| tristanC | SpamapS: beside the image management part, it should work as expected | 00:19 |
|---|---|---|
| tristanC | though i only tested it with the freenium plan, so it may needs some adjustment for more intensive usage | 00:20 |
| SpamapS | Yeah just that I don't have to do all the baseline stuff of being the second driver. ;) | 00:23 |
| *** yolanda_ has joined #zuul | 00:24 | |
| tristanC | also, we should probably have a global zuul auth section in nodepool.yaml to register the zuul ssh key | 00:24 |
| tristanC | it would be easier for driver to setup zuul access | 00:25 |
| *** yolanda__ has quit IRC | 00:27 | |
| tristanC | tobiash: note that there is a spec to implement cli action over zuul-web api: https://review.openstack.org/562321 | 00:46 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: Implement an Amazon EC2 driver https://review.openstack.org/535558 | 01:17 |
| *** bhavik1 has joined #zuul | 03:37 | |
| *** CrayZee has quit IRC | 03:38 | |
| openstackgerrit | Merged openstack-infra/zuul master: Don't try to delete non-existing local refs https://review.openstack.org/563982 | 03:47 |
| *** swest has joined #zuul | 05:00 | |
| *** swest has quit IRC | 05:01 | |
| *** swest has joined #zuul | 05:02 | |
| *** bhavik1 has quit IRC | 05:26 | |
| *** AJaeger has quit IRC | 06:14 | |
| *** AJaeger has joined #zuul | 06:17 | |
| openstackgerrit | Felix Schmidt proposed openstack-infra/zuul master: Add start and end timestamp to task result in zuul_json callback https://review.openstack.org/563888 | 06:34 |
| *** AJaeger has quit IRC | 06:35 | |
| *** AJaeger has joined #zuul | 06:38 | |
| openstackgerrit | Felix Schmidt proposed openstack-infra/zuul master: Add start and end timestamp to task result in zuul_json callback https://review.openstack.org/563888 | 07:24 |
| *** jpena|off is now known as jpena | 07:42 | |
| *** electrofelix has joined #zuul | 07:58 | |
| *** ssbarnea_ has joined #zuul | 08:00 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: reporter: do not expect branch attribute in Tag object https://review.openstack.org/560700 | 08:10 |
| tobiash | tristanC: fixed your test case ^ | 08:10 |
| tobiash | you missed the sql reporting in the tag pipeline | 08:11 |
| tristanC | tobiash: thanks! | 08:14 |
| *** xinliang has quit IRC | 08:49 | |
| *** xinliang has joined #zuul | 08:51 | |
| *** xinliang has quit IRC | 08:51 | |
| *** xinliang has joined #zuul | 08:51 | |
| mordred | fungi: to be fair to npm devs (what the heck am I saying) the upath lib declares a max boundary which us installing node10 violates - so I'd say actually the system there is working and possibly prevented us from something being broken by the release | 09:58 |
| mordred | (like, prevented something from being broken at runtime by throwing a compile-time assert, which is way better than "we're goingto update a thing you can't reasonably have more than one version of without a way of you choosing behavior" :) | 09:59 |
| *** gregdek has quit IRC | 11:19 | |
| *** gregdek has joined #zuul | 11:20 | |
| *** mattclay has quit IRC | 11:33 | |
| *** mattclay has joined #zuul | 11:33 | |
| *** jpena is now known as jpena|lunch | 11:35 | |
| *** mario-goulart has joined #zuul | 11:46 | |
| *** pwhalen_ has joined #zuul | 12:11 | |
| *** pwhalen has quit IRC | 12:11 | |
| *** yolanda_ has quit IRC | 12:23 | |
| *** gothicmindfood has quit IRC | 12:23 | |
| *** jpena|lunch is now known as jpena|off | 12:25 | |
| *** yolanda_ has joined #zuul | 12:26 | |
| *** gothicmindfood has joined #zuul | 12:26 | |
| *** eventingmonkey has quit IRC | 12:26 | |
| *** eventingmonkey has joined #zuul | 12:28 | |
| *** jpena|off is now known as jpena | 12:41 | |
