| *** saneax has quit IRC | 00:13 | |
| *** rfolco has quit IRC | 01:29 | |
| *** swest has quit IRC | 01:32 | |
| *** ysandeep|away is now known as ysandeep|rover | 01:44 | |
| *** swest has joined #zuul | 01:46 | |
| *** ysandeep|rover is now known as ysandeep|rover|b | 02:36 | |
| *** Goneri has quit IRC | 03:49 | |
| *** bhavikdbavishi has joined #zuul | 03:50 | |
| *** bhavikdbavishi1 has joined #zuul | 03:53 | |
| *** bhavikdbavishi has quit IRC | 03:54 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 03:54 | |
| *** ysandeep|rover|b is now known as ysandeep|rover | 04:19 | |
| *** evrardjp has quit IRC | 04:36 | |
| *** evrardjp has joined #zuul | 04:37 | |
| *** y2kenny has quit IRC | 05:43 | |
| *** dpawlik has joined #zuul | 06:06 | |
| *** jamesmcarthur has quit IRC | 06:07 | |
| *** saneax has joined #zuul | 06:08 | |
| *** igordc has quit IRC | 06:21 | |
| *** avass is now known as Guest53619 | 06:52 | |
| *** avass has joined #zuul | 06:52 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Adds roles to install and run hashicorp packer https://review.opendev.org/709292 | 06:57 |
|---|---|---|
| *** jcapitao has joined #zuul | 07:06 | |
| tobiash | tristanC, corvus: how shall we proceed with https://review.opendev.org/619156 (zk auth). Can we help there? | 07:11 |
| tobiash | bolg: ^ | 07:11 |
| *** jpena|off is now known as jpena | 07:16 | |
| *** bhavikdbavishi has quit IRC | 07:23 | |
| *** gtema has joined #zuul | 07:23 | |
| *** jamesmcarthur has joined #zuul | 07:41 | |
| *** tosky has joined #zuul | 07:44 | |
| *** jamesmcarthur has quit IRC | 07:47 | |
| openstackgerrit | Daniel Pawlik proposed zuul/zuul-jobs master: [DNM] - testing updated software-factory images https://review.opendev.org/717210 | 07:48 |
| *** rpittau|afk is now known as rpittau | 07:56 | |
| *** ysandeep|rover is now known as ysandeep|rover|l | 08:03 | |
| *** bhavikdbavishi has joined #zuul | 08:04 | |
| *** dpawlik has quit IRC | 08:11 | |
| *** dpawlik has joined #zuul | 08:11 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Make reporting asynchronous https://review.opendev.org/691253 | 08:25 |
| *** harrymichal has joined #zuul | 08:27 | |
| *** ysandeep|rover|l is now known as ysandeep|lunch | 08:32 | |
| *** jamesmcarthur has joined #zuul | 08:43 | |
| *** jamesmcarthur has quit IRC | 08:48 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Make reporting asynchronous https://review.opendev.org/691253 | 08:50 |
| *** rmart04 has joined #zuul | 08:50 | |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: Add support for ansible callbacks https://review.opendev.org/717260 | 09:00 |
| *** hashar has joined #zuul | 09:03 | |
| *** jamesmcarthur has joined #zuul | 09:07 | |
| *** ysandeep|lunch is now known as ysandeep|rover | 09:09 | |
| *** jamesmcarthur has quit IRC | 09:12 | |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: Add support for RedHat platforms on install-podman https://review.opendev.org/716578 | 09:14 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: WIP: Store unparsed branch config in Zookeeper https://review.opendev.org/705716 | 09:29 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Prepare Zookeeper for scale-out scheduler https://review.opendev.org/717269 | 09:29 |
| zbr | tobiash: tristanC : can you help me with https://review.opendev.org/#/c/690057 ? | 09:30 |
| zbr | is see additional changes proposed to this role and multi-platform testing is really needed. | 09:31 |
| *** jamesmcarthur has joined #zuul | 09:47 | |
| *** jamesmcarthur has quit IRC | 09:55 | |
| *** jamesmcarthur has joined #zuul | 10:04 | |
| *** bhavikdbavishi has quit IRC | 10:25 | |
| *** rpittau is now known as rpittau|bbl | 10:29 | |
| *** jcapitao is now known as jcapitao_lunch | 10:40 | |
| *** jamesmcarthur has quit IRC | 10:47 | |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: Add support for ansible callbacks https://review.opendev.org/717260 | 11:00 |
| *** bhavikdbavishi has joined #zuul | 11:03 | |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: Add support for ansible callbacks https://review.opendev.org/717260 | 11:04 |
| *** ysandeep|rover is now known as ysandeep|break | 11:05 | |
| *** jamesmcarthur has joined #zuul | 11:14 | |
| *** jamesmcarthur has quit IRC | 11:20 | |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: Add support for ansible callbacks https://review.opendev.org/717260 | 11:31 |
| *** jpena is now known as jpena|lunch | 11:38 | |
| *** ysandeep|break is now known as ysandeep|rover | 11:48 | |
| tristanC | zbr: commented | 11:53 |
| tristanC | tobiash: corvus: iirc we decided to not implement zk-auth as tls is enough to prevent annonymous access (which is consistent with gearman). Thus, unless it is still benefical, I would like to abandon the zk-auth changes. | 11:55 |
| tobiash | tristanC: ah thanks, I lost track on this topic | 11:55 |
| tristanC | tobiash: i did an initial zk tls setup using the shellscript in the operator, and we decided that using the cert-manager crd would be better. | 12:01 |
| tristanC | tobiash: would you mind having a look at this change, it is quite useful to work with the operator integration tests locally: https://review.opendev.org/#/c/714163/6 | 12:02 |
| *** jcapitao_lunch is now known as jcapitao | 12:07 | |
| *** rpittau|bbl is now known as rpitau | 12:08 | |
| *** rpitau is now known as rpittau | 12:09 | |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Prepare Zookeeper for scale-out scheduler https://review.opendev.org/717269 | 12:10 |
| tobiash | tristanC: lgtm | 12:10 |
| tobiash | tristanC: I've responded to your question on https://review.opendev.org/708812 | 12:14 |
| tristanC | tobiash: according to https://opendev.org/zuul/zuul/src/branch/master/tox.ini#L59 , mypy is not running in the tests | 12:15 |
| tobiash | tristanC: ah you mean the test folder | 12:15 |
| tobiash | good catch | 12:15 |
| *** jamesmcarthur has joined #zuul | 12:15 | |
| tristanC | tobiash: and my second issue is that mypy is not checking code inside untyped function, unless we run it in --strict. thus it seems like some annotation are not validated and they may become incorrect | 12:17 |
| tristanC | tobiash: however, that is great to hear they are already useful within some IDE, thus their addition lgtm | 12:19 |
| tobiash | well strict is not possible unless we annotate all of zuul | 12:19 |
| *** jamesmcarthur has quit IRC | 12:20 | |
| zbr | tristanC: how about running "pip install tox" regardless if "tox" command is present or not? | 12:21 |
| zbr | i am not aware of other way to assure it is upgraded when needed | 12:21 |
| tristanC | tobiash: for example, those annotations are not tested by mypy https://review.opendev.org/#/c/709735/8/zuul/scheduler.py@340 , because the constructor arguments are not typed | 12:22 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: WIP: Store unparsed branch config in Zookeeper https://review.opendev.org/705716 | 12:22 |
