| *** rlandy has quit IRC | 00:28 | |
| *** y2kenny has quit IRC | 00:52 | |
| *** swest has quit IRC | 01:16 | |
| *** wuchunyang has joined #zuul | 01:23 | |
| *** wuchunyang has quit IRC | 01:26 | |
| *** wuchunyang has joined #zuul | 01:27 | |
| *** swest has joined #zuul | 01:33 | |
| *** rfolco|rover has quit IRC | 01:49 | |
| *** bhavikdbavishi has joined #zuul | 02:21 | |
| *** bhavikdbavishi1 has joined #zuul | 03:08 | |
| *** bhavikdbavishi has quit IRC | 03:10 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 03:10 | |
| *** wuchunyang has quit IRC | 03:20 | |
| *** bhavikdbavishi has quit IRC | 03:27 | |
| *** bhavikdbavishi has joined #zuul | 03:35 | |
| *** wuchunyang has joined #zuul | 04:14 | |
| *** wuchunyang has quit IRC | 04:18 | |
| *** bhavikdbavishi has quit IRC | 04:18 | |
| *** bhavikdbavishi has joined #zuul | 04:20 | |
| *** dmellado has quit IRC | 04:25 | |
| *** evrardjp has joined #zuul | 04:33 | |
| swest | ianw: I have a fix for a race condition in diskimage-builder that seems to have made a reappearance (launchpad issue link in commit message) https://review.opendev.org/#/c/728824/ | 04:49 |
|---|---|---|
| ianw | swest: can we update the changelog to be a little more descriptive of what it does? | 05:01 |
| ianw | man say "-u Update partition mappings." ... but doesn't really say when or where or what that means | 05:01 |
| *** wuchunyang has joined #zuul | 05:03 | |
| swest | ianw: ok, will update the changelog, thanks! | 05:16 |
| *** ysandeep|away is now known as ysandeep | 05:22 | |
| *** dustinc has quit IRC | 05:26 | |
| *** saneax_AFK has joined #zuul | 05:29 | |
| *** wuchunyang has quit IRC | 05:37 | |
| *** sgw has quit IRC | 05:42 | |
| *** hashar has joined #zuul | 05:53 | |
| *** bhavikdbavishi1 has joined #zuul | 06:06 | |
| *** bhavikdbavishi has quit IRC | 06:08 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 06:08 | |
| *** bhavikdbavishi has quit IRC | 06:43 | |
| *** wuchunyang has joined #zuul | 06:52 | |
| *** saneax_AFK is now known as saneax_ | 07:00 | |
| *** olaph has quit IRC | 07:06 | |
| *** jcapitao has joined #zuul | 07:07 | |
| *** bhavikdbavishi has joined #zuul | 07:20 | |
| *** bhavikdbavishi has quit IRC | 07:25 | |
| *** bhavikdbavishi has joined #zuul | 07:27 | |
| *** bhavikdbavishi has quit IRC | 07:32 | |
| *** tosky has joined #zuul | 07:32 | |
| *** bhavikdbavishi has joined #zuul | 07:49 | |
| *** rpittau|afk is now known as rpittau | 07:54 | |
| *** jpena|off is now known as jpena | 07:59 | |
| *** ysandeep is now known as ysandeep|afk | 08:04 | |
| mhu | AJaeger, done, thanks for the heads-up | 08:06 |
| *** nils has joined #zuul | 08:14 | |
| AJaeger | thanks, mhu | 08:16 |
| *** sshnaidm|afk is now known as sshnaidm | 08:22 | |
| *** wuchunyang has quit IRC | 08:25 | |
| *** fbo|off is now known as fbo | 08:38 | |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: WIP gitlab: implement git push support https://review.opendev.org/734159 | 08:54 |
| *** hashar has quit IRC | 09:06 | |
| *** bhavikdbavishi has quit IRC | 09:06 | |
| *** bhavikdbavishi has joined #zuul | 09:09 | |
| *** harrymichal has joined #zuul | 09:26 | |
| avass | anyone that has some time over that wants to take a look at: https://review.opendev.org/#/c/733675/ ? :) | 09:31 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: implement git push support https://review.opendev.org/734159 | 09:33 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: implement git push support https://review.opendev.org/734159 | 09:37 |
| *** ysandeep|afk is now known as ysandeep | 09:39 | |
| *** rpittau is now known as rpittau|bbl | 10:11 | |
| *** dennis_effa has joined #zuul | 10:20 | |
| *** dmellado has joined #zuul | 10:28 | |
| *** jcapitao is now known as jcapitao_lunch | 10:29 | |
| *** wuchunyang has joined #zuul | 10:46 | |
