| *** jkilpatr has joined #zuul | 00:08 | |
| *** jkilpatr has quit IRC | 00:45 | |
| *** jkilpatr has joined #zuul | 00:57 | |
| tristanC | mordred: pabelanger: re-using min-ready and max-concurrency, the static node driver may be able to provide concurrent job execution on the same system | 04:00 |
|---|---|---|
| tristanC | mordred: about the oci driver, it is indeed re-using host fs so that containers are easy to manage and as light as possible | 04:02 |
| tristanC | and it seems like simple label such as 'oci-centos-7' should be enough, and easier to implement compared to 'oci:centos-7' | 04:05 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Implement a static driver for Nodepool https://review.openstack.org/468624 | 04:06 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Extend Nodepool configuration syntax to support multiple drivers https://review.openstack.org/468751 | 04:06 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Collect request handling implementation in an OpenStack driver https://review.openstack.org/468750 | 04:06 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Abstract Nodepool provider management code https://review.openstack.org/468749 | 04:06 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Add support for custom ssh port https://review.openstack.org/468752 | 04:07 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Implement an OpenContainer driver https://review.openstack.org/468753 | 04:07 |
| *** hashar has joined #zuul | 07:02 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Implement a static driver for Nodepool https://review.openstack.org/468624 | 07:20 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Extend Nodepool configuration syntax to support multiple drivers https://review.openstack.org/468751 | 07:20 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Collect request handling implementation in an OpenStack driver https://review.openstack.org/468750 | 07:20 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Abstract Nodepool provider management code https://review.openstack.org/468749 | 07:20 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Add support for custom ssh port https://review.openstack.org/468752 | 07:20 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool feature/zuulv3: Implement an OpenContainer driver https://review.openstack.org/468753 | 07:20 |
| *** openstackgerrit has quit IRC | 08:18 | |
| *** fbouliane has quit IRC | 09:27 | |
| *** Shrews has quit IRC | 09:27 | |
| *** SpamapS has quit IRC | 09:27 | |
| *** rcarrillocruz has quit IRC | 09:27 | |
| *** jlk has quit IRC | 09:27 | |
| *** rbergeron has quit IRC | 09:27 | |
| *** fbouliane has joined #zuul | 09:49 | |
| *** Shrews has joined #zuul | 09:49 | |
| *** SpamapS has joined #zuul | 09:49 | |
| mordred | tristanC: nod | 10:06 |
| *** rcarrillocruz has joined #zuul | 10:13 | |
| *** jlk has joined #zuul | 10:13 | |
| *** rbergeron has joined #zuul | 10:13 | |
| *** jkilpatr has quit IRC | 10:40 | |
| *** jkilpatr has joined #zuul | 10:56 | |
| *** dkranz has joined #zuul | 13:21 | |
| tobiash | hi | 13:23 |
| tobiash | is it possible to specify the location of the playbooks in the according zuul.yaml? | 13:23 |
| tobiash | (or .zuul.yaml) | 13:23 |
| tobiash | maybe some projects don't want to have a folder playbooks in the repo root but maybe in ci/playbooks | 13:24 |
| mordred | tobiash: yes, I believe so - although scanning the code it's not immediately clear how we exposed that | 13:27 |
| mordred | tobiash: nope. I'm lying - playbooks seems to be hardcoded right now | 13:29 |
| mordred | tobiash: I think it is the intent that it be configurable though | 13:29 |
| tobiash | mordred: ah, just thought I'm blind ;) | 13:29 |
| mordred | :) | 13:29 |
| tobiash | but good to know that that's something that will be configurable at some point in time | 13:31 |
| tobiash | maybe I have time for this topic after my vacation in a few weeks | 13:31 |
| *** morgan has quit IRC | 13:37 | |
| tobiash | yay, another decoding issue: http://paste.openstack.org/show/610971/ | 13:45 |
| pabelanger | tobiash: Yes, I've been meaning to submit a patch. I use the test folder currently in ansible roles for testing, would love to update .zuul.layout to account for that | 13:53 |
| tobiash | pabelanger: ok, cool | 13:57 |
| * mordred has a feeling humans are going to have strong opinions as to where they keep their playbooks | 14:17 | |
| pabelanger | https://review.openstack.org/#/c/468942/ easy fix is people would like to review nodepool. Currently console-log is broken syntax for nodepool | 14:34 |
| tobiash | how does console streaming (and saving to the console.log) work with multi-node tests? | 15:54 |