| tristanC | zbr: perhaps if the upgrade_tox toggle is set, then the role should run the upgrade command, whenever tox is already installed or not | 12:23 |
| zbr | is not so easy, you could also have additional constraints | 12:23 |
| zbr | if you do a "tox>1.2.3 tox-pyenv tox-docker" kind of stuff. | 12:24 |
| tristanC | zbr: then it's no longer just upgrading tox is it? | 12:24 |
| zbr | tristanC: tox plugins are key for running tox in certain use-cases, i do see them as part of the same tool | 12:26 |
| bolg | tristanC: I am very in favor of using those annotations (https://review.opendev.org/708812) it really helps navigating the code. What IDE are you using? | 12:27 |
| zbr | default is w/o plugins, just "tox", but others will need to enable extra plugins. for example core tox is unable to find pyenv pythons. | 12:27 |
| tristanC | zbr: understood, but this is triggered by a variable named 'upgrade_tox', which sounds missleading if it is used to install tox plugins | 12:27 |
| zbr | also there is a plugin named tox-venv which enable it to use venv instead of the more problematic virtualenv :D | 12:28 |
| *** rlandy has joined #zuul | 12:28 | |
| tristanC | zbr: or at least the commit message says so | 12:28 |
| bolg | tristanC, tobias: my approach would be to add them part by part as it is touched | 12:28 |
| zbr | tristanC: i did not say that your remark was not correct, but i am looking for a more reliable solution | 12:29 |
| zbr | in the end I could add a conditonal: if params are unchanged, use old behavior, else do the new. | 12:29 |
| zbr | at least, this approach would minimize risks | 12:29 |
| bolg | tristanC: regarding https://review.opendev.org/#/c/709735/8/zuul/scheduler.py@340 yes you are right, buy IDE is at least not confusing all possible classes implementing those properties :) | 12:30 |
| tristanC | bolg: bolg: don't get me wrong, i am a big fan of type annotation. I'm using emacs with flycheck-mypy, which is using the mypy check implementation, and without the `--strict` argument, mypy simply skip function that are not typed, thus i don't benefit from https://review.opendev.org/#/c/709735/8/zuul/scheduler.py@340 | 12:31 |
| tristanC | bolg: and regarding the addition to test file, since zuul ci doesn't check them (mypy only run on the zuul/ directory, not the tests/), then there is a risk the annotation bitrot | 12:32 |
| bolg | tristanC: undestood. I am trying to add them to the relevant code whenever i touch it. Regarding the test, yeah I agree there are bit redundant there | 12:34 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Improve typings for driver event ingestion https://review.opendev.org/717297 | 12:36 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: WIP: Driver event ingestion https://review.opendev.org/717299 | 12:43 |
| *** jpena|lunch is now known as jpena | 12:45 | |
| avass | corvus: I started with this: https://review.opendev.org/#/c/717260/4 but I get the feeling that the way we write ansible.cfg should be overhauled instead | 12:49 |
| *** jamesmcarthur has joined #zuul | 12:51 | |
| avass | corvus: let's say someone wants to use the full_skip callback: https://docs.ansible.com/ansible/latest/plugins/callback/full_skip.html | 12:55 |
| *** jamesmcarthur has quit IRC | 12:56 | |
| avass | corvus: then we'd have to make sure the config is added to the [defaults] section in ansible.cfg and not create another section called [defaults] | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Scheduler test app manager https://review.opendev.org/708812 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Use scheduler manager consistently in tests https://review.opendev.org/709542 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Refactor executor_client in tests https://review.opendev.org/709672 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Refactor merge_client in tests https://review.opendev.org/709676 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Refactor nodepool in tests https://review.opendev.org/709703 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Refactor zookeeper in tests https://review.opendev.org/709709 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Consolidate scheduler pause/exit as hibernation https://review.opendev.org/709723 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Refactor `self.event_queues` in tests https://review.opendev.org/709990 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Scheduler's pause/resume functionality https://review.opendev.org/709735 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Separate connection registries in tests https://review.opendev.org/712958 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Prepare Zookeeper for scale-out scheduler https://review.opendev.org/717269 | 12:56 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: WIP: Store unparsed branch config in Zookeeper https://review.opendev.org/705716 | 12:56 |
| avass | so instead of adding a [ansible_callback "<callback_name>"] to zuul.conf we could just add [ansible_config "<ansible.cfg section name>"] that adds data to that section | 12:57 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Improve typings for driver event ingestion https://review.opendev.org/717297 | 12:57 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: WIP: Driver event ingestion https://review.opendev.org/717299 | 12:57 |
| zbr | tristanC: thanks for ensure-tox comments, going back to implement them. Maybe more important is https://review.opendev.org/#/c/716578/ re install-podman | 12:58 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect merger to Zookeeper https://review.opendev.org/716221 | 12:59 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect executor to Zookeeper https://review.opendev.org/716262 | 12:59 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect fingergw to Zookeeper https://review.opendev.org/716875 | 12:59 |
| avass | corvus: that would make timeouts/retries and things like that configurable as well | 12:59 |
| *** rfolco has joined #zuul | 13:01 | |
| bolg | tristanC, tobiash, corvus: thanks for your reviews on the scale-out-scheduler topic, since most of them just prepare tests and do not introduce breaking changes, do you think we could go forward with them or something missing? | 13:02 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: Add phoronix-test-suite job https://review.opendev.org/679082 | 13:03 |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: tox: allow tox to be upgraded https://review.opendev.org/690057 | 13:11 |
| tristanC | bolg: the topic looks solid to me, at least until the `Scheduler's pause/resume functionality` change | 13:12 |
