how do we get that one reviewed and approved, given that I wrote most of it but you uploaded the null driver fix? | 23:12 |
| jlk | in v3 land, when somebody pushes a change to a config repo, is zuul supposed to "notice" this and reconfigure? | 23:20 |
| pabelanger | untrusted yes | 23:22 |
| pabelanger | trusted you need to reload zuul | 23:23 |
| pabelanger | jlk: you mean zuul.yaml right? | 23:24 |
| jlk | yes | 23:24 |
| pabelanger | now you have me thinking | 23:25 |
| pabelanger | for openstack-infra, we have only added projects to our zuul.yaml in a config repo. I do know you need to have the projects in the tenant before you can add them to zuul.yaml otherwise zuul -1 as a configuration error | 23:27 |
| pabelanger | but, being trusted jobs, zuul will only notice them after they merge, zuul should reconfigure automatically | 23:28 |
| pabelanger | but jeblair knows best | 23:28 |
| *** jamielennox|away is now known as jamielennox | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!