| clarkb | in a v2.5 world just the primary node gets console streamed. In v3 I believe each host gets its own console stream | 15:55 |
| tobiash | just tried this out and it generated one console.log with the console content of one node | 15:55 |
| tobiash | (in zuulv3) | 15:55 |
| clarkb | or rather its all run through central ansible (the executor) so all the console streaming from each host ends up back at the executor | 15:56 |
| clarkb | so its aggregated rather than single point | 15:56 |
| tobiash | I'll check how your base jobs work | 15:56 |
| *** hashar is now known as hasharAway | 15:57 | |
| *** morgan has joined #zuul | 15:58 | |
| tobiash | so I'm doing the same like the first two steps here: | 15:58 |
| tobiash | https://review.openstack.org/gitweb?p=openstack-infra/openstack-zuul-roles.git;a=blob;f=roles/prepare-workspace/tasks/main.yaml;h=4d42b2dff3ab9e72e3618271615177437cd32aa8;hb=HEAD | 15:58 |
| jlk | o/ | 16:00 |
| tobiash | ah, I think I know my error | 16:00 |
| tobiash | http://paste.openstack.org/show/610995/ | 16:00 |
| tobiash | log collection only leaves one console.log alive... | 16:00 |
| *** openstackgerrit has joined #zuul | 16:10 | |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Add console-log to config-validate https://review.openstack.org/468942 | 16:10 |
| jlk | SpamapS: I added a number of tasks to this story, did I do it right? https://storyboard.openstack.org/?#!/story/2000774 | 16:18 |
| SpamapS | jlk: aye, looks good | 16:30 |
| SpamapS | jlk: thanks! | 16:31 |
| tobiash | do we still have races in zuul tests? | 16:37 |
| tobiash | just had a failure of tests.unit.test_zuultrigger.TestZuulTriggerParentChangeEnqueued.test_zuul_trigger_parent_change_enqueued | 16:37 |
| tobiash | (locally) | 16:38 |
| jlk | If we do, they're somewhat deterministic races | 16:38 |
| tobiash | rerun of the test was successful | 16:38 |
| SpamapS | There are a few that are performance related. | 16:42 |
| SpamapS | like, internal stuff times out. | 16:42 |
| tobiash | rerun the whole suite was green the second time | 16:43 |
| SpamapS | I have seen a couple flakes in the gate/check tests too though, so it's likely there are a few races hiding in the weeds. | 16:43 |
| SpamapS | nothing that is well known and common | 16:43 |
| SpamapS | we should maybe dig around in kibana a little to see if there are any obvious patterns | 16:44 |
| SpamapS | but I think the rate still low, might not have enough data and _a lot_ has changed recently | 16:44 |
| tobiash | probably | 16:44 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Decode the console log stream https://review.openstack.org/469181 | 16:49 |
| tobiash | Another py3 decoding fix ^^^ | 16:49 |
| pabelanger | tobiash: your running ansible under python35? | 16:52 |
| tobiash | pabelanger: yes | 16:57 |
| tobiash | :) | 16:57 |
| pabelanger | tobiash: nice | 16:57 |
| tobiash | my zuulv3 test setup is py3 only currently ;) | 16:57 |
| pabelanger | Ya, I'd love for us to do that too | 16:58 |
| pabelanger | soon(tm) | 16:58 |
| tobiash | and it runs some simple jobs already... | 16:58 |
| tobiash | but it crashes every now and then with some py3 encoding issue | 16:58 |
| tobiash | (but all in zuul so far) | 16:59 |
| tobiash | and with fixes in review | 16:59 |
| tobiash | or merged | 16:59 |
| *** bhavik has joined #zuul | 17:11 | |
| *** harlowja has joined #zuul | 17:14 | |
| SpamapS | pabelanger: weren't there some tests in the review queue for that particular code path? | 17:16 |
| SpamapS | http://zuulv3-dev.openstack.org/logs/1ed5cccc9371461c963e7808004e6370/tox/py35-1.log.txt | 17:17 |
| SpamapS | Could not determine repository location of /home/zuul/src/git.openstack.org/openstack-infra/zuul | 17:17 |
| tobiash | SpamapS: is that a deterministic failure? | 17:21 |
| tobiash | didn't see this error yet | 17:22 |
| pabelanger | SpamapS: I'd have to check again, but don't think so | 17:23 |
| SpamapS | pabelanger: ^ that tox fail seems to happen periodically. Any idea? | 17:25 |
| tobiash | has the url_pattern been removed from the zuul config? | 17:30 |
| pabelanger | checking zuulv3-dev to see | 17:30 |
| tobiash | don't find any real occurrence in the code | 17:31 |
| tobiash | but it is still in some fixtures | 17:31 |
| tobiash | on post playbook failure the job defined failure url is not used for gerrit reporting | 17:34 |
| pabelanger | not sure, don't see anything obvious | 17:35 |
| pabelanger | does autohold work for zuulv3? | 17:35 |
| pabelanger | I know we changed it | 17:35 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Use util functions to make lists of things https://review.openstack.org/469192 | 17:39 |