| bolg | tristanC: yep thats what i meant :) till there we could go forward with it | 13:13 |
| bolg | tristanC: thank for the comments on the pause/resume I will look into it | 13:19 |
| *** bhavikdbavishi has quit IRC | 13:26 | |
| tobiash | tristanC, corvus: I've added a question on zk tls (https://review.opendev.org/712531) | 13:42 |
| openstackgerrit | Merged zuul/zuul-operator master: Add nodepool external config https://review.opendev.org/715311 | 13:52 |
| openstackgerrit | Merged zuul/zuul-operator master: Adapt the integration playbook to be usable locally https://review.opendev.org/714163 | 13:52 |
| *** jamesmcarthur has joined #zuul | 13:52 | |
| mordred | tobiash: I believe it can be used directly - if you see https://review.opendev.org/#/c/712817/7/doc/source/examples/zoo.cfg | 13:53 |
| mordred | tobiash: (iirc, this was a thing that was discovered as the work progressed, that pem encoding as an option in addition to jks format | 13:53 |
| AJaeger | here's a zuul-jobs change to add phoronix test jobs, it had previously two +2s and got a small update- anybody wants to review, please? https://review.opendev.org/#/c/679082/ | 13:55 |
| *** zxiiro has joined #zuul | 13:57 | |
| openstackgerrit | Merged zuul/zuul-operator master: Add nodepool kubernetes pod label to integration test https://review.opendev.org/715316 | 13:57 |
| *** jamesmcarthur has quit IRC | 13:57 | |
| *** Goneri has joined #zuul | 13:58 | |
| *** ysandeep|rover is now known as ysandeep|away | 13:59 | |
| tobiash | mordred: ah cool, thx | 14:10 |
| tristanC | tobiash: would you mind having a look to https://review.opendev.org/706639 it's a spec update to define job volume for the zuul CRD | 14:23 |
| tobiash | tristanC: do you want to change to camelcase in the spec as well? | 14:25 |
| *** cdearborn has joined #zuul | 14:25 | |
| tristanC | tobiash: yep, that's done in https://review.opendev.org/#/c/707190/1 | 14:27 |
| tobiash | oh that's the followup | 14:27 |
| tobiash | k | 14:27 |
| openstackgerrit | Merged zuul/zuul master: Scheduler test app manager https://review.opendev.org/708812 | 14:28 |
| openstackgerrit | Merged zuul/zuul master: Use scheduler manager consistently in tests https://review.opendev.org/709542 | 14:28 |
| openstackgerrit | Merged zuul/zuul master: Refactor executor_client in tests https://review.opendev.org/709672 | 14:28 |
| openstackgerrit | Merged zuul/zuul master: Refactor merge_client in tests https://review.opendev.org/709676 | 14:28 |
| openstackgerrit | Merged zuul/zuul master: Refactor nodepool in tests https://review.opendev.org/709703 | 14:28 |
| openstackgerrit | Merged zuul/zuul master: Refactor zookeeper in tests https://review.opendev.org/709709 | 14:29 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Add release note for zookeeper tls support https://review.opendev.org/717312 | 14:29 |
| tobiash | tristanC, corvus, mordred: does that release note look right for zk tls? ^ | 14:30 |
| *** jamesmcarthur has joined #zuul | 14:30 | |
| *** y2kenny has joined #zuul | 14:30 | |
| corvus | avass: hrm, do we really want to open the possibility of setting the "defaults" section? we've wanted very tight control over how zuul interacts with ansible. i think maybe anything that goes in the defaults section is something we should keep hard-coded in the executor for now.... | 14:31 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Add release note for zookeeper tls support https://review.opendev.org/717312 | 14:31 |
| corvus | tobiash: that note looks good, but i also wanted to make the note in zuul have doc links; and then maybe even go back and revise the nodepool note to include them as well. | 14:33 |
| mnaser | so, an update on the zuul website gatsby-ifying | 14:35 |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: Add support for RedHat platforms on install-podman https://review.opendev.org/716578 | 14:35 |
| corvus | tobiash: the other thing to know about the zk tls stack is that we hit a problem with the intermediate registry which caused https://review.opendev.org/712817 to fail; i've added debugging to the registry server to try to track down the problem if it happens again, but the problem is on upload, and that blob is now permanently corrupt. we can probably rebase it after its parents land and it might work then | 14:36 |
| corvus | (however, it passed previous testing, so i'm confident in it) | 14:36 |
| mnaser | https://i.imgur.com/OMVUC28.png -- this is what it looks like, and that's gatsby-ified :) | 14:36 |
| corvus | avass: i sent you a private message here in irc; can you check that you got it? | 14:36 |
| mnaser | i will push up a patch shortly with the wip and i think i'll takea that opportunity to add some stuff to zuul/zuul-jobs to build a gatsby site | 14:37 |
| corvus | mnaser: that looks pretty much the same to my eyes :) | 14:37 |
| mnaser | corvus: that's exactly the goal ;) | 14:37 |
| tobiash | corvus: ah ok | 14:39 |
| corvus | tobiash: i think maybe just link to the howtos/zookeeper.rst file? | 14:40 |
| *** EmilienM is now known as EvilienM | 14:40 | |
| corvus | i guess we'll need an anchor at the top of that to :ref | 14:41 |
| tobiash | corvus: yes, fighting with local doc build on mac... | 14:41 |
| corvus | tobiash: also could link to :attr:`zookeeper` maybe | 14:41 |
| AJaeger | mnaser: want to review 679082 again, please? | 14:42 |
| *** jamesmcarthur has quit IRC | 14:44 | |
| *** jamesmcarthur has joined #zuul | 14:44 | |
| AJaeger | corvus, zuul-maint, do we want to merge the stack for changing the roles install-X to ensure-X by avass starting at https://review.opendev.org/#/c/716663/ - so, merge quickly before we get more merge conflicts. Or want to wait until a certain point? | 14:45 |
| *** sgw has joined #zuul | 14:49 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Add release note for zookeeper tls support https://review.opendev.org/717312 | 14:50 |
| *** avass is now known as Guest97814 | 14:50 | |
| *** avass has joined #zuul | 14:50 | |
| corvus | AJaeger: i was thinking maybe start merging those on monday? give folks a chance to raise any concerns on the mailing list | 14:50 |
| avass | coruvs: ah sorry, logged off my work computer | 14:52 |
| AJaeger | corvus: that works - let me leave a comment on the first change saying that | 14:52 |
| avass | corvus: I'm not sure why we want such tight control over ansible.cfg to be honest | 14:53 |
| corvus | avass: most of the settings we have set there are due to lessons we learned the hard way. things about fact caching, and retries, etc. i'm open to allowing more customization, but maybe if there isn't a pressing need, we should keep control so we can make sure it works reliably well for all users. | 14:55 |
