| *** jamesmcarthur has joined #zuul | 00:02 | |
| *** jamesmcarthur has quit IRC | 00:21 | |
| *** Defolos has quit IRC | 00:23 | |
| *** Defolos has joined #zuul | 00:25 | |
| *** jamesmcarthur has joined #zuul | 00:26 | |
| *** Defolos has quit IRC | 00:34 | |
| *** jamesmcarthur has quit IRC | 01:05 | |
| *** tosky has quit IRC | 01:06 | |
| *** jamesmcarthur has joined #zuul | 01:06 | |
| *** jamesmcarthur has quit IRC | 01:09 | |
| *** jamesmcarthur_ has joined #zuul | 01:09 | |
| *** jamesmcarthur_ has quit IRC | 01:26 | |
| *** jamesmcarthur has joined #zuul | 01:27 | |
| *** jamesmcarthur has quit IRC | 01:29 | |
| *** jamesmcarthur_ has joined #zuul | 01:29 | |
| *** jamesmcarthur_ has quit IRC | 01:39 | |
| *** jamesmcarthur has joined #zuul | 01:40 | |
| *** jamesmcarthur has quit IRC | 01:54 | |
| *** jamesmcarthur has joined #zuul | 01:54 | |
| *** jamesmcarthur has quit IRC | 02:09 | |
| *** jamesmcarthur has joined #zuul | 02:11 | |
| *** jamesmcarthur has quit IRC | 02:29 | |
| *** jamesmcarthur has joined #zuul | 02:29 | |
| *** jamesmcarthur has quit IRC | 03:21 | |
| *** jamesmcarthur has joined #zuul | 04:30 | |
| *** sshnaidm|afk has quit IRC | 04:38 | |
| *** sshnaidm|afk has joined #zuul | 04:51 | |
| *** jamesmcarthur has quit IRC | 05:29 | |
| *** evrardjp has quit IRC | 05:34 | |
| *** evrardjp has joined #zuul | 05:35 | |
| *** raukadah is now known as chandankumar | 05:43 | |
| *** sgw has joined #zuul | 06:39 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: install-javascript-packages: add tox_constraints_file https://review.opendev.org/709414 | 06:48 |
|---|---|---|
| *** AJaeger has quit IRC | 06:59 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: install-javascript-packages: add tox_constraints_file https://review.opendev.org/709414 | 07:00 |
| *** AJaeger has joined #zuul | 07:03 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: install-javascript-packages: add tox_constraints_file https://review.opendev.org/709414 | 07:25 |
| *** dpawlik has joined #zuul | 07:29 | |
| *** sshnaidm|afk has quit IRC | 07:41 | |
| *** Defolos has joined #zuul | 07:47 | |
| *** saneax has joined #zuul | 07:47 | |
| *** sshnaidm|afk has joined #zuul | 07:55 | |
| mnaser | yay, foreground option landed! | 08:03 |
| mnaser | testing it right now | 08:04 |
| *** sshnaidm|afk is now known as sshnaidm | 08:04 | |
| *** paulalbertella is now known as reiterative | 08:07 | |
| mnaser | yay, working nicely locally | 08:14 |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-helm master: Drop args from container spec https://review.opendev.org/709424 | 08:17 |
| *** tosky has joined #zuul | 08:18 | |
| *** jcapitao has joined #zuul | 08:23 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: install-javascript-packages: add tox_constraints_file https://review.opendev.org/