| *** wuchunyang has quit IRC | 10:52 | |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: GitHub Reporter: Fix `Reviewed-by` in Merge Commit Message https://review.opendev.org/734580 | 10:57 |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: GitHub Reporter: Fix `Reviewed-by` in Merge Commit Message https://review.opendev.org/734580 | 11:00 |
| openstackgerrit | Sagi Shnaidman proposed zuul/zuul-jobs master: Add ansible collection roles https://review.opendev.org/730360 | 11:03 |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: GitHub Reporter: Fix `Reviewed-by` in Merge Commit Message https://review.opendev.org/734580 | 11:04 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: implement git push support https://review.opendev.org/734159 | 11:07 |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: GitHub Reporter: Fix `Reviewed-by` in Merge Commit Message https://review.opendev.org/734580 | 11:10 |
| openstackgerrit | Sagi Shnaidman proposed zuul/zuul-jobs master: Add jobs for testing ensure-ansible https://review.opendev.org/734584 | 11:12 |
| avass | zbr: just noticed a small spelling error in 690057 :) | 11:15 |
| zbr | :( | 11:16 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: implement git push support https://review.opendev.org/734159 | 11:21 |
| *** threestrands has quit IRC | 11:27 | |
| avass | zuul-jobs-maint: How do we want to enforce variable namespaces, we use tox_executable in ensure-tox, tox and fetch-tox-output (and that makes sense). | 11:31 |
| avass | so doing something like that in https://review.opendev.org/#/c/730360/9 for ansible_galaxy_executable seems to me like it should make more sense than passing three different variables for install/build/publish as well. | 11:32 |
| *** jpena is now known as jpena|lunch | 11:32 | |
| avass | oh, unless we want to namespace it and default the namespaced variable it to the value of ansible_galaxy_executable | 11:35 |
| *** bhavikdbavishi has quit IRC | 11:36 | |
| AJaeger | avass: For such common variables like ansible_galaxy_executable, we could go with the "earliest" - so, tie the namespace to the ensure-X variant. In other words, I think we should pick one good namespace if possible | 11:39 |
| avass | yeah that sounds like one option | 11:52 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: gitlab: implement git push support https://review.opendev.org/734159 | 11:54 |
| *** rfolco has joined #zuul | 11:57 | |
| *** rpittau|bbl is now known as rpittau | 12:11 | |
| *** rfolco is now known as rfolco|rover | 12:11 | |
| *** bhavikdbavishi has joined #zuul | 12:14 | |
| *** jcapitao_lunch is now known as jcapitao | 12:18 | |
| *** bhavikdbavishi has quit IRC | 12:18 | |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: GitHub Reporter: Fix `Reviewed-by` in Merge Commit Message https://review.opendev.org/734580 | 12:21 |
| *** rlandy has joined #zuul | 12:22 | |
| *** Goneri has joined #zuul | 12:25 | |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: GitHub Reporter: Fix `Reviewed-by` in Merge Commit Message https://review.opendev.org/734580 | 12:28 |
| *** Goneri has quit IRC | 12:30 | |
| *** dennis_effa has quit IRC | 12:30 | |
| *** Goneri has joined #zuul | 12:33 | |
| *** sgw has joined #zuul | 12:33 | |
| *** jpena|lunch is now known as jpena | 12:40 | |
| *** bhavikdbavishi has joined #zuul | 12:56 | |
| *** saneax_ is now known as saneax_AFK | 12:59 | |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: web UI: user login with OpenID Connect https://review.opendev.org/734082 | 13:03 |
| tobiash | avass: you're doing windows jobs as well right? | 13:23 |
| tobiash | if that's correct I think you may want to pay attention to https://docs.ansible.com/ansible/latest/user_guide/windows_performance.html | 13:23 |
| tobiash | we had some performance problems regarding huge per task overheads with some images and it turned out that the .net optimization wasn't deterministic in our image builds | 13:24 |