| * tobiash remembers many lessons we learned the hard way | 14:57 | |
| *** jamesmcarthur has quit IRC | 15:00 | |
| *** jamesmcarthur has joined #zuul | 15:00 | |
| avass | corvus: well that seems like different things to me, it's still possible to default to values that are known to work well and recommend users to not touch those, or blacklist something | 15:01 |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: tox: allow tox to be upgraded https://review.opendev.org/690057 | 15:01 |
| avass | corvus: it's not like it's needed, but adding configuration for callbacks exclusively seems like something that can be handled in a more general way | 15:02 |
| corvus | avass: there's a risk here -- the more users customize that, the harder it is for us to know that it's going to work well and that any future changes we make aren't going to break. i'm not completely opposed, but i think we should have a very compelling use case. | 15:02 |
| corvus | avass: if you did the [ansible_config "section_name"] approach, you'd still have to merge the callback whitelist | 15:05 |
| avass | corvus: no since the callback_whitelist would be in [ansible_config "defaults"] :) | 15:06 |
| avass | unless there's a callback hardcoded in zuul of course | 15:06 |
| corvus | avass: then you'd have to hope the user whitelists the zuul callbacks | 15:06 |
| corvus | there is | 15:06 |
| *** klindgren_ has quit IRC | 15:06 | |
| corvus | i mean, it's not impossible, i'm just pointing out that the general approach isn't necessarily "cleaner" or "easier" | 15:06 |
| *** klindgren has joined #zuul | 15:07 | |
| corvus | avass: i like the approach in 717260 | 15:07 |
| avass | corvus: we'd have to make sure that callback_config_section can't be set to [defaults] or [ssh_connection] in that case | 15:09 |
| avass | I guess the easy approach is to only allow [callback_<name>] | 15:09 |
| corvus | what's the difference between callback_plugins and callback_whitelist? | 15:10 |
| *** jcapitao has quit IRC | 15:11 | |
| avass | corvus: not sure, but I was wondering why ara didn't need to be whitelisted | 15:11 |
| corvus | oh _plugins is the path, right? | 15:11 |
| avass | corvus: that's my guess | 15:11 |
| corvus | ah, the zuul plugin is enabled by setting stdout_callback, so that doesn't collide after all | 15:12 |
| corvus | avass: iiuc, i think maybe builtin plugins need to be enabled by adding to whitelist, but non-built-in-plugins only need to be in the callback_plugins path? | 15:13 |
| *** yoctozepto has quit IRC | 15:17 | |
| avass | corvus: something else I think we want but not necessary need is to add env variables to ansible or configure callbacks per tenant | 15:18 |
| avass | I'll check | 15:18 |
| corvus | avass: we might need to merge callback_plugins no matter what | 15:18 |
| avass | corvus: my guess is yes: https://docs.ansible.com/ansible/latest/reference_appendices/config.html#default-callback-whitelist | 15:18 |
| avass | corvus: that's if we don't want to force people to install the callbacks in the callback directory :) | 15:18 |
| *** yoctozepto8 has joined #zuul | 15:18 | |
| corvus | avass: it's probably worth looking at how the ara plugin is installed and how we find out its directory. see getAraCallbackPlugin | 15:18 |
| corvus | i think we'd want folks to install plugins by adding them to the extra packages that zuul-manage-ansible installs | 15:18 |
| avass | I'll take a look at that | 15:21 |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: Add support for ansible callbacks https://review.opendev.org/717260 | 15:26 |
| avass | corvus: ^ how about we do that instead? | 15:26 |
| avass | corvus: and we only support adding [callback_$foo] to ansible.cfg for now | 15:27 |
| corvus | avass: sounds good to me | 15:27 |
| avass | corvus: do we want to filter the name in [ansible_callback "<name>"] so it's not possible to add something to [defaults], or should we ust not support that? | 15:29 |
| avass | ^ since I guess you can do [ansible_callback "nocows = False"] otherwise :) | 15:31 |
| *** igordc has joined #zuul | 15:32 | |
| corvus | avass: oh i thought your change would only write out "[callback_nocows = False]" in that case? | 15:32 |
| avass | corvus: it still needs to add it to the whitelist so it would be something like callback_whitelist =\n nocows = False | 15:33 |
| avass | with nocows = False being indented of course, but I think I checked yesterday and configloader still sees that as a separate value as long as there's an equal sign in there | 15:35 |
| fungi | i'm glad i'm not the only one who appreciates cows | 15:36 |
| * mordred feels like we're missing olaph | 15:36 | |
| corvus | avass: ok. well, we should probably filter it anyway just to keep anything unexpected from happening :) | 15:37 |
| fungi | yep, i miss him every day | 15:37 |
| avass | corvus: I'm looking forward to the "ensureNoCows" testcase :) | 15:37 |
| corvus | mordred: have a way to send him a message? http://paste.openstack.org/show/791596/ | 15:39 |
| corvus | avass: heh, i'm kind of surprised we don't have one already :) | 15:40 |
| corvus | though, i bet at this point, if we turned on cows, a bunch of tests would break | 15:40 |
| *** rpittau is now known as rpittau|afk | 15:43 | |
| mordred | corvus: sent! | 15:44 |
| *** gtema has quit IRC | 15:48 | |
| *** ysandeep|away is now known as ysandeep | 15:49 | |
| *** ysandeep is now known as ysandeep|away | 15:56 | |
| y2kenny | For the build summary page, the web interface fetch zuul-manifest.json from the log server. Does the log server have to be the same as the web UI server? I am getting "Network Error (Unable to fetch URL...)" even though I can copy and paste the url and the file is there. | 16:10 |
| fungi | y2kenny: it does not have to be the same (we certainly don't run it that way in opendev) but you may need to do some work to convince browsers it's not a cross-site/cross-domain exploit | 16:12 |
| fungi | typically using cors headers | 16:12 |
| tobiash | mordred: did you validate that the ansi color code removal worked for the line comments? | 16:12 |
| tobiash | judging from my test case the regex doesn't seem to match the escape sequences | 16:13 |
| mordred | tobiash: yes - lemme see if I can find the patch | 16:13 |
| y2kenny | I was wondering about that but I thought fetch/get is not part of the cross-site stuff (but I haven't touch those in a long time.) fungi, when you sing using cors headers, which part of the system do I need to modify? | 16:14 |
| mordred | tobiash: https://review.opendev.org/#/c/716604/ | 16:14 |
| mordred | tobiash: the pylint comments there had ansi codes | 16:14 |
| y2kenny | said* | 16:14 |
| tobiash | weird | 16:14 |
| mordred | y2kenny: the system hosting the log files | 16:14 |
| y2kenny | ok | 16:15 |