| jlk | SpamapS: do commit hints referring to story/tasks on storyboard cause some automatic task manipulation, or do I have to follow that up manually to twiddle the status bit? | 17:40 |
| jlk | oh cool, automatically happens. | 17:42 |
| SpamapS | jlk: if you have a Story: and Task: yes, it's fully automated. | 17:42 |
| jlk | :thumbsup: | 17:42 |
| SpamapS | Story: will just put it in the info log of the story, but having both fully automates, which is nice. | 17:42 |
| SpamapS | jlk: also note that there's a "boardtty" for when you get tired of storyboard's somewhat sluggish JS UI ;) | 17:43 |
| SpamapS | though boardtty needs some work.. the "assign to" dropdown is unusable. | 17:43 |
| SpamapS | sorry, 'boartty' | 17:43 |
| SpamapS | I started working on that but got lost in some needed model changes :-P | 17:44 |
| jlk | BoarttyMcBoarttyface. | 17:44 |
| SpamapS | Yeah we never should have had that poll... ;) | 17:45 |
| jeblair | tobiash: in v3, we shouldn't copy console.log anymore; instead the console logs for all of the nodes (in a multinode job) get streamed and interleaved into ansible_log.txt on the executor, and that should generally be considered "the log" for a build. there is a bug with that right now that mordred is looking into -- we only get output from the first command task. | 18:05 |
| tobiash | jeblair: ok, good to know | 18:05 |
| tobiash | thanks | 18:05 |
| jeblair | tobiash: seccuss-url is a job attribute in v3, so right now, the strategy for log links is to set that (in the base job) to the same location where ansible would put the logs (probably also in the base job). unfortunately that means duplicating the value in two different places/systems (zuul config + ansible). | 18:07 |
| jeblair | tobiash: i'd like to move to a system where we have the executor read data back from ansible runs (maybe it just loads a json file from a known location) so that it can learn the log url from the job content itself. then we can remove that configuration from success-url. it just reports whatever ansible said it should for the logs. | 18:08 |
| tobiash | jeblair: I noticed that the failure-url is not taken into account if the job result is POST_FAILURE | 18:08 |
| jeblair | tobiash: that might be an oversight | 18:08 |
| tobiash | jeblair: so one of the post playbooks (probably the last one) should push the logs somewhere and leave a json file behind with the url? | 18:12 |
| jeblair | tobiash: yeah, that's what i'm thinking. we can use the mechanism to provide any data back to zuul in the future. we can start with log links (and maybe result comments), but it could eventually also be the way we get jobs to leave line comments as well. | 18:14 |
| tobiash | that would be cool :) | 18:14 |
| *** bhavik has quit IRC | 18:15 | |
| tobiash | jeblair: currently in the process of enhancing the test_playbook test with a check for success and failure urls | 18:18 |
| tobiash | http://paste.openstack.org/show/611005/ | 18:18 |
| tobiash | that's the current behavior | 18:19 |
| tobiash | jeblair: does it make sense to fix that or better wait for the new solution? | 18:20 |
| openstackgerrit | Paul Belanger proposed openstack-infra/nodepool feature/zuulv3: Add boot-from-volume support for nodes https://review.openstack.org/464283 | 18:23 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: WIP: Add web-based console log streaming https://review.openstack.org/463353 | 18:28 |
| Shrews | jeblair: ^^^ using a thread now, as you asked. Had to do some finagling of the event loop to get that to work | 18:29 |
| Shrews | now gotta figure out how to test it :/ | 18:38 |
| Shrews | but first... tea | 18:38 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix and test report urls for post fail and timeout https://review.openstack.org/469214 | 18:44 |
| *** bhavik has joined #zuul | 18:47 | |
| jeblair | tobiash: it probably makes sense to fix now; any unknown non-success result should probably use the failure pattern | 18:59 |
| jeblair | Shrews: w00t! hopefully with it running in a thread it should be easier to get it up and running in tests like we do with most of the other daemons (merger, etc) | 18:59 |
| tobiash | jeblair: ok, so just replace with else here https://review.openstack.org/#/c/469214/1/zuul/model.py | 19:00 |
| jeblair | tobiash: yeah, i think so | 19:01 |
| tobiash | will update the patch in a minute | 19:01 |
| *** bhavik has quit IRC | 19:04 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix and test report urls for unknown failures https://review.openstack.org/469214 | 19:05 |