| tobiash | so we run the first in our base playbook and added the second to our image builds | 13:25 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: web UI: user login with OpenID Connect https://review.opendev.org/734082 | 13:28 |
| avass | oh wow, yeah that will be useful | 13:28 |
| avass | most of our windows jobs are currently just one massive task that takes up to an hour to complete because of legacy, but we're planning on breaking it down and that might be needed for that | 13:30 |
| avass | tobiash: thanks! | 13:30 |
| tobiash | avass: this bit us also in pre/post playbooks and especially when using the fetch module | 13:32 |
| tobiash | because the fetch module runs one powershell process per file it fetches... | 13:32 |
| tobiash | and it turned out to be pretty bad fetching html results with multi second powershell startup penalty per file... | 13:33 |
| avass | wow :) | 13:33 |
| tobiash | it was also non-trivial to track this down ;) | 13:34 |
| avass | we're moving away from windows as much as possible, but we're sadly still stuck on windows because of third party software that doesn't have linux versions :( | 13:36 |
| tobiash | I know that pain... | 13:36 |
| avass | we even experimented running things under wine... :) | 13:39 |
| tobiash | ok, we didn't go that far :) | 13:40 |
| *** gmann is now known as gmann_afk | 13:52 | |
| *** bhavikdbavishi has quit IRC | 14:02 | |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: tox: allow tox to be upgraded https://review.opendev.org/690057 | 14:20 |
| mhu | corvus, hello! Do the -1s still hold on https://review.opendev.org/#/c/728061/ and https://review.opendev.org/#/c/728410/ ? | 14:26 |
| corvus | mhu: oh weird, i wonder why that -1 wasn't cleared when you uploaded a new ps | 14:28 |
| corvus | (on 061) | 14:28 |
| corvus | mhu: oh -- 061 seems to have gone back to ps1? | 14:29 |
| corvus | that's neat, i don't think i've ever seen gerrit do that. but it looks correct -- ps4 is the same as ps1, which means my -1 on ps1 is valid again | 14:30 |
| mhu | corvus, ugh, sorry about that, I'll fix it right away | 14:31 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: CLI: Fix errors with the REST client https://review.opendev.org/728061 | 14:33 |
| *** gmann_afk is now known as gmann | 14:36 | |
| mhu | I must have mixed my rebases with my cherrypicks ... which is why I don't want to have long-living topics, I'm prone to errors :) | 14:37 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: REST API: improve tenant scoping of autohold, authorizations https://review.opendev.org/728118 | 14:41 |
| fungi | i didn't realize gerrit would allow you to reupload an identical commit from an earlier patchset | 14:47 |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: web UI: user login with OpenID Connect https://review.opendev.org/734082 | 14:49 |
| corvus | fungi: i think it does if theres an intervening patchset that's different | 14:50 |
| mhu | ugh, most of my patches are affected ... | 14:50 |
| openstackgerrit | Sagi Shnaidman proposed zuul/zuul-jobs master: Add jobs for testing ensure-ansible https://review.opendev.org/734584 | 14:57 |
| *** bhavikdbavishi has joined #zuul | 15:02 | |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: CLI: add autohold-info, autohold-delete via REST https://review.opendev.org/728410 | 15:05 |
| *** jcapitao is now known as jcapitao_afk | 15:09 | |
| avass | corvus: yeah gerrit keeps votes in case you restore changes :) | 15:14 |
| *** olaph has joined #zuul | 15:20 | |
| *** harrymichal has quit IRC | 15:34 | |
| *** y2kenny has joined #zuul | 15:36 | |
| y2kenny | how do you guys usually run a specific unit test for nodepool with tox? (TestDriverStatic.test_static_basic for example) | 15:38 |
| clarkb | y2kenny: tox -epy38 -- test_name_regex_here | 15:38 |
| *** jcapitao_afk is now known as jcapitao | 15:38 | |
| clarkb | for your example I'd do something like `tox -epy38 -- test_static_basic` | 15:38 |