| tobiash | tristanC, corvus: I re-reviewed the mqtt driver and revoked my +2 because we should think if we want to change the message format slightly | 12:51 |
| *** hashar has joined #zuul | 12:55 | |
| *** pwhalen_ is now known as pwhalen | 12:55 | |
| *** pwhalen has quit IRC | 12:56 | |
| *** pwhalen has joined #zuul | 12:56 | |
| tristanC | tobiash: that sounds good, i'll have a look tomorrow | 13:08 |
| fungi | mordred: yeah, seems sane enough. i just couldn't let the close timing of pip 10 and npm 10 pass unnoticed | 13:09 |
| fungi | (pip 10 is not especially broken either, they've just made some choices in the name of progress which require us to modernize a questionable practice) | 13:11 |
| *** maxamillion has quit IRC | 13:12 | |
| *** maxamillion has joined #zuul | 13:13 | |
| *** hashar has quit IRC | 13:15 | |
| *** maxamillion has quit IRC | 13:24 | |
| *** maxamillion has joined #zuul | 13:24 | |
| mordred | fungi: oh - TOTALLY :) | 14:09 |
| *** dkranz has joined #zuul | 14:24 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: [WIP] Colors in job-output.txt https://review.openstack.org/564066 | 14:29 |
| tobiash | corvus, clarkb: I think it would be good to have 560700 in the next release as it fixes sql reporting on tag triggered pipelines | 14:39 |
| corvus | tobiash: well, we just restarted yesterday -- can we maybe get that one in the next release? | 14:43 |
| tobiash | corvus: oh, didn't notice that you already restarted | 14:44 |
| tobiash | Next release is fine then | 14:44 |
| corvus | tobiash: we also just missed your merger fix because of the node snafu :( | 14:45 |
| tobiash | corvus: as I'm not tag driven I'm fine with having it in master :) | 14:46 |
| Shrews | mordred: fyi, there is a sort of mini-discussion on https://review.openstack.org/536862 when you get a chance. tl;dr it would be nice if there was a way to make ttrun work the way it used to before stestr | 14:53 |
| Shrews | mordred: but if there isn't an answer other than "set env vars when running", i guess i can live with it. just wondering if you had any other ideas | 14:55 |
| mordred | Shrews: I should go read that ... in the zuul change I had to do something in the base test case - but the intent is that ttrun works the way it used to before stestr | 14:58 |
| mordred | Shrews: so let's consider the fact that it doesn't right now to be a -1 worthy bug | 14:58 |
| Shrews | okie dokie | 14:59 |
| *** yolanda_ has quit IRC | 15:01 | |
| *** gothicmindfood has quit IRC | 15:01 | |
| *** jhesketh has quit IRC | 15:01 | |
| *** xinliang has quit IRC | 15:01 | |
| *** electrofelix has quit IRC | 15:01 | |
| *** sshnaidm|off has quit IRC | 15:01 | |
| *** pbrobinson has quit IRC | 15:01 | |
| *** toabctl has quit IRC | 15:01 | |
| *** nguyenhai has quit IRC | 15:01 | |
| *** rcarrillocruz has quit IRC | 15:01 | |
| *** zigo has quit IRC | 15:01 | |
| *** snapiri has quit IRC | 15:01 | |
| *** adam_g has quit IRC | 15:01 | |
| *** ChanServ has quit IRC | 15:01 | |
| *** nguyenhai has joined #zuul | 15:01 | |
| *** rcarrillocruz has joined #zuul | 15:01 | |
| *** zigo has joined #zuul | 15:01 | |
| *** snapiri has joined #zuul | 15:01 | |
| *** adam_g has joined #zuul | 15:01 | |
| *** ChanServ has joined #zuul | 15:01 | |
| *** barjavel.freenode.net sets mode: +o ChanServ | 15:01 | |
| *** nguyenhai has quit IRC | 15:02 | |
| clarkb | Shrews: mordred the idea I had was to check if invoked under testtools.run or subunit.run and change the behavior default in the tests based on that | 15:02 |
| clarkb | I think you can check the argv to get that info | 15:02 |
| *** jhesketh has joined #zuul | 15:03 | |
| mordred | clarkb: if it's like zuul I think there's actually likely a bug we can fix before needing to get that fancy - although that's a good idea if the other thing doesn't work | 15:03 |