| jlk | jeblair: we talked last week about oddness with requiring change like things (votes, etc..) on push like events. I've tested, and confirmed that it causes a traceback in the github code. I'm looking at gerrit now, and just confirmed that gerrit driver will traceback too on checking for say 'current_patchset' on a ref-updated event. | 19:08 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: Add config option for executor process user https://review.openstack.org/460671 | 19:08 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: Allow for using build UUID as temp job dir name https://review.openstack.org/456685 | 19:08 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: Add a finger protocol log streamer https://review.openstack.org/456721 | 19:08 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul feature/zuulv3: Allow for specifying root job directory https://review.openstack.org/456691 | 19:08 |
| *** ianw has joined #zuul | 19:20 | |
| jlk | jeblair: do you think it makes sense, in the context of a push / ref-update event, that current_patchset should always equate to true, or should it do something to look at the ref that is updated, and see if the sha of the update is still the latest sha on the ref? | 19:23 |
| jeblair | jlk: maybe always be true? | 19:45 |
| jlk | I'm drafting a mail for the list, to get some discussion on it, or at least dump my brain out somewhere other than IRC :D | 19:46 |
| jeblair | jlk: (it doesn't really make sense; i almost think that always returning true or always returning false are almost equally valid) | 19:46 |
| rbergeron | jlk: i dunno if you saw this or if it's old news, but github is moving from rest api to graphsomethingsomething | 20:16 |
| rbergeron | https://developer.github.com/v4/ | 20:16 |
| rbergeron | "Github is moving from a REST API to one based on GraphQL. Single request => several actions." | 20:16 |
| rbergeron | dunno if that... affects anything or is useful in the context of the mail you just sent, but: i merely bringeth possibly useful information, | 20:16 |
| mordred | yah. just in time for us to have finished the rest port :) | 20:16 |
| rbergeron | and hope that others can make use of things | 20:16 |
| rbergeron | and now i really go to do the errands | 20:17 |
| jeblair | jlk: question on https://review.openstack.org/463420 | 20:22 |
| mordred | jeblair: oh - so we want an "app" not an "integratoin" now? | 20:23 |
| jeblair | mordred: apparently? it looks like they've been updating their docs (which mostly say "app" in the title and "integration" buried deep in the text) | 20:24 |
| mordred | glorious | 20:24 |
| mordred | to be fair - "app" makes a bit more sense than "integration" | 20:24 |
| jeblair | mordred: you can buy apps for your projects too. | 20:25 |
| mordred | jeblair: what about for my tenants? | 20:25 |
| jeblair | mordred: i'm imagining in-app purchases. you could sell positive CI results. | 20:25 |
| jeblair | also, if we could somehow do gold farming that would be just great. | 20:26 |
| SpamapS | so, I finally got time to show Zuul to the Kubernetes sig-testing group | 20:40 |
| SpamapS | It went well, lots of good questions from people who have basically been going through some of the same growing pains as OpenSack. | 20:41 |
| *** jkilpatr has quit IRC | 20:44 | |
| jeblair | SpamapS: oh good! | 20:47 |
| pabelanger | SpamapS: great talk, was able to watch | 20:47 |
| dmsimard | pabelanger: I'm seeing that happen pretty often recently, does that ring you a bell ? | 20:51 |
| dmsimard | OpenStackCloudException: Tried to delete floating ip 38.145.33.45 associated with server 97f2998d-5103-4ee0-b114-fdc2563316fd but there was an error deleting it. Not deleting server | 20:51 |
| dmsimard | On the cloud, the VM no longer has a floating IP attached | 20:52 |
| dmsimard | and in the meantime the floating IP has been re-associated with another server | 20:52 |
| pabelanger | dmsimard: might be work posting the debug.log to pastebin and have mordred peak too | 20:52 |
| pabelanger | I'd have to check and see if that happens for tripleo-test-cloud-rh1 | 20:53 |
| dmsimard | this is against rdo cloud | 20:53 |
| pabelanger | right, but nodepool.o.o might see it with tripleo since it also is using FIP | 20:54 |
| dmsimard | ah | 20:54 |
| dmsimard | pabelanger: but.. basically: http://paste.openstack.org/raw/611023/ | 20:54 |
| mordred | SpamapS: yay! | 20:55 |
| pabelanger | dmsimard: that debug log will have more logging for shade | 20:55 |
| pabelanger | which we'd need to see | 20:55 |
| pabelanger | dmsimard: is this nodepool? | 20:55 |
| pabelanger | or manually running commands | 20:56 |
| mordred | dmsimard, pabelanger: hrm. why did shade think that server had 38.145.33.45 associated with it in the first place | 20:56 |
| dmsimard | pabelanger: nodepool is leaving VMs running due to that floating IP error so I look manually. | 20:56 |
| pabelanger | dmsimard: ya, nodepool logs will be helpful | 20:56 |