| fungi | y2kenny: i think it'll be on the site where the logs are hosted, so say that requests from $other_site are allowed | 16:15 |
| mordred | y2kenny: it needs to publish cors headers allowing the domain where the dashboard is hosted | 16:15 |
| clarkb | y2kenny: the webserver serving the content (log server in your case?) needs to set access-control-allow-origin header to include the domain that is retrieving the content (the dashboard url) or you can * | 16:15 |
| tobiash | mordred: oh, looks like this missed the binary char in front of the sequence https://review.opendev.org/#/c/716623/4/roles/tox/files/cinder-pylint-output.txt | 16:15 |
| y2kenny | I will dig into that. Thanks fungi, mordred, clarkb | 16:15 |
| fungi | you bet, glad i knew the answer for a change ;) | 16:16 |
| mordred | tobiash: oh - might have been copy-paste :( | 16:16 |
| tobiash | yeah, something seems to have filtered it | 16:17 |
| corvus | mordred, tristanC: in https://review.opendev.org/716305 i noticed in the preview that the refresh button has changed; do we want to try to get it to look like it currently does, or go with the new look? | 16:32 |
| tristanC | y2kenny: is there something setting the Content-Security-Policy header on your zuul-web ui? | 16:32 |
| tristanC | corvus: oh indeed... well i don't mind the new look :) | 16:33 |
| tristanC | there is proably a css expression that would make it look better, but i don't know which one | 16:35 |
| mordred | corvus: yeah - I don't mind the new look, but also would be fine trying to get it to match the old look | 16:35 |
| *** evrardjp has quit IRC | 16:36 | |
| *** evrardjp has joined #zuul | 16:37 | |
| corvus | tobiash: which patch were you running with that caused a leak? | 16:37 |
| corvus | tobiash: memory leak | 16:37 |
| tobiash | corvus: https://review.opendev.org/714852 | 16:38 |
| corvus | tobiash: ok. i was about to look at https://review.opendev.org/712939 and wanted to make sure that wasn't it :) | 16:38 |
| tobiash | it hasn't merged | 16:38 |
| tobiash | corvus: that one is in prod since three weeks :) | 16:39 |
| *** y2kenny has quit IRC | 16:41 | |
| *** yoctozepto8 is now known as yoctozepto | 16:43 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul-jobs master: Support multiple matchers when parsing tox output https://review.opendev.org/716263 | 16:46 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul-jobs master: Don't silently ignore exceptions when parsing tox output https://review.opendev.org/716766 | 16:46 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul-jobs master: Strip source dir from file comments https://review.opendev.org/716264 | 16:47 |
| tobiash | mnaser: ^ | 16:47 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul-jobs master: Ignore absolute paths after stripping work dir https://review.opendev.org/717042 | 16:47 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul-jobs master: Add test cases for tox line comment parsing https://review.opendev.org/717341 | 16:47 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-operator master: Update dhall-kubernetes version https://review.opendev.org/717342 | 16:47 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-operator master: Update dhall-kubernetes version to v4.0.0 https://review.opendev.org/717342 | 16:51 |
| *** bhavikdbavishi has joined #zuul | 16:57 | |
| *** gtema has joined #zuul | 17:11 | |
| *** jpena is now known as jpena|off | 17:14 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-operator master: Add job_volumes CR spec attribute https://review.opendev.org/706642 | 17:14 |
| *** bhavikdbavishi1 has joined #zuul | 17:18 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-operator master: Update attributes to camelCase https://review.opendev.org/707193 | 17:19 |
| *** gtema has quit IRC | 17:19 | |
| *** gtema has joined #zuul | 17:20 | |
| *** bhavikdbavishi has quit IRC | 17:20 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 17:20 | |
| *** harrymichal has quit IRC | 17:21 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul-jobs master: Add test cases for tox line comment parsing https://review.opendev.org/717341 | 17:22 |
| *** gtema has quit IRC | 17:25 | |
| *** jamesmcarthur has quit IRC | 17:26 | |
| *** jamesmcarthur has joined #zuul | 17:29 | |
| *** dpawlik has quit IRC | 17:30 | |
| *** jamesmcarthur has quit IRC | 17:33 | |
| *** jamesmcarthur has joined #zuul | 17:34 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul-jobs master: Add test cases for tox line comment parsing https://review.opendev.org/717341 | 17:34 |
| openstackgerrit | Clark Boylan proposed zuul/nodepool master: Check dib image build timeout more often https://review.opendev.org/717357 | 17:39 |
| clarkb | fungi: ianw ^ I think that might be at least semi related to the dib process leaks | 17:39 |
| zbr | clarkb: corvus : does install-podman at https://review.opendev.org/#/c/716578/ look ready? | 17:41 |
| clarkb | zbr: I think it should be ensure-podman to go along with the refactor that avass is doing | 17:42 |
| clarkb | zbr: the idea being things might be preinstalled or not and ensure comunicates that a bit better | 17:42 |
| zbr | :( | 17:42 |
| zbr | clarkb: i was not aware of ensure-podman work, but i do support it. still why not using a symlink instead for the rename? | 17:48 |
| zbr | and more important: that refactoring does not bring any "meat" for me at the table as it does not enable me to install podman on redhat platforms. | 17:49 |
| zbr | if we merge avass work soon, like today, i am happy to rebase my work on top of it. | 17:50 |
| zbr | as long we endup with a zuul role that makes podman available on the node, i will be very happy. | 17:50 |
| tobiash | zbr: afair above avass work will be merged on monday | 17:51 |
| clarkb | I guess we can rebase the other direction | 17:52 |
| zbr | did I understand correctly that the direction is to replace install-* with ensure-* in order to make it more obvious for users that roles are about getting to a state and not about an "action" | 17:52 |
| clarkb | zbr: that was my understanding | 17:53 |
| zbr | for me ensure ~= install, at least from the configuration management point of view. | 17:53 |
| zbr | obiously that if something is already installed (and working), you should succeed :D | 17:54 |
| clarkb | zbr: part of it is we had both mixed in depending on the role. So the goal was to be consistent and ensure was chosen as it better reflects the inteint | 17:54 |
| clarkb | *intent | 17:54 |
| zbr | nothing against, it already faced at least one confused user. | 17:54 |
| *** hashar is now known as hasharDinner | 17:55 | |
| zbr | what is not fully clear to me is why having both, i would have used symlinks and only document that we will remove them after a while. | 17:57 |