| *** yolanda_ has joined #zuul | 15:03 | |
| *** gothicmindfood has joined #zuul | 15:03 | |
| mordred | like - I have this working the way we want without needting to set env vars in several repos already | 15:03 |
| Shrews | oh good | 15:04 |
| *** xinliang has joined #zuul | 15:04 | |
| *** electrofelix has joined #zuul | 15:04 | |
| *** sshnaidm|off has joined #zuul | 15:04 | |
| *** pbrobinson has joined #zuul | 15:04 | |
| *** toabctl has joined #zuul | 15:04 | |
| *** nguyenhai_ has joined #zuul | 15:04 | |
| *** weshay has quit IRC | 15:04 | |
| mordred | *but* - I like the way clarkb thinks :) | 15:05 |
| *** nguyenhai_ has quit IRC | 15:05 | |
| *** nguyenhai_ has joined #zuul | 15:05 | |
| *** weshay has joined #zuul | 15:06 | |
| Shrews | mordred: this seems to be an error message from yarn in the nodepool-zuul-functional job: http://logs.openstack.org/45/564045/2/check/nodepool-zuul-functional/c4a75b8/job-output.txt.gz#_2018-04-24_20_51_45_137565 | 15:24 |
| Shrews | not quite sure how to unravel that | 15:25 |
| *** rlandy has joined #zuul | 15:25 | |
| clarkb | Shrews: I think that may have been the issue corvus fixed yesterday. Does it complain about upack and >4 < 9? | 15:25 |
| clarkb | if so a recheck should get it moving | 15:25 |
| Shrews | clarkb: yeah | 15:25 |
| Shrews | cool. thx | 15:25 |
| clarkb | basically nodejs 10.0.0 released. upack doesn't support >9 nodejs. Corvus updated our nodeenv install to install 8.x (latest stable) | 15:26 |
| corvus | well, second latest stable :) | 15:30 |
| clarkb | er right | 15:31 |
| dims | anyone have an example of how i can specify a canonical_hostname in my .zuul.yaml? | 15:43 |
| clarkb | dims: I think that is determined by the zuul server configuration not the per project configuration | 15:44 |
| clarkb | dims: it has to do with where the repo is hosted canonically from zuul's view which is in the connection configuration | 15:45 |
| *** andrea06590 has joined #zuul | 15:45 | |
| pabelanger | zuul.project.canonical_hostname is the current value of it, from the inventory | 15:45 |
| dims | clarkb : kubernetes uses vanity urls. github.com/kubernetes/cloud-provider-openstack uses k8s.io/kubernetes/cloud-provider-openstack as package name ... i wrongly assumed ? | 15:45 |
| mordred | dims: wow, that's awesome | 15:46 |
| dims | guess i have to try some gymnastics .. | 15:46 |
| clarkb | in this case the canonical name would be github.com because that is wher the repo lives right? | 15:47 |
| clarkb | but I guess that breaks the golang idea of git repos are the thing and they are all on github expcet for when they are not | 15:47 |
| corvus | clarkb: golang doesn't require github | 15:48 |
| dims | they tell folks to checkout code under $GOPATH/src/k8s.io ... | 15:48 |
| corvus | (and the golang format is exactly what zuul produces -- on purpose) | 15:49 |
| clarkb | corvus: iirc there is a whitelist of urls that you have to get yourself added into to function with some tool | 15:49 |
| corvus | dims: maybe you could set up a special connection to github for k8s, and set the canonical hostname on that connection to k8s.io? | 15:49 |
| clarkb | so its not arbitrary urls but ya they are of that format | 15:49 |
| clarkb | `go get`? | 15:49 |
| *** andrea06590 has quit IRC | 15:49 | |
| dims | in the openlab zuul run, /home/zuul/src/github.com/kubernetes/cloud-provider-openstack is where the code gets put down | 15:50 |
| dims | corvus : y, i have to try some stuff | 15:50 |
| corvus | dims, mrhillsman: https://zuul-ci.org/docs/zuul/admin/drivers/github.html#attr-%3Cgithub%20connection%3E.canonical_hostname | 15:51 |