| dmsimard | I'll increase verbosity, not seeing much right now | 20:57 |
| jlk | rbergeron: I saw that and it's an open tab for me to start reading. | 21:08 |
| mordred | dmsimard: I will be interested in what you learn | 21:09 |
| dmsimard | mordred: getting sidetracked right now, our nodepool's logging.conf is pretty darn messy too | 21:10 |
| dmsimard | let you know when I find something | 21:10 |
| mordred | dmsimard: kk. thanks | 21:11 |
| jlk | jeblair: ugh, yeah I didn't see they already deprecated "integration_id". I'll rebase. | 21:12 |
| clarkb | pabelanger: dmsimard fwiw the reason we don't delete the server if the floating ip fails to delete is that only good way of knowing if the fip is in use is via the server record so idea is they have to be deleted together | 21:13 |
| dmsimard | clarkb: makes sense, something is off though -- it's like there's a race condition where the floating ip is unbound first, allocated to another server and then it tries to delete the original server and fails | 21:17 |
| *** jkilpatr has joined #zuul | 21:21 | |
| jeblair | omg jlk! i have reviewed your entire github stack. the integration/apps thing was the only non +2. cc mordred | 21:26 |
| jlk | hells yeah! | 21:26 |
| jlk | jeblair: in that case, should I wait for mordred 's +2/W and just do a follow up patch to s/integration_id/app_id/ ? | 21:27 |
| jlk | instead ore rebasing 10~ patches? | 21:27 |
| jeblair | jlk: if it's all the same to you, i'd say fix+rebase now; i'll re-add any +2s of mine that get dropped and mordred or i can +w the stack | 21:31 |
| jlk | alrighty | 21:31 |
| mordred | jlk: I agree with all of the things | 21:32 |
| mordred | jeblair: yay! | 21:32 |
| mordred | jeblair, jlk: I'll go do the followup review pass as soon as the rebase hits | 21:32 |
| *** jkilpatr has quit IRC | 21:45 | |
| *** jkilpatr has joined #zuul | 21:45 | |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Include exc_info in reporter failure https://review.openstack.org/460765 | 21:46 |
| mordred | jlk, jeblair: wow. no kidding. the rest of that stack looks great | 21:46 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add cachecontrol to requests to github https://review.openstack.org/461587 | 21:47 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Implement github trigger requirement status https://review.openstack.org/463386 | 21:47 |
| jlk | incoming..... | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Use util functions to make lists of things https://review.openstack.org/469192 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Include tenant name in github context https://review.openstack.org/463871 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Implement github pipeline req of current-patchset https://review.openstack.org/468589 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Implement github pipepline req of open https://review.openstack.org/468213 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Save installation ids to a cache and fetch them per project https://review.openstack.org/463424 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Catch integration key file not found in zuul https://review.openstack.org/463425 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Create github client each time it's called. https://review.openstack.org/463421 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Use app_id with github https://review.openstack.org/463420 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Refactor integration key loading to always load https://review.openstack.org/463423 | 21:54 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Decode JSON body once for requests https://review.openstack.org/463422 | 21:54 |
| *** dougbtv has quit IRC | 22:07 | |
| pabelanger | confirmed locally nodepool-builder was running under python3.5 :) | 22:09 |
| pabelanger | I know we test it with nodepool-dsvm, but still cool having it work locally | 22:09 |
| *** hasharAway has quit IRC | 22:17 | |
| jeblair | mordred: i just saw https://review.openstack.org/465718 and realized i duped some (but not all) of that. sorry. | 22:20 |
| mordred | jeblair: no worries at all! | 22:20 |
| mordred | jlk, jeblair: BOOM. just +A'd the stack | 22:28 |
| * jlk dances | 22:29 | |
| jeblair | jlk, jesusaur: just reviewed https://review.openstack.org/449794 which depends on an abandoned change | 22:31 |