| clarkb | zbr: as a transition. Symlinks may work though ime they should be avoided for artifacts like this as half they time they don't do what you expect | 17:58 |
| clarkb | they break python packaging for example | 17:58 |
| clarkb | and wouldn't work on windows machines | 17:59 |
| zbr | clarkb: as long you do not deploy on windows they work even with packages :D | 17:59 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-operator master: Add job_volumes CR spec attribute https://review.opendev.org/706642 | 18:00 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-operator master: Update attributes to camelCase https://review.opendev.org/707193 | 18:00 |
| zbr | does zuul-jobs work on windows? last time i check it was impossible to run tox -e linters even on BSD systems due to use of incompatible find args | 18:02 |
| clarkb | zbr: they are ansible roles and ansible runs on windows. I don't expect they would be used in a zuul context though | 18:03 |
| fungi | zbr: i would consider reasonable portability adaptations to roles/playbooks/jobs folks might want to run on those platforms as welcome | 18:04 |
| fungi | but only if folks are really trying to use them from zuul | 18:04 |
| zbr | fungi: zbr: the only reason why you didn't see a PR already is that I was waiting for the new ansible-lint release in order to make it. | 18:04 |
| fungi | and not just gratuitous portability work for the sake of it even if nobody might every use it | 18:04 |
| zbr | my desktop is a fruity BSD one so I am in the first line of finding such bugs, most of them are very easy to fix. | 18:05 |
| fungi | i do run a lot of openbsd, but i've not been bold enough to try using it for my workstation | 18:06 |
| fungi | i mostly rely on it for headless network devices | 18:06 |
| zbr | macos | 18:06 |
| fungi | oh, "fruit" | 18:07 |
| fungi | got it | 18:07 |
| zbr | the rotten one | 18:07 |
| fungi | heh | 18:07 |
| zbr | btw, if you have time to read (and vote) on https://github.com/ansible/ansible-lint/issues/716 it would be great | 18:08 |
| zbr | i started to work on adding a "molecule_whitelist" to it in order to avoid hacks like we needed for zuul_return, but progress is kinda slow, maybe I will have something ready by monday. | 18:10 |
| zbr | errata: "module_whitelist" | 18:11 |
| openstackgerrit | Merged zuul/zuul master: Protect getCachedChanges from concurrent modification https://review.opendev.org/715270 | 18:11 |
| openstackgerrit | Merged zuul/zuul master: Install unzip on all platforms https://review.opendev.org/714919 | 18:15 |
| corvus | zbr: there's a thread on the zuul-discuss list: http://lists.zuul-ci.org/pipermail/zuul-discuss/2020-April/001203.html | 18:16 |
| corvus | zbr: i'm looking for a review from mnaser on https://review.opendev.org/716578 before we proceed since he had concerns earlier | 18:18 |
| zbr | hehe, yesterday mail is still backlog. i will reply to show my support for the move, is a good one. | 18:18 |
| corvus | zbr: cool, thx :) | 18:19 |
| corvus | we're just waiting till monday since it's a big move in case anyone had any concerns, and they had a mail backlog :) | 18:20 |
| corvus | zbr: and probably a review from tristanC too | 18:21 |
| corvus | (re the podman change) | 18:21 |
| openstackgerrit | James E. Blair proposed zuul/nodepool master: Fix shutdown ordering https://review.opendev.org/717134 | 18:26 |
| openstackgerrit | James E. Blair proposed zuul/nodepool master: Add GCE driver tests https://review.opendev.org/717133 | 18:26 |
| *** rmart04 has quit IRC | 18:38 | |
| *** jamesmcarthur has quit IRC | 18:39 | |
| *** hasharDinner has quit IRC | 18:55 | |
| *** hashar has joined #zuul | 18:57 | |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-website master: WIP: switch website to Gatsby https://review.opendev.org/717371 | 19:05 |
| clarkb | Looking into that nodepool builder behavior with dib processes leaking I think my change is not necessary (its redundant for how control flow happens there and I missed it on my first pass) | 19:11 |
| clarkb | what I have noticed is that we do p.wait() and the subprocess docs warn this could deadlock with using PIPE | 19:11 |
| clarkb | I wonder if we are deadlocking there which would also explain the lack of logging about timeouts | 19:12 |
| clarkb | tobiash: ^ hvae you ever seen that with dib builds under nodepool-builder? basically over time disk-image-create processes accumulate | 19:12 |
| mnaser | zbr: i hate to be annoying but given we're moving towards ensure-* -- could you revise that so we don't have to rename it after? | 19:12 |
| mnaser | sorry, i know you've been waiting to land that for a while | 19:12 |
| mnaser | btw, for those interested, you can clone that gatsby change locally and run 'yarn install' and you should be good to go with it running locally | 19:12 |
| mnaser | i will be working on some zuul jobs/roles | 19:13 |
| mnaser | and take advantage of that to add some generic jobs in zuul-jobs | 19:13 |
| clarkb | mnaser: ignoring the netlify tooling of the potential future, what does gatsy workflow look like for us? its edit markdown, run build, copy/publish output basically the same workflow as sphinx ya? | 19:14 |
| fungi | mnaser: yeah, it sounds like zbr is already on board with the install->ensure rename and is rebasing onto that | 19:15 |
| tobiash | clarkb: I didn't notice leaked dib-image-create processes, but I can have a look on our builder tomorrow if I find any | 19:16 |
| clarkb | tobiash: thanks | 19:16 |
| openstackgerrit | Merged zuul/zuul master: Consolidate scheduler pause/exit as hibernation https://review.opendev.org/709723 | 19:23 |
| *** bhavikdbavishi has quit IRC | 19:23 | |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-website master: WIP: switch website to Gatsby https://review.opendev.org/717371 | 19:24 |
| *** bhavikdbavishi has joined #zuul | 19:25 | |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-website master: WIP: switch website to Gatsby https://review.opendev.org/717371 | 19:26 |
| *** gtema has joined #zuul | 19:26 | |
| openstackgerrit | Merged zuul/zuul master: Refactor `self.event_queues` in tests https://review.opendev.org/709990 | 19:27 |
| *** jamesmcarthur has joined #zuul | 19:28 | |
| *** jamesmcarthur has quit IRC | 19:28 | |
| *** jamesmcarthur has joined #zuul | 19:29 | |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-website master: WIP: switch website to Gatsby https://review.opendev.org/717371 | 19:29 |
| mnaser | clarkb: yep | 19:30 |
| *** gtema has quit IRC | 19:30 | |
| mnaser | clarkb: now just to clarify that we will not be committing built output though, so you dont really need to run build, you can use 'yarn develop' which will start a local server too | 19:31 |
| clarkb | right we don't commit sphinx output either | 19:31 |
| mnaser | yeah | 19:31 |