| corvus | that's what we use in openstack to trick our zuul in to calling things "git.openstack.org/openstack/nova" instead of "review.openstack.org/openstack/nova" | 15:51 |
| corvus | i... never expected it to be used with github though. but it should be supported. | 15:52 |
| dims | corvus : fun right? :) | 15:52 |
| corvus | it will mean you'll have more than one connection to github. one for just k8s.io projects, and another for github.com projects. | 15:52 |
| dims | corvus : ack will try that | 15:53 |
| dims | thanks | 15:53 |
| mordred | dims: does go get github.com/kubernetes/cloud-provider-openstack do the riht thing? | 15:53 |
| * mordred is curious if there is a general declaration mechansim somewhere | 15:53 | |
| dims | mordred : best practice there is to use "go get k8s.io/kubernetes/cloud-provider-openstack" | 15:53 |
| mordred | GOTCHA | 15:54 |
| mordred | so yeah - what corvus said | 15:54 |
| dims | as long i don't have to re-register the github app thingy | 15:54 |
| dims | do i have to? | 15:54 |
| mrhillsman | nope | 15:55 |
| corvus | dims: that's a really good question. the app id is associated with the connection, but let me see what happens if you have 2 connections with the same app id | 15:56 |
| dims | thanks corvus ! | 15:56 |
| dims | mrhillsman : you caught the full conversation right? | 15:57 |
| mrhillsman | well I thought no because. Note that Zuul will still only communicate with the GitHub server identified by server; | 15:57 |
| mrhillsman | I'm reading back but I think I got it | 15:57 |
| dims | thanks mrhillsman | 15:57 |
| mrhillsman | As I was reading, the src/github.com is problematic because best practice is k8s.io | 15:58 |
| mrhillsman | but now that I think about it, may have to register another app :( | 15:59 |
| clarkb | mrhillsman: sounds like corvus is going to test it | 15:59 |
| corvus | dims: the github app specifies the zuul connection name, so yeah, a second connection needs a second app | 16:00 |
| mrhillsman | and it would make sense to have to | 16:00 |
| corvus | clarkb: well, i went to read code :) | 16:00 |
| corvus | mrhillsman: yep | 16:00 |
| *** acozine1 has joined #zuul | 16:00 | |
| mordred | a short term workaround would be to just put "ln -s src/github.com src/k8s.io" in before doing other things | 16:00 |
| corvus | it does seem like it would be nice to be able to make an alias like this in the tenant config. but there's a lot of stuff in zuul that hangs off the canonical hostname, so that will take some careful thought to make sure it's possible. | 16:01 |
| mordred | that way go will find the source code repos at $GOPATH/src/k8s.io when it expects to | 16:01 |
| mordred | corvus: ++ | 16:01 |
| corvus | mordred: yeah, that might not be a bad idea, if it's possible to incorporate those into the relevant jobs. | 16:02 |
| dims | mordred : y that's what we did and "dep ensure" croaks | 16:04 |
| mrhillsman | I need to catch up on this cause ^ | 16:04 |
| dims | mrhillsman : don't want to do a second app right away .... should be the last resort | 16:05 |
| mrhillsman | I need to check the jobs as I am not sure how/when that symlink is created | 16:05 |
| mrhillsman | Gotcha dims | 16:05 |
| dims | corvus : mordred : can we muck with zuul's data to hack this in? | 16:06 |
| pabelanger | dims: you can write a new role that updates prepare-workspaces to sync the repos into the right location, if symlinks fail | 16:06 |
| clarkb | fwiw k8s.io/kubernetes/cloud-provider-openstack 404s for me | 16:06 |
| dims | to the existing connection | 16:06 |
| clarkb | if I http(s) it | 16:06 |
| *** hashar has joined #zuul | 16:06 | |
| clarkb | dims: I think you can probably apply that configuration corvus linked above but it will likely break anyone else using that connection expecting github.com | 16:07 |
