| *** mattw4 has quit IRC | 00:16 | |
| mnaser | ok i pretty much have these helm charts done (mostly) and currently being cleaned up | 00:47 |
|---|---|---|
| mnaser | does anyone know of a nice small tool that can do things like "watch for file and send signal" | 00:47 |
| mnaser | i have a way of restarting all containers if the config files change, but that is obviously terrible because a tenant config change isnt worth restarting all containers, i can write something that wathces the file and sends a signal but i feel like a small tool like that .. must exist somewhere | 00:48 |
| clarkb | mnaser inotify is the syscall | 00:51 |
| fungi | inotify? | 00:51 |
| fungi | d'oh, jinx! | 00:51 |
| clarkb | inotify-wait is a tool that uses it | 00:51 |
| fungi | yep | 00:52 |
| mnaser | right, but i was hoping to avoid writing something that uses inotify but if inotify-wait is a thing ill look into that | 00:52 |
| mnaser | https://linux.die.net/man/1/inotifywatch | 00:52 |
| mnaser | oh cool | 00:52 |
| mnaser | ok so i can watch for that and send signals to zuul | 00:52 |
| clarkb | er watch maybe thats it | 00:52 |
| fungi | though in theory zuul might could be told how to watch files for changes in a similar manner | 00:53 |
| fungi | rather than relying on an external process to watch them and signal its process | 00:53 |
| mnaser | fungi: i thought about that but i figured folks might want to do a controlled reload generally | 00:53 |
| fungi | sure | 00:53 |
| fungi | seems like if it were a feature it would need to be able to be disabled | 00:54 |
| mnaser | ah yep | 00:54 |
| *** jamesmcarthur has joined #zuul | 01:06 | |
| *** jamesmcarthur has quit IRC | 01:10 | |
| *** bhavikdbavishi has joined #zuul | 02:42 | |
| *** irclogbot_0 has quit IRC | 02:56 | |
| *** irclogbot_2 has joined #zuul | 03:16 | |
| *** rlandy|bbl is now known as rlandy | 03:20 | |
| *** rlandy has quit IRC | 03:25 | |
| *** reiterative has quit IRC | 05:34 | |
| *** pcaruana has joined #zuul | 05:43 | |
| *** bhavikdbavishi1 has joined #zuul | 05:56 | |
| *** bhavikdbavishi has quit IRC | 05:57 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 05:57 | |
| *** raukadah is now known as chandankumar | 06:06 | |
| *** saneax has joined #zuul | 06:31 | |
| *** aluria has joined #zuul | 07:21 | |
| *** avass has joined #zuul | 07:36 | |
| *** saneax has quit IRC | 07:43 | |
| *** irclogbot_2 has quit IRC | 07:44 | |
| *** irclogbot_2 has joined #zuul | 07:46 | |
| *** jcapitao|afk has joined #zuul | 07:57 | |
| *** jcapitao|afk is now known as jcapitao | 08:11 | |
| *** tosky has joined #zuul | 08:14 | |
| *** saneax has joined #zuul | 08:20 | |
| *** themroc has joined #zuul | 08:27 | |
| *** jpena|off is now known as jpena | 08:38 | |
| *** hashar has joined #zuul | 08:59 | |
| *** reiterative has joined #zuul | 09:11 | |
| *** bhavikdbavishi has quit IRC | 09:23 | |
| *** tosky_ has joined #zuul | 10:07 | |
| *** tosky is now known as Guest97727 | 10:07 | |
| *** tosky_ is now known as tosky | 10:07 | |
| *** Guest97727 has quit IRC | 10:08 | |
| *** bhavikdbavishi has joined #zuul | 10:47 | |
| *** bhavikdbavishi1 has joined #zuul | 10:49 | |
| *** bhavikdbavishi has quit IRC | 10:51 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 10:51 | |
| *** armstrongs has joined #zuul | 11:05 | |
| *** armstrongs has quit IRC | 11:14 | |
| *** jcapitao is now known as jcapitao|afk | 11:16 | |
| *** sshnaidm|afk is now known as sshnaidm | 11:21 | |
| *** jpena is now known as jpena|lunch | 11:58 | |
| *** avass has quit IRC | 12:09 | |
| *** jcapitao|afk is now known as jcapitao | 12:54 | |
| *** rlandy has joined #zuul | 12:57 | |
| *** Goneri has quit IRC | 13:00 | |
| *** BanHammor has joined #zuul | 13:07 | |
| *** jpena|lunch is now known as jpena | 13:11 | |
| BanHammor | Hi folks! My zuul job starts a Docker container, does some work with the container, and then its job.cleanup-run shuts down the container. | 13:11 |
| BanHammor | I want to move away from a fixed container name to a randomly-generated one, and i'm doing that by writing down the container name in a set_fact | 13:12 |