| mnaser | oops our yarn role is broken | 19:32 |
| mnaser | Warning: apt-key output should not be parsed (stdout is not a terminal)\ngpg: failed to start agent '/usr/bin/gpg-agent': No such file or directory\ngpg: can't connect to the agent: No such file or directory | 19:32 |
| AJaeger | anybody wants to review the tests for the parser jobs https://review.opendev.org/#/c/717341/? Then we can merge the complete stack. mordred , corvus, what do you think? | 19:36 |
| corvus | AJaeger: i'd love an ack from mnaser on https://review.opendev.org/717341 but then i agree, merge it all | 19:38 |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-jobs master: install-yarn: add coverage for all platforms https://review.opendev.org/717375 | 19:38 |
| clarkb | mnaser: gpg2 runs an agent daemon. I'm not sure if that is a hard dep or not, but if it is a hard dep you'll need to ensure that tool is installed I bet | 19:38 |
| clarkb | (its annoying bceause if you want ot load a key for a short period of time or off removable media you have to go find that process and kill it | 19:38 |
| mnaser | AJaeger, corvus: oh that's awesome. i think i would like to make a small revision on top of it where we add tests/__init__.py inside 'files' for the golangci-lint jobs. just to make sure that didn't break | 19:39 |
| mnaser | let me make a patch on top of it to check that its ok | 19:40 |
| AJaeger | tobiash: one question on 717341: I don't see where the tests run. The tests look fine but I must be blind or something is broken ;( | 19:41 |
| AJaeger | mnaser: sure, go for it. | 19:41 |
| mnaser | AJaeger: it should run inside the tox-pyXX envs | 19:42 |
| * AJaeger just removed his +2 after not finding the test results in the jobs that run. | 19:42 | |
| AJaeger | mnaser: those were not triggered | 19:42 |
| AJaeger | mnaser: AH! | 19:42 |
| * AJaeger is probably blind, will check | 19:42 | |
| tobiash | AJaeger: https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_b36/717341/3/check/tox-py37/b36c608/testr_results.html | 19:42 |
| mnaser | AJaeger: wait, nevermind, that also means the golang ones are running already too | 19:43 |
| mnaser | i potato'd there ;) | 19:43 |
| mnaser | +2 from me, thanks tobiash ! | 19:43 |
| tobiash | It's tested in the py* tests | 19:43 |
| AJaeger | tobiash: yeah, found them finally. Thanks | 19:43 |
| AJaeger | mnaser: so, I'll +A, ok? | 19:43 |
| corvus | AJaeger: beat you to it | 19:43 |
| AJaeger | corvus: no extra work for me -great ;) Thanks! | 19:44 |
| tobiash | \o/ | 19:45 |
| mnaser | hmm | 19:45 |
| openstackgerrit | Merged zuul/zuul-jobs master: Add phoronix-test-suite job https://review.opendev.org/679082 | 19:46 |
| mnaser | does anyone know why our install-yarn role backs out if there is no yarn.lock file? | 19:46 |
| corvus | mordred: ^ | 19:46 |
| mordred | uh | 19:46 |
| mnaser | i can understand bailing out in the actual 'yarn' aka run yarn role | 19:47 |
| mnaser | but in install-yarn that seems much -- seee https://opendev.org/zuul/zuul-jobs/src/branch/master/roles/install-yarn/tasks/main.yaml | 19:47 |
| mordred | mnaser: I think maybe because we also use it in some python jobs - and we're only wanting to opportunitistically install yarn if it's needed | 19:47 |
| mordred | but - that might not be the best design choice | 19:47 |
| *** dpawlik has joined #zuul | 19:48 | |
| mnaser | seems like 'javascript' is the only playbook which references it | 19:48 |
| mordred | I thnk what those python jobs that want it really want is "install-yarn-if-project-needs-it" | 19:48 |
| mnaser | https://opendev.org/zuul/zuul-jobs/src/branch/master/playbooks/javascript/pre.yaml is the only reference ive found for usage | 19:48 |
| mnaser | let me check hound | 19:48 |
| mordred | playbooks/release/pre.yaml in zuul | 19:48 |
| mnaser | ozj https://opendev.org/openstack/openstack-zuul-jobs/src/branch/master/playbooks/javascript/pre.yaml too | 19:48 |
| corvus | mnaser: (side note: be aware the error zuul is reporting on 717371 is due to the fact that the zuul-website-media repo uses the same jobs -- they work cooperatively to publish the site) | 19:48 |
| mordred | but yeah - I don't see any places that use it where it being conditional is useful | 19:48 |
| mnaser | corvus: oh thanks, i wasn't sure right off the bat so figured "leave them in there for now" was my workaround :p | 19:49 |
| corvus | mnaser: yeah, i think that's good, then we can update -media to work again | 19:50 |
| mnaser | ok id be in favour of dropping yarn.lock referencing in install-yarn, because it doesn't seem like i can find a case where it maeks sense | 19:50 |
| corvus | we're not going to upload any giant binary blobs in the next little bit | 19:50 |
| mnaser | thanks for the context mordred :> | 19:50 |
| corvus | mnaser: we run playbooks/javascript/pre.yaml in python jobs | 19:51 |
| corvus | mnaser: well, looks like only tox-nodejs-npm | 19:51 |
| corvus | that sounds like it's the sort of job we could assume yarn is wanted? | 19:51 |
| mnaser | corvus: i would agree with that | 19:52 |
| corvus | mnaser: yeah, i think i'm on board | 19:52 |
| mnaser | it also follows our consistency with other roles | 19:52 |
| mnaser | i.e. we do install tox regardless of you having a tox.ini file or not | 19:52 |
| clarkb | ++ the role including side can make it conditional if it needs that | 19:54 |
| clarkb | but the role itself should maybe be less conditional unless the necessary bits are already there (in this case yarn is installed already so noop) | 19:54 |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 19:56 |
| mnaser | cool, so that should also properly test it too | 19:57 |
| openstackgerrit | Merged zuul/zuul-jobs master: Support multiple matchers when parsing tox output https://review.opendev.org/716263 | 20:00 |
| openstackgerrit | Merged zuul/zuul-jobs master: Don't silently ignore exceptions when parsing tox output https://review.opendev.org/716766 | 20:00 |
| openstackgerrit | Merged zuul/zuul-jobs master: Strip source dir from file comments https://review.opendev.org/716264 | 20:00 |
| openstackgerrit | Merged zuul/zuul-jobs master: Ignore absolute paths after stripping work dir https://review.opendev.org/717042 | 20:00 |
| corvus | mnaser, clarkb, fungi: i think the gpg thing is more widespread | 20:01 |
| corvus | https://zuul.opendev.org/t/zuul/build/5d5fbbc8e3894e1e821bba396c991e22 | 20:01 |
| corvus | the last ps of that change did not fail that job | 20:01 |
| mnaser | hmm | 20:01 |
| mnaser | have we changed opendev's image recently? | 20:01 |
| clarkb | mnaser: should be every day :) | 20:01 |
| corvus | https://review.opendev.org/716785 was recent | 20:02 |
| mnaser | hmm | 20:02 |
| openstackgerrit | Merged zuul/zuul-jobs master: Add test cases for tox line comment parsing https://review.opendev.org/717341 | 20:02 |