| y2kenny | clarkb: no need for the path to the test file? | 15:38 |
| y2kenny | oh ok | 15:39 |
| y2kenny | thanks | 15:39 |
| clarkb | y2kenny: its a python re.search() regex so it can match anywhere in the test name string iirc | 15:39 |
| clarkb | if there are other tests with the name test_static_basic you can further qualify it with the class name | 15:39 |
| fungi | and you can get fancy with regular expression syntax there if you like | 15:39 |
| *** hashar has joined #zuul | 15:40 | |
| y2kenny | great. That worked, thanks! The solutions on google/stackoverflow includes the test file path for some reason and those didn't work (caused invocation error of some kind.) | 15:40 |
| *** bhavikdbavishi has quit IRC | 15:41 | |
| clarkb | y2kenny: it depends on the test runner itself (not really tox). Nose requires a path iirc others want the python import style paths | 15:44 |
| y2kenny | clarkb: ok... I am totally new to serious python development so I have no idea (only wrote some python scripts here and there.) | 15:45 |
| clarkb | y2kenny: no worries. | 15:45 |
| openstackgerrit | James E. Blair proposed zuul/nodepool master: Add more detail to TESTING.rst https://review.opendev.org/734631 | 15:45 |
| clarkb | thinking out loud here maybe we should add examples oh ya corvus ^ | 15:45 |
| corvus | we had them in the zuul repo, but not nodepool | 15:45 |
| *** y2kenny has quit IRC | 15:46 | |
| clarkb | corvus: avass points out a copy paste thing in that change. Otherwise I think we can land it | 15:47 |
| clarkb | might also want to s/py35/py38/ | 15:47 |
| *** y2kenny has joined #zuul | 15:47 | |
| corvus | clarkb: i thought about that but thought that might could wait until we drop 3.5 | 15:48 |
| clarkb | corvus: wfm | 15:48 |
| openstackgerrit | James E. Blair proposed zuul/nodepool master: Add more detail to TESTING.rst https://review.opendev.org/734631 | 15:48 |
| mhu | corvus, https://review.opendev.org/#/c/728061/ and https://review.opendev.org/#/c/728410/ should be good to go now | 16:09 |
| *** rpittau is now known as rpittau|afk | 16:09 | |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: A Zuul reporter for Elasticsearch https://review.opendev.org/644927 | 16:12 |
| *** jcapitao has quit IRC | 16:24 | |
| *** nils has quit IRC | 16:38 | |
| *** hashar has quit IRC | 16:43 | |
| *** hashar has joined #zuul | 16:46 | |
| zbr | clarkb: can we please do https://review.opendev.org/#/c/731605 ? | 16:47 |
| openstackgerrit | Merged zuul/nodepool master: Add more detail to TESTING.rst https://review.opendev.org/734631 | 16:47 |
| clarkb | zbr: fwiw I agreed with ianw that we should have comments explaining the devitation for expected behavior at least for that particular pause example because the behavior there isn't going to be immediately obvious to most | 16:48 |
| clarkb | zbr: what we want to avoid is someone cleaning that up inappropriately | 16:48 |
| clarkb | and I don't think that is beyond the scope of the change | 16:49 |
| zbr | in last version I added explanation: https://review.opendev.org/#/c/731605/3/test-playbooks/dstat-graph.yaml | 16:50 |
| clarkb | ya I know I just wanted to clarify that. I'm doing a proper review of the other changes really quick before approving | 16:50 |
| zbr | and again: i am not trying to hide the dirt, i am trying to avoid adding more | 16:50 |
| zbr | the fact that this code went in without complaints should be a warning, that we did a poor review. ansible itself is raising warnings on all these calls. | 16:51 |
| clarkb | zbr: out of curiousity what is the issue with https://review.opendev.org/#/c/731605/3/test-playbooks/python/fetch-sphinx-output.yaml that ansible complains about? the file redirection requires shell (command won't work) | 16:52 |
| clarkb | zbr: well the pause example is a good point for why that isn't universally the case | 16:52 |