| BanHammor | The question is - can I pass the fact from the run playbook to the cleanup-run playbook? | 13:12 |
| *** jamesmcarthur has joined #zuul | 13:29 | |
| *** avass has joined #zuul | 13:34 | |
| *** jamesmcarthur has quit IRC | 13:34 | |
| *** Goneri has joined #zuul | 13:46 | |
| *** jamesmcarthur has joined #zuul | 13:53 | |
| *** tosky_ has joined #zuul | 14:23 | |
| *** tosky is now known as Guest27890 | 14:23 | |
| *** tosky_ is now known as tosky | 14:23 | |
| *** Guest27890 has quit IRC | 14:26 | |
| *** sugaar has quit IRC | 14:30 | |
| *** sugaar has joined #zuul | 14:39 | |
| *** rfolco has joined #zuul | 14:43 | |
| *** jamesmcarthur has quit IRC | 14:51 | |
| *** jamesmcarthur has joined #zuul | 14:56 | |
| *** jpena is now known as jpena|off | 15:04 | |
| clarkb | BanHammor: I think you have to write that data to disk on the executor then load it again in cleanup? a better option may be to use the build uuid valur to name your container | 15:10 |
| clarkb | that should be random but fixed to the job currently running | 15:10 |
| BanHammor | ooh that's a nice idea! | 15:11 |
| BanHammor | (unfortunately, because of constraints i'm working with, kind of impossible, but really neat) | 15:11 |
| *** jpena|off is now known as jpena | 15:13 | |
| tobiash | BanHammor: you could try zuul_return, I think zuul should forward that variable to the next playbook | 15:21 |
| Shrews | i think we really need a set/get example of using zuul_return. I was looking for one and could not find it | 15:23 |
| *** electrofelix has joined #zuul | 15:24 | |
| Shrews | (in our docs, that is) | 15:27 |
| BanHammor | tobiash, it seems like more of a job-to-zuul thing than a playbook-to-playbook thing? | 15:31 |
| fungi | yeah, i was looking for more detail on zuul_return as well, to see if it said that would allow you to pass variables to subsequent playbooks but it only talks about passing them to child jobs | 15:36 |
| fungi | (also, yes, an example would be an awesome addition to the docs) | 15:37 |
| *** bhavikdbavishi has quit IRC | 15:54 | |
| *** jamesmcarthur has quit IRC | 15:56 | |
| *** jamesmcarthur has joined #zuul | 15:57 | |
| *** jamesmcarthur has quit IRC | 16:02 | |
| openstackgerrit | David Shrewsbury proposed zuul/zuul master: Docs: Fix zuul variable references https://review.opendev.org/700014 | 16:09 |
| tobiash | BanHammor, fungi: I thought there was a change that forwards this to playbools as well, but I might be wrong | 16:11 |
| tobiash | but maybe my memory mixed it up with https://review.opendev.org/642857 | 16:12 |
| *** themroc has quit IRC | 16:16 | |
| openstackgerrit | David Shrewsbury proposed zuul/zuul master: Docs: Cleanup jobs doc https://review.opendev.org/700014 | 16:26 |
| *** jamesmcarthur has joined #zuul | 16:27 | |
| *** jamesmcarthur has quit IRC | 16:27 | |
| *** jamesmcarthur has joined #zuul | 16:28 | |
| *** armstrongs has joined #zuul | 16:34 | |
| *** armstrongs has quit IRC | 16:43 | |
| *** mattw4 has joined #zuul | 16:43 | |
| *** jamesmcarthur has quit IRC | 16:51 | |
| *** hashar has quit IRC | 17:16 | |
| Shrews | tfw your paste of a filename ends up being an accidental paste of a URL and 'vi' starts downloading it | 17:20 |
| clarkb | Shrews: new xterm/xfce-terminal pops up a warning when I try pasting anything remotely dangerous and makes me review it before adding it to my shell | 17:21 |
| Shrews | i mean, i know vi can do that with urls, but i seem to always forget and get very surprised for a moment when i accidentally do that | 17:23 |
| clarkb | Shrews: it is really useful for reading large zuul logs that cause your browser to bog down and be sad | 17:23 |
| openstackgerrit | Merged zuul/zuul master: Add note to supercedent with regards to file filters https://review.opendev.org/698733 | 17:25 |
| clarkb | corvus: I haven't seen any complaining post opendev zuul restart | 17:26 |
| clarkb | I guess today is the pi release day? | 17:26 |
| corvus | cool, i'll get started on that soon | 17:26 |
| clarkb | then I'll go update the annual report draft to get the number of releases correct :) | 17:27 |
| *** chandankumar has quit IRC | 17:31 | |
| corvus | commit 84f6ea667c3453a75a7a0210ee08228c9eec167a (HEAD -> master, tag: 3.14.0, origin/master, origin/HEAD) | 17:33 |