| dims | clarkb : y understood | 16:07 |
| dims | pabelanger : that sounds promising | 16:08 |
| mrhillsman | dims do you have a link to dep ensure output? | 16:08 |
| dims | "both /home/zuul/src/k8s.io/cloud-provider-openstack and /home/zuul/src/github.com/kubernetes/cloud-provider-openstack are in the same GOPATH /home/zuul" | 16:08 |
| dims | http://logs.openlabtesting.org/logs/39/139/3d22a1a45af104946b6cac6c03f89151fc214161/cloud-provider-openstack-unittest/cloud-provider-openstack-unittest/823ad14/job-output.txt.gz | 16:08 |
| mordred | mrhillsman, dims: you could also mv the dir if dep ensure croaks on symlinks | 16:08 |
| mrhillsman | pabelanger: I think that is good | 16:08 |
| dims | mordred : yep | 16:08 |
| mrhillsman | Yeah, we have a few options for sure | 16:08 |
| mrhillsman | I need to look at the playbooks | 16:09 |
| mrhillsman | jobs | 16:09 |
| corvus | dims: the quickest thing is going to be doing it in the jobs; so maybe mv as mordred suggests. even the "easiest" changes to zuul's config are not going to be trivial. this is fundamental stuff :) | 16:09 |
| clarkb | wait why have url scoped "packages" if you can't have two packages with the same name on different hosts | 16:09 |
| *** andreaf_ has joined #zuul | 16:09 | |
| mordred | I'd do mv as a short-term rather than a modified prepare-workspaces role - just so that it's a small surface area | 16:09 |
| mrhillsman | Whatever lol | 16:09 |
| mordred | what corvus said | 16:09 |
| mrhillsman | Ok mordred | 16:09 |
| dims | gotcha mordred corvus | 16:10 |
| *** andreaf has quit IRC | 16:10 | |
| mordred | like, assume the base job is going to put things into zuul locations - then make a k8s-base job for things that are going to clone k8s repos that does mvs (you might want to only do mv src/github.com/kubernetes src/k8s.io/kubernetes) | 16:10 |
| corvus | clarkb: maybe it's because they both declare they are the same package? | 16:11 |
| *** andreaf_ is now known as andreaf | 16:11 | |
| clarkb | corvus: maybe? I thought the url was part of the code declaration too (which is why dims needs to address this in this way I expect) | 16:11 |
| dims | mrhillsman : will check back later ... need to get on a call | 16:11 |
| mordred | well - I think the tooling assumption is that you would have used go get to clone the repos | 16:12 |
| clarkb | it seems like a weird design choice to scope thigns then ignore the scoping | 16:12 |
| mordred | and if you did that, having both would indicate an error | 16:12 |
| mordred | that you should probably know about | 16:12 |
| mrhillsman | dims: ok | 16:12 |
| corvus | clarkb: right, but both github/foo and k8s/foo have a file in them that says "i am k8s/foo"... that's what i'm *assuming* is the problem | 16:12 |
| mrhillsman | well that is true in the case of go proper | 16:13 |
| mordred | reconciling two systems that want to clone git for you turns out to be fun :) | 16:13 |
| SpamapS | I have a question about trusted/untrusted roles... | 16:13 |
| mrhillsman | but k8s practice is to use their namespace | 16:13 |
| SpamapS | Yesterday I tried to make a generic "fetch all the journalctl outputs" role and put it in my shared jobs repo (godaddy-zuul-jobs) | 16:13 |
| mrhillsman | anywho, thanks for the insight gents | 16:14 |
| SpamapS | But when I pulled that role in on my base job.. it failed because zuul.executor.log_root was readonly. | 16:14 |
| mordred | sounds like a nice role | 16:14 |
| corvus | SpamapS: log_root shouldn't be readonly for anyone | 16:14 |
| SpamapS | Is there something I missed? Is there an intermediary place to dump logs that will get pulled in later? | 16:14 |
| corvus | SpamapS: can you paste the error? | 16:15 |