| clarkb | zbr: and why being better about commenting exceptions is a good idea | 16:52 |
| clarkb | similar qusetion with https://review.opendev.org/#/c/731605/3/test-playbooks/python/fetch-sphinx-tarball.yaml is ansible mad about shell vs command there? | 16:53 |
| zbr | yep, that is why ansible team added the warn param, so user can make an informed decision, usually. | 16:53 |
| corvus | clarkb: have we decided that we do want to lint test-playbooks? | 16:58 |
| clarkb | corvus: I was under the impression we were linting them already? | 16:58 |
| corvus | clarkb: nope, that's the child change of the change you just approved: https://review.opendev.org/731471 | 16:59 |
| clarkb | ah, I didn't realize | 16:59 |
| corvus | i mean, i guess i'm okay with it if everyone else is. but we decided once already not too because ansible-lint produced too many busy-work warnings that we didn't think were important for test-playbooks. | 17:00 |
| clarkb | I see that now we find dirs called "playbooks" not test-playbooks to lint them | 17:00 |
| clarkb | does ansible raise those warnings at runtime too? | 17:01 |
| clarkb | if so then I can see how having clean linting on the test-playbooks would be a good idea | 17:01 |
| zbr | yes it does | 17:01 |
| corvus | yep | 17:01 |
| clarkb | because that will be added noise when trying to udnerstand why a test is not working | 17:01 |
| zbr | and tbh, having two different qa makes no sense, test code is as important as the code itself | 17:01 |
| corvus | frankly, i'm tempted to just turn off that option in ansbile entirely | 17:01 |
| clarkb | corvus: the option to emit warnings? | 17:02 |
| corvus | yep | 17:02 |
| corvus | they've never caused us to do anything other than add "warn: false" | 17:02 |
| zbr | these are there with a reason. | 17:02 |
| zbr | the real issue is that you cannot disable specific warnings, that is the real issue. | 17:03 |
| zbr | because some of them are.... less useful than others. | 17:03 |
| zbr | btw, regarding the idea of switching from py35 to py38 directly, i do not think is wise, i know lots of issues with py38 and the gap is too big (IMHO). | 17:05 |
| clarkb | zbr: zuul supports a range currently 3.5 - 3.8 | 17:06 |
| clarkb | the suggestion was to update the docs to say 3.8 rather than 3.5 since 3.5 will be going away well before 3.8 | 17:06 |
| zbr | clarkb: i guess you are not aware what happens with ansible when you run on 3.8... | 17:06 |
| *** jpena is now known as jpena|off | 17:06 | |
| clarkb | zbr: we are, we had to revert the change to bump zuul to 3.8 though it was working fine for us iirc | 17:06 |
| mordred | clarkb: we rolled back to 3.7 as part of tracking down memory leak | 17:07 |
| clarkb | mordred: I know, but we still test with 3.8 and 3.8 was working for opendev | 17:07 |
| mordred | but then we actually removed jemalloc and that fixed it - so I don't know that we currently have any experience ourselves with a 3.8 issue | 17:07 |
| *** mhayden has left #zuul | 17:07 | |
| clarkb | we rolled back to 3.7 due to the issue that tobiash ? had | 17:07 |
| mordred | and - even with that - the executors run on 3.5 currently | 17:07 |
| clarkb | oh right it was executor specific beacuse ansible | 17:08 |
| mordred | yeah | 17:08 |
| zbr | funny read around py38 and netaddr: https://www.reddit.com/r/ansible/comments/gx202c/ansible_team_closing_and_locking_valid_bugs/ | 17:08 |
| clarkb | zuul does test with 3.8 though including actually running ansible aiui | 17:08 |
| zbr | py36/py37 are ok/safe, but not py38, yet. | 17:08 |
| mordred | in any case - I was at least not personally aware that there was a functional issue with ansible and 3.8 | 17:08 |
| clarkb | (there are functional ansible tests in the zuul test suite) | 17:08 |
| mordred | good that our images are on 3.7 I suppose | 17:08 |
| clarkb | mordred: we rolled back to 3.7 due to that issue | 17:09 |