| corvus | that look right? | 17:33 |
| clarkb | commit and tag number both look correct to me | 17:33 |
| corvus | pushed | 17:34 |
| clarkb | corvus: https://review.opendev.org/#/c/699857/ is semi related (pushes ansible things forward a bit more) | 17:36 |
| corvus | done | 17:38 |
| clarkb | using codesearch to grep around for more instances of the issues I had to fix in zuul-jobs I seem to only find them in playbooks/roles that are in ansible things not run by zuul | 17:40 |
| clarkb | ah here is at least one that might trip on the zuul side | 17:40 |
| clarkb | overall though not as bad as I feared | 17:40 |
| *** jpena is now known as jpena|off | 17:44 | |
| clarkb | zuulians: I've added some edits around user stats to the annual report draft. If you have reviewd it already another quick look would be great. If you haven't reviewed it yet any help is much appreciated | 18:05 |
| clarkb | https://etherpad.openstack.org/p/zuul-2019-annual-report-data you can find it there as well as fungi's stats collection | 18:05 |
| clarkb | one neat thing in ^ is Zuul has contributions from all continents but antarctica over 2019 | 18:05 |
| clarkb | mordred: ^ want to become a south pole resident to fix that for us? | 18:06 |
| Shrews | corvus: it seems our zuul_return plugin supports a 'path' and 'file' argument, though I don't see either one being used anywhere in our system. path will always be to the 'results.json' file, and i have NO idea what 'file' is supposed to have been used for. Is there a backstory on the usage of those? | 18:06 |
| *** panda has quit IRC | 18:08 | |
| corvus | Shrews: path is there in case someone wants to abuse zuul_return to operate on other files (maybe some weird nested thing?). file is so you can return a bunch of data without loading it into ansible | 18:11 |
| corvus | (ie, have a script that outputs the return data, then just "zuul_return: file: /path/to/output" or something | 18:11 |
| Shrews | ohhhhh... enlightenment | 18:12 |
| *** panda has joined #zuul | 18:13 | |
| corvus | iow, path has no intended use, it's just there "in case" and to avoid hardcoding something. file does have a use, but afaik, it's never been used in anger. possibly because it's not documented? :) | 18:13 |
| corvus | and probably because we've never gotten around to finishing up the inline pep8 job | 18:14 |
| Shrews | definitely not documented :) I was working on trying to understand it better to hopefully add some of that | 18:14 |
| corvus | ++ | 18:14 |
| *** rfolco is now known as rfolco|bbl | 18:18 | |
| *** jcapitao has quit IRC | 18:31 | |
| openstackgerrit | David Shrewsbury proposed zuul/zuul master: Add some docs to the zuul_return action plugin https://review.opendev.org/700049 | 18:45 |
| Shrews | Just a few things that would have helped me out ^^^ | 18:47 |
| clarkb | Shrews: that has pep8 failures and they don't appear related to the ansible version update (job still ran under 2.8) | 18:57 |
| Shrews | i didn't think we ran pep8 on the ansible stuff. meh | 19:18 |
| openstackgerrit | David Shrewsbury proposed zuul/zuul master: Add some docs to the zuul_return action plugin https://review.opendev.org/700049 | 19:20 |
| *** electrofelix has quit IRC | 19:28 | |
| *** dmsimard has quit IRC | 19:41 | |
| *** dmsimard has joined #zuul | 19:41 | |
| tobiash | mordred: I've responded on https://review.opendev.org/695001 | 19:58 |
| *** sshnaidm is now known as sshnaidm|afk | 20:00 | |
| *** saneax has quit IRC | 20:00 | |
| *** pabelanger has quit IRC | 20:05 | |
| *** maxamillion has quit IRC | 20:05 | |
| *** maxamillion has joined #zuul | 20:05 | |
| *** irclogbot_2 has quit IRC | 20:06 | |
| *** irclogbot_0 has joined #zuul | 20:11 | |
| *** pabelanger has joined #zuul | 20:29 | |
| openstackgerrit | Merged zuul/nodepool master: 'keys' must be defined for host-key-checking: false https://review.opendev.org/698029 | 20:36 |
| openstackgerrit | Merged zuul/zuul master: Docs: Cleanup jobs doc https://review.opendev.org/700014 | 20:49 |
| *** jamesmcarthur has joined #zuul | 20:59 | |
| *** jamesmcarthur has quit IRC | 21:16 | |