| SpamapS | Hm maybe I misinterpreted the error. | 16:15 |
| mordred | SpamapS: (did you accidentally run the role on the remote node expecting it to run on the executor, and thus zuul.executor.log_root didn't actually exist?) | 16:15 |
| mordred | or something like that - but yeah, if you can share error that would be neat | 16:16 |
| SpamapS | ah no | 16:16 |
| SpamapS | it was not log_root, good catch | 16:16 |
| SpamapS | it was zuul.project.src_dir | 16:16 |
| mordred | SpamapS: also ... I AM SORRY - I was supposed to have improved log collection months ago and still haven't | 16:16 |
| corvus | write your error message on an index card and put it in a white envelope and carnac the magnificent error message oracle will decipher it for you | 16:17 |
| SpamapS | http://paste.openstack.org/show/719936/ | 16:18 |
| SpamapS | here are the relevant sections | 16:18 |
| SpamapS | I think I'm using src_dir as scratch space | 16:18 |
| SpamapS | and that doesn't work :-P | 16:18 |
| SpamapS | 2018-04-24 17:24:26.312446 | POST-RUN START: [trusted : github.secureserver.net/cloudplatform/project-config/playbooks/base/post.yaml@master] | 16:19 |
| SpamapS | running trusted.. so it's not that | 16:19 |
| corvus | SpamapS: zuul_work_dir is generally used in roles that expect to run on the remote node, so it has no prefix, it just starts with "src/", under the assumption that the task is running in the remote user's home directory | 16:20 |
| corvus | SpamapS: in this case, when running on the executor, it just ends up with a relative path, and pwd is usually the playbook | 16:21 |
| corvus | SpamapS: so that's actually trying to write to something outside the work dir.. like /trusted/project_0/project-config/src/github.... etc | 16:22 |
| corvus | SpamapS: since you're operating on the executor, you probably want to do what you said earlier: use log_root instead | 16:22 |
| SpamapS | I'm not sure why this was running on localhost actually. | 16:27 |
| SpamapS | oh derp | 16:29 |
| SpamapS | this is in my base post playbook | 16:29 |
| corvus | SpamapS: oh you want to run it on the remote node to do a journalctl export first? | 16:29 |
| SpamapS | which only does stuff on localhost. | 16:29 |
| SpamapS | derp derp derp | 16:29 |
| openstackgerrit | Monty Taylor proposed openstack-infra/nodepool master: Switch to stestr https://review.openstack.org/536862 | 16:29 |
| SpamapS | corvus: right | 16:29 |
| SpamapS | I think I have it now | 16:29 |
| SpamapS | I was lazy, not reading the playbook | 16:29 |
| corvus | SpamapS: you should be able to have that playbook run tasks on "all:" hosts | 16:29 |
| SpamapS | need to have a 'hosts: all' play before the 'hosts: localhost' | 16:29 |
| corvus | ya, cool. | 16:29 |
| mordred | Shrews: ^^ I do not have internet bandwidth enough to validate that patch | 16:29 |
| * SpamapS recedes into the shadows | 16:29 | |
| mordred | Shrews: but I believe it will provide you with more joy | 16:30 |
| mordred | SpamapS: come back! we like you! | 16:30 |
| SpamapS | corvus: you make an excellent rubber ducky | 16:30 |
| corvus | i make bathtime lots of fun! | 16:30 |
| corvus | i'm... gonna go work on memory caches and stuff. | 16:30 |
| openstackgerrit | Stephen Finucane proposed openstack-infra/zuul-jobs master: Default warning-is-error to True for non-legacy Sphinx projects https://review.openstack.org/559348 | 16:32 |
| Shrews | mordred: i demand ponies, full of joy and goodness and joyness | 16:32 |
| *** mrhillsman has quit IRC | 16:43 | |
| *** mrhillsman has joined #zuul | 16:43 | |
| *** ssbarnea_ has quit IRC | 17:08 | |
| *** jpena is now known as jpena|off | 17:12 | |
| *** hashar is now known as hasharAway | 17:13 | |