| avass | corvus: yeah, I can agree that ansible-lint is mostly busy-work, don't think I've fixed something that would have caused a bug somewhere. | 17:09 |
| mordred | clarkb: I disagree | 17:09 |
| avass | well, except for the custom rules that is | 17:09 |
| mordred | clarkb: we rolled back to 3.7 because of the memory leak and never re-rolled forward | 17:09 |
| clarkb | mordred: the memory leak was present in 3.7 and 3.8. We discovered that removing the malloc replacement concurrent with fixing the issue that tobiash raised which is how those two things happened together | 17:10 |
| clarkb | iirc | 17:10 |
| mordred | gotcha. I believe I was more focused on the memory leak and remained unaware that there was an actual functional issue with ansible+3.8 | 17:10 |
| mordred | but now I know :) | 17:10 |
| clarkb | mordred: see Ibfe1b226630c8db202a294590c6435666afcb62d | 17:10 |
| mordred | oh - ah - it's with ansible 2.7 | 17:11 |
| mordred | now I remember this | 17:11 |
| corvus | zbr, clarkb: anyway, just wanted to point out that https://review.opendev.org/731605 is in service of https://review.opendev.org/731471 which has a -1 regarding the utility of linting test-playbooks. i'm not going to -2 it, and if any zuul-jobs-maint expresses explicit support for it, i'll probably return to a +-0. but it does seem a little cart-before-the-horse. | 17:11 |
| corvus | avass: the custom rules are great. :) | 17:11 |
| corvus | (yes we need to merge the cleanup before the enforce; but we just haven't gotten consensus on the enforce yet) | 17:12 |
| mordred | clarkb: incidentally, starting with 2.8 we can _use_ collections, although the ansible-galaxy command cannot install them until 2.9 - but if we install things into the right paths they'll work from 2.8 | 17:12 |
| zbr | corvus: thanks for not fighting against it. | 17:12 |
| clarkb | corvus: I think my biggest thing is avoiding unnecessary noise in testing is desireable. Warnings that look like errors have a habit of being blamed for problems when they are bengign | 17:12 |
| clarkb | and that impacts others ability to debug | 17:12 |
| *** fbo is now known as fbo|off | 17:12 | |
| mordred | so as soon as we drop 2.7 - we can start the "move zuul's extra ansible modules into collections" project | 17:12 |
| zbr | in fact that move can start now, but gradually. | 17:16 |
| corvus | clarkb: understood. i rarely see the output, even when debugging, and like perhaps a sizable plurality of ansible users have just trained myself to ignore the chatty warnings. as you can see from the change, they come up *a lot* in the test playbooks, because we're doing *a lot* of git commands. just doesn't seem worth it to me. | 17:16 |
| avass | mordred: \o/ | 17:17 |
| clarkb | corvus: ya, and as zbr says it would be nice if we could enable deprecation warnings and disable "you're using git wrong" warnings | 17:18 |
| avass | mordred: guess I'll continue working on my ansible collection change soonish | 17:18 |
| corvus | clarkb: i agree with that | 17:18 |
| clarkb | the deprecation warnings are at least useful for thsoe of us trying to keep up with ansible | 17:18 |
| zbr | there is a feature request for this on ansible, i know for sure. | 17:18 |
| avass | aren't deprecation warnings already present in ansible? | 17:19 |
| openstackgerrit | Merged zuul/zuul-jobs master: test-playbooks: avoid warnings with shell/command https://review.opendev.org/731605 | 17:20 |
| clarkb | avass: they are, but they'll get lost if we disable all warnings. They also tend to get lost among the other warnings if you don't clean them up | 17:20 |
| avass | ah yeah | 17:20 |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: Enable linting of test-playbooks https://review.opendev.org/731471 | 17:24 |
| *** hashar has quit IRC | 17:25 | |
| *** saneax_AFK has quit IRC | 17:26 | |
| tobiash | clarkb, mordred: I didn't read all backlog but this reproduced the problem of py38 with the quickstart job that used py38 at that time: https://review.opendev.org/#/c/727155 | 17:27 |