| * jeblair decommissions "Github Zuul v3" dashboard and restarts gertty | 22:32 | |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Use app_id with github https://review.openstack.org/463420 | 22:33 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Create github client each time it's called. https://review.openstack.org/463421 | 22:33 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Decode JSON body once for requests https://review.openstack.org/463422 | 22:34 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Refactor integration key loading to always load https://review.openstack.org/463423 | 22:34 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Save installation ids to a cache and fetch them per project https://review.openstack.org/463424 | 22:34 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Catch integration key file not found in zuul https://review.openstack.org/463425 | 22:34 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Include tenant name in github context https://review.openstack.org/463871 | 22:34 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Implement github pipepline req of open https://review.openstack.org/468213 | 22:35 |
| jesusaur | whoo, the github stack is merging | 22:35 |
| *** Zara has quit IRC | 22:35 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Convert some leftover changeish mentions to ref https://review.openstack.org/465718 | 22:35 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Implement github pipeline req of current-patchset https://review.openstack.org/468589 | 22:36 |
| *** SotK has quit IRC | 22:36 | |
| jeblair | mordred: +2 thx! | 22:36 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Use util functions to make lists of things https://review.openstack.org/469192 | 22:37 |
| jeblair | jlk: ^ that's the tail! | 22:37 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Log and collect stats for events consistently https://review.openstack.org/468664 | 22:40 |
| jlk | we'll have to re-deploy and see where it falls over. But jamielennox|away is gone all week | 22:41 |
| jeblair | jlk, mordred: well, it might be about time for us to point zuulv3-dev at github too, so you might have some compatriots there :) | 22:42 |
| mordred | indeed! | 22:42 |
| mordred | I can think of two different jobs I'd like to run just off the top of my head | 22:43 |
| jeblair | note that i still don't consider zuulv3 secure or correct yet, so baby steps are warranted. | 22:43 |
| jlk | We also don't have depends-on yet, so things will be a bit difficult for cross-repo type workflows | 22:45 |
| mordred | jlk: they will - but "does this PR to ansible break zuul" would be a good start even without depends-on | 22:46 |
| jeblair | yeah, and that's something we need to pull up into the driver interface. i don't think it'll be too hard. we should look at tobiash's dependency-cycle-detection-in-pipeline change first; it may help and affect how we implement that. | 22:47 |
| mordred | jeblair: speaking of tobiash... I just respun https://review.openstack.org/468664 - but tobiash has a comment on PS1 that is potentially worth not missing | 22:48 |
| jeblair | mordred: yeah, i agree, and https://review.openstack.org/468554 is partly based on his suggestions | 22:50 |
| mordred | jeblair: ah neat! | 22:51 |
| jeblair | maybe that's the start of a pattern? i dunno, it's new to me. :) | 22:54 |
| jeblair | seemed to work pretty well in that case. | 22:55 |
| jeblair | will probably work well in the other distributed workers too (merger, etc). | 22:55 |
| jeblair | the scheduler / pipeline managers may need a slightly different approach (since they are single threaded) | 22:55 |
| mordred | jeblair: btw - jesusaur left a comment on https://review.openstack.org/#/c/461176 | 23:01 |
| *** SotK has joined #zuul | 23:03 | |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Store initial repo state in the merger https://review.openstack.org/461176 | 23:10 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Use previously stored repo state on executor https://review.openstack.org/461177 | 23:11 |
| jeblair | mordred, jesusaur: oops, too late for my -1 on my own change | 23:11 |
| jeblair | jesusaur: your question led me to find an error :) | 23:11 |
| mordred | jeblair: yay! | 23:11 |
| jeblair | i think we can just fix with a followup; it's not a regression | 23:12 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add default-branch property to projects https://review.openstack.org/467334 | 23:12 |
| mordred | jeblair: cool | 23:12 |
| jeblair | i'm looking into the failures of the last 3 changes in that stack (which seem legit) | 23:13 |
| *** dougbtv has joined #zuul | 23:14 | |
| jesusaur | jeblair: yay, i'm helping! | 23:14 |
| *** dkranz has quit IRC | 23:19 | |
| jlk | the "my changes" screen in gertty is so so much smaller now! | 23:35 |
| openstackgerrit | Jesse Keating proposed openstack-infra/zuul feature/zuulv3: Remove pipeline argument from various report fncts https://review.openstack.org/464252 | 23:41 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!