| clarkb | corvus: followup from yesterday's zuul restart, I've not noticed any issues | 17:14 |
| *** acozine1 has quit IRC | 17:24 | |
| mordred | Shrews: I have boxes of rhinos | 17:36 |
| corvus | clarkb: cool, so maybe let's go ahead and tag it? | 17:46 |
| clarkb | corvus: I think so | 17:46 |
| corvus | wow, there's 33 changes from 3.0.1 to 623533ab85ddf406bcac16452a638a706682775a | 17:49 |
| corvus | i'm going to tag http://git.zuul-ci.org/cgit/zuul/commit/?id=623533ab85ddf406bcac16452a638a706682775a as 3.0.2 -- it's what openstack-infra has been running since yesterday | 17:50 |
| corvus | https://zuul-ci.org/docs/zuul/releasenotes.html#in-development doesn't show any major changes | 17:52 |
| corvus | clarkb, fungi, mordred, tobiash, Shrews: ^ sound good? | 17:53 |
| tobiash | corvus: lgtm | 17:54 |
| fungi | yep, 3.0.2 seems fine | 17:54 |
| clarkb | ya largely bug fixes I think | 17:54 |
| fungi | haven't seen any regressions in production today anyway | 17:54 |
| corvus | i also think nodepool has seen enough changes to warrant a release, but i'm not sure if we're running master. maybe we should work on that next. (i don't think it needs to be coordinated with zuul, just while we're thinking about it) | 17:54 |
| fungi | right, we agreed we weren't going to sync nodepool release times nor version numbers to zuul | 17:55 |
| clarkb | ya likely needs a similar global restart and watch it for ~12 hours then call it good | 17:55 |
| fungi | but a nodepool release also seems warranted | 17:55 |
| corvus | zuul 3.0.2 pushed | 17:55 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Add regex support to project stanzas https://review.openstack.org/535713 | 17:57 |
| tobiash | clarkb: that should address your comments ^ | 17:58 |
| *** ssbarnea_ has joined #zuul | 18:01 | |
| *** pwhalen has quit IRC | 18:04 | |
| *** pwhalen has joined #zuul | 18:08 | |
| *** ssbarnea_ has quit IRC | 18:12 | |
| *** ssbarnea_ has joined #zuul | 18:13 | |
| Shrews | i think there are some big enough nodepool changes that i'd like to see that run for a few days at least before another release | 18:22 |
| *** ssbarnea_ has quit IRC | 18:32 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: Fix race in test_hold_expiration_no_default https://review.openstack.org/564314 | 18:55 |
| Shrews | mordred: ^^ is the reason your stestr change failed | 18:55 |
| Shrews | rechecking it for you | 18:56 |
| Shrews | oh shoot, that's wrong | 18:57 |
| Shrews | grrrr | 18:57 |
| *** ssbarnea_ has joined #zuul | 19:03 | |
| *** myoung is now known as myoung|dr | 19:03 | |
| *** bhavik1 has joined #zuul | 19:09 | |
| Shrews | hrm, i'm not certain how this test has been working at all, actually | 19:14 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool master: Fix race in test_hold_expiration_no_default https://review.openstack.org/564314 | 19:25 |
| Shrews | that should do it | 19:25 |
| *** lsell has joined #zuul | 19:33 | |
| *** bhavik1 has quit IRC | 20:10 | |
| *** myoung|dr is now known as myoung | 20:41 | |
| pabelanger | Oooh, a logo | 21:17 |
| *** openstackgerrit has quit IRC | 21:20 | |
| *** ssbarnea_ has quit IRC | 21:55 | |
| *** hasharAway has quit IRC | 22:02 | |
| *** rlandy has quit IRC | 22:11 | |
| *** threestrands has joined #zuul | 22:17 | |
| *** openstackgerrit has joined #zuul | 22:45 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: [WIP] Cache configuration objects in addition to YAML dicts https://review.openstack.org/564061 | 22:45 |
| corvus | that's really close to ready, but i'm going to split out some changes to make it more reviewable | 22:46 |
| *** threestrands_ has joined #zuul | 23:35 | |
| *** threestrands has quit IRC | 23:38 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!