| mordred | tobiash: nod. but it's fine on 2.8 - just not 2.7 yes? | 17:29 |
| fungi | #status log updated intermediate cert bundle for openstackid.org and openstackid-dev.o.o to remove the expired addtrust ca cert, as its presence causes problems for requests on older python versions | 17:35 |
| openstackstatus | fungi: finished logging | 17:35 |
| tobiash | mordred: yes | 17:35 |
| fungi | oops, wrong channel, sorry for the noise | 17:36 |
| avass | corvus: I believe I talked about wanting to have project specific nodes a week or two back. but I realized what I really want is project specific, post-review nodes to get rid of secrets and nested ansible plays. :) | 17:38 |
| avass | or just post-review nodes too be honest, the projects could be in a separate tenant | 17:39 |
| fungi | so the ability to limit what pipelines certain node labels can be used in, rather than allowing anyone to add a particular node label> | 17:41 |
| fungi | ? | 17:41 |
| avass | yeah | 17:42 |
| avass | something like deployment nodes | 17:42 |
| clarkb | would project ssh keys work for that? | 17:42 |
| clarkb | similar to how we use zuul for CD in opendev | 17:42 |
| fungi | we've effectively done that in our deployment by authorizing project-specific ssh keys on nodes | 17:42 |
| clarkb | we add host at job runtime off of the executor, then the job content runs against the target host | 17:43 |
| clarkb | fungi: jinx | 17:43 |
| fungi | yeah, it doesn't stop anyone from adding those to their jobs, but their jobs won't be able to actually connect to those nodes | 17:43 |
| avass | I guess you keep the ssh key as a secret ? | 17:45 |
| clarkb | avass: its built into zuul, each project has a project specific ssh key | 17:46 |
| clarkb | avass: you can query the api for the public key material for a specific project, add that to your hosts, and then jobs running on that project can use the key (and its post review only iirc) | 17:47 |
| avass | that's the same key that is used for encryption right? | 17:47 |
| avass | didn't know you had access to that during a job | 17:48 |
| *** dustinc has joined #zuul | 17:48 | |
| avass | I guess I missed this: https://zuul-ci.org/docs/zuul/reference/jobs.html#project-key | 17:48 |
| *** ysandeep is now known as ysandeep|away | 17:49 | |
| avass | well, that solves that. thanks :) | 17:50 |
| corvus | avass: sorry i was away for your question, but glad you found a solution :) | 17:52 |
| corvus | avass: we also have tenant keys on the roadmap, but not yet implemented | 17:53 |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: tox: allow tox to be upgraded https://review.opendev.org/690057 | 17:54 |
| y2kenny | so I was just looking at how nodepool get its information to the executor. Looks like everything is passed via "connection_port" and then there are some special logics in the executor that maps nodepool's data structure into executor/ansible's data structure. Is that a correct understanding? | 18:05 |
| *** Defolos has joined #zuul | 18:06 | |
| *** Defolos has quit IRC | 18:06 | |
| *** sshnaidm is now known as sshnaidm|afk | 18:07 | |
| zbr | avass: was my last comment on https://review.opendev.org/#/c/703065/9/roles/revoke-sudo/tasks/main.yaml ok? | 18:10 |
| avass | zbr: yeah | 18:13 |
| avass | zbr: but I believe the current idea is to be able to define that as a nodepool variable, same with toggling whether to revoke sudo or not. | 18:14 |
| avass | but I'm not sure how much work is needed to be able to pass variables from nodepool.yaml to ansible | 18:15 |
| avass | something like: http://paste.openstack.org/show/794540/ where there would be a nodepool.<var> like there are zuul.<var> | 18:21 |
| avass | but I'm not sure if there should be specific parameters that can be set in nodepool or if it should be possible to set any variable like that paste | 18:23 |