| mnaser | we have a -plain image now? | 20:02 |
| clarkb | gnupg hard deps on gpg-agent on bionic | 20:02 |
| clarkb | (according to ubuntu package search) | 20:03 |
| clarkb | we shouldn't be able to install gnupg without gpg-agent | 20:03 |
| *** saneax has quit IRC | 20:03 | |
| corvus | didn't we used to output the image build info in the logs? | 20:03 |
| clarkb | mnaser: thats related to python things (basically no more pip and virtualenv in the root install) | 20:03 |
| corvus | here we go: https://zuul.opendev.org/t/zuul/build/5d5fbbc8e3894e1e821bba396c991e22/log/zuul-info/zuul-info.ubuntu-bionic.txt#6 | 20:04 |
| corvus | 2020-04-03 17:53 | 20:04 |
| corvus | so indeed, that is a newly built image, after https://review.opendev.org/716785 merged | 20:04 |
| mnaser | it seems that gnupg is a recommended package only? | 20:04 |
| mnaser | https://nb01.openstack.org/ubuntu-bionic-0000104264.log ctrl+f gnupg | 20:05 |
| corvus | (i'm not 100% sure 716785 was actually deployed, that's not clear since that's currently work in progress in #opendev) | 20:05 |
| mnaser | and gnupg2 is also just suggested | 20:05 |
| clarkb | mnaser: but `gpg` is tere according to the warning message | 20:06 |
| clarkb | er the line after the warning | 20:06 |
| clarkb | gpg is saying "I can't start my agent" | 20:06 |
| * clarkb checks gnupg2 too deps | 20:07 | |
| clarkb | gnupg2 deps gnupg which deps gpg-agent | 20:07 |
| clarkb | (those are all requires too not suggests or recommends) | 20:07 |
| mnaser | so | 20:07 |
| clarkb | fungi: ^ seems like you keep up with debian packaging, any idea how we can have a gpg intsalled without a gpg-agent binary? | 20:07 |
| mnaser | https://nb01.openstack.org/ubuntu-bionic-plain-0000000031.log installed gnupg | 20:08 |
| mnaser | https://nb01.openstack.org/ubuntu-bionic-plain-0000000043.log did not | 20:08 |
| corvus | should we switch to #opendev? | 20:08 |
| mnaser | good point | 20:08 |
| corvus | (this is starting to look like an opendev problem) | 20:08 |
| openstackgerrit | Merged zuul/zuul-jobs master: helm: collect kubernetes logs in post https://review.opendev.org/715709 | 20:14 |
| *** y2kenny has joined #zuul | 20:26 | |
| mnaser | hmm | 20:26 |
| mnaser | i guess we'll have to add install-nodejs for install-yarn | 20:27 |
| mnaser | almost every single platform is broken because it either doesnt ship node or ships one thats too old | 20:27 |
| y2kenny | tristanC: hey sorry, I had a 3 hrs meeting. How do I check the Content-Security-Header for the zuul webui? | 20:27 |
| fungi | y2kenny: you could use a tool like wget or curl to inspect the http headers, or your browser may have a debugging console (either built in or as a popula "web developer" extension) | 20:31 |
| y2kenny | ok... so the content security policy is on the zuul web UI side and not the log serving side? | 20:31 |
| tristanC | y2kenny: yes, that's something that can be set by a static file provider to prevent further html/javascript action from happening | 20:36 |
| fungi | right, it may or may not be present, if it is present you may have to adjust it | 20:36 |
| fungi | if it's not present then it's not the cause of the problem | 20:36 |
| y2kenny | tristanC: I don't see that header from either side. But I do have a 404 on job-output.json.gz | 20:37 |
| y2kenny | even though I have job-output.json in the directory | 20:37 |
| y2kenny | and the network tab shows the get for zuul-manifest.json is also successful | 20:37 |
| tristanC | y2kenny: that's normal, iirc the webui try the .gz version first, then foldback to the uncompressed version | 20:37 |
| tristanC | y2kenny: what's the issue again? | 20:37 |
| y2kenny | when I load the build summary, I get an error stating Network error | 20:38 |
| y2kenny | (hold on... the pop up disappeared) | 20:38 |
| y2kenny | Network Error (Unable to fetch URL, check your network connectivity, browser plugins, ad-blockers, or try to refresh this page) ...1d5ecd8/zuul-manifest.json | 20:39 |
| y2kenny | I think adding the CORS header on the log serving side gave me more stuff now | 20:39 |
| y2kenny | but the error still pops up | 20:39 |
| y2kenny | (I get the pretty UI :D) | 20:39 |
| fungi | for opendev's deployment, our logs are being served from a different site (actually a bunch of different sites in different domains) so we configure those sites to include "Access-Control-Allow-Origin: *" in the headers | 20:40 |
| y2kenny | (the console tab) | 20:40 |
| fungi | which is the most common cause of the problem | 20:40 |
| y2kenny | I added the header | 20:40 |
| fungi | okay, so if you fetch a zuul-manifest.json file you see that included in the headers of the response | 20:41 |
| y2kenny | and now I see the console tab but for some reason I still get pop up error... even though I think the webui actually picked up all the necessary info | 20:41 |
| fungi | maybe a cached response? | 20:42 |
| y2kenny | I thought something may be cached but shift-reload didn't remove it.... trying on Chrome (which never loaded the page before) it's fine... yea... it's probably some cache thing | 20:44 |
| tristanC | corvus: btw, i added a job-volume test in https://review.opendev.org/706642 . that's something we'd like to start using and it would be convenient if the feature get merged soon. | 20:44 |
| y2kenny | um.... let me see what happens if I try to load the page in firefox private mode | 20:44 |
| y2kenny | private mode also no error... so it's probably some lingering thing. Anyway, thanks guys. | 20:45 |
| fungi | these days the only way to be sure you've cleared a browser cache is to nuke it from orbit | 20:50 |
| fungi | take an angle grinder to your ssd and polish off the bits | 20:50 |
| y2kenny | I remember the days when a super strong magnet would've done it... ;) | 20:51 |
| fungi | yup, i used to have to use a degaussing table to clear our sensitive media at a job many beers ago | 20:53 |
| *** EvilienM is now known as EmilienM | 21:07 | |
| *** jamesmcarthur has quit IRC | 21:23 | |
| *** jamesmcarthur has joined #zuul | 21:24 | |
| *** jamesmcarthur has quit IRC | 21:29 | |
| *** bhavikdbavishi has quit IRC | 21:30 | |
| *** jamesmcarthur has joined #zuul | 21:36 | |
| *** hashar has quit IRC | 22:05 | |
| *** armstrongs has joined #zuul | 22:08 | |
| *** rlandy has quit IRC | 22:12 | |
| *** armstrongs has quit IRC | 22:18 | |
| *** dpawlik has quit IRC | 22:32 | |
| *** igordc has quit IRC | 22:36 | |
| *** igordc has joined #zuul | 22:36 | |
| *** tosky has quit IRC | 23:16 | |
| *** jamesmcarthur has quit IRC | 23:19 | |
| *** jamesmcarthur has joined #zuul | 23:20 | |
| *** jamesmcarthur has quit IRC | 23:25 | |
| *** jamesmcarthur has joined #zuul | 23:29 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