| *** pcaruana has quit IRC | 21:23 | |
| *** Goneri has quit IRC | 21:35 | |
| *** jamesmcarthur has joined #zuul | 21:36 | |
| *** shanemcd- has quit IRC | 21:36 | |
| *** shanemcd has joined #zuul | 21:38 | |
| *** rfolco|bbl has quit IRC | 22:06 | |
| mnaser | Project <..> not in pipeline <Pipeline check> for change <Change 0x7fbf91b85d50 ... 1,726373902bb8218dd0bf0bb99e2e91d72c0505e1> | 22:09 |
| mnaser | is this possibly because there is no 'noop' jobs defined so it doesnt exist at all? | 22:09 |
| mnaser | this is a PR from an unprotected branch into a protected branch on github. | 22:09 |
| mnaser | (where this specific PR adds zuul.d/project.yaml) | 22:10 |
| corvus | mnaser: you don't actually need a noop job, you can have an empty job list, but unless/until there's a project-pipeline stanza for a project, zuul will ignore changes | 22:10 |
| corvus | (otherwise, it might be acting on changes to a project that has no intention of actually using zuul) | 22:11 |
| corvus | mnaser: one way to avoid this is to put into a config-project a project stanza with an empty job list that matches all projects | 22:11 |
| clarkb | zuul's jobs are running under ansible 2.9 and seem happy from what I've seen so far | 22:11 |
| mnaser | corvus: yeah i was thinking of just doing that but i guess i'll avoid it till i figure out the possible implications | 22:11 |
| corvus | (but that would explicitly say "every project uses the check pipeline") | 22:12 |
| corvus | mnaser: the main downside is that if zuul is configured to report merge conflicts (which it usually is for check), then it would do so if someone submits a pr with a conflict | 22:13 |
| corvus | mnaser: otherwise, even if it's in the pipeline with no jobs, it'll still keep quiet. | 22:13 |
| mnaser | corvus: ok gotcha | 22:14 |
| corvus | mnaser: if you don't want to do that, the other general approach would be to merge a change which adds a project stanza with no jobs or the noop job; then subsequent changes will be self-testing | 22:15 |
| jlk | quick question from a group that might likely know. I'm trying to update a tox file so that on win32, for the py34 environment, one of the deps versions gets overridden. What I'm seeing is that it does an _additive_ dependency rather than a replacement. https://github.com/sigmavirus24/github3.py/pull/978 is my attempt thus far. | 22:15 |
| jlk | (yes I know this isn't zuul, but y'all are pretty intimately familiar w/ Tox) | 22:15 |
| mnaser | corvus: yeah i think that's the best "opt-in" approach | 22:15 |
| clarkb | jlk: I think if you change line 11 to add ; sys_platform != 'win32' then you'll get what you want | 22:16 |
| jlk | oooh | 22:16 |
| clarkb | or I guess you have to accomodate the py34 as well | 22:16 |
| clarkb | but thats the general idea make a line that matches when the other doesnt | 22:17 |
| clarkb | and vice versa | 22:17 |
| corvus | is that the same as requirements files, or is the tox "deps" thing different? | 22:17 |
| clarkb | corvus: at least on older tox that list was given verbatim to pip | 22:17 |
| clarkb | corvus: which is why our lines will look like -r requirements.txt | 22:18 |
| corvus | gotcha | 22:18 |
| jlk | wait, hrm. | 22:18 |
| jlk | the problem is the base 'pytest' line and the new line that starts with py34 | 22:19 |
| jlk | I don't know how to exclude the base pytest line only on py34 | 22:19 |
| clarkb | jlk: https://www.python.org/dev/peps/pep-0496/#id7 that may help | 22:19 |
| clarkb | I would do pytest>=2.3.5, <5.0.0 ; sys_platform != 'win32' or ( sys_platform == 'win32' and python_version != '3.4') | 22:21 |
| jlk | ahhh | 22:21 |
| clarkb | and pytest>=2.3.5, <4.6.7 ; sys_platform == 'win32' and python_version == '3.4' | 22:22 |
| jlk | the python_version one was what I wanted | 22:22 |
| *** coldtom has quit IRC | 22:32 | |
| *** jamesmcarthur has quit IRC | 22:34 | |
| *** aluria has quit IRC | 22:49 | |
| *** tosky has quit IRC | 23:22 | |
| jlk | clarkb: that definitely worked; the version I wanted got installed. Sadly that wasn't the package that seems to be breaking my tests :( | 23:27 |
| *** Goneri has joined #zuul | 23:52 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!