| avass | or if maybe they should be stored in zuul.nodepool.<var> since 'nodepool' could already be used somewhere | 18:31 |
| *** iurygregory has quit IRC | 18:43 | |
| mordred | avass: we set a nodepool dict on every host | 18:44 |
| mordred | avass: https://zuul.opendev.org/t/openstack/build/84cd43655f734934a4e232c2f2dd9667/log/zuul-info/inventory.yaml#9 | 18:45 |
| avass | mordred: ah, nice :) | 18:45 |
| *** iurygregory has joined #zuul | 18:47 | |
| mordred | avass: also - didn't we have a discussion in the past about setting a flag on hosts in nodepool to allow driving something like that? | 18:48 |
| avass | mordred: yeah | 18:48 |
| avass | mordred: I just haven't looked into it yet | 18:48 |
| mordred | kk. just making sure I wasn't misremembering | 18:48 |
| mordred | it seems like a valid use case :) | 18:49 |
| fungi | we've used this, for example, to cross reference job failures to specific node providers, or even to match certain log messages to specific host_id values (which can identify a particular bad hypervisor host) | 18:50 |
| Shrews | ohai! friendly fyi for an easy fix for zuul callback plugins someone might want to put together: https://github.com/ansible/ansible/issues/69954 | 19:07 |
| *** ianychoi__ has joined #zuul | 19:16 | |
| *** ianychoi_ has quit IRC | 19:19 | |
| guillaumec | Shrews: hello! it does indeed :) | 19:28 |
| mordred | Shrews: fascinating | 19:29 |
| mordred | Shrews: and thanks! | 19:29 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Rework quick-start and prepare for other tutorials https://review.opendev.org/732066 | 19:54 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "gate your first patch" https://review.opendev.org/732067 | 19:54 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "Use zuul jobs" https://review.opendev.org/732068 | 19:54 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "gate pipeline" https://review.opendev.org/732069 | 19:54 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "job secrets" https://review.opendev.org/732070 | 19:54 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "job dependencies" https://review.opendev.org/732071 | 19:54 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "Use zuul jobs" https://review.opendev.org/732068 | 20:08 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "gate pipeline" https://review.opendev.org/732069 | 20:08 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "job secrets" https://review.opendev.org/732070 | 20:08 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "job dependencies" https://review.opendev.org/732071 | 20:08 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Detect Gerrit gate pipelines with the wrong connection https://review.opendev.org/733929 | 20:27 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Rework quick-start and prepare for other tutorials https://review.opendev.org/732066 | 21:10 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Rework quick-start and prepare for other tutorials https://review.opendev.org/732066 | 21:50 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "gate your first patch" https://review.opendev.org/732067 | 21:50 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "Use zuul jobs" https://review.opendev.org/732068 | 21:50 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "gate pipeline" https://review.opendev.org/732069 | 21:50 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "job secrets" https://review.opendev.org/732070 | 21:50 |
| openstackgerrit | Guillaume Chauvel proposed zuul/zuul master: tutorial: Add "job dependencies" https://review.opendev.org/732071 | 21:50 |
| *** rfolco|rover has quit IRC | 22:05 | |
| *** armstrongs has joined #zuul | 22:38 | |
| *** armstrongs has quit IRC | 22:47 | |
| *** tosky has quit IRC | 22:59 | |
| *** ysandeep|away is now known as ysandeep | 23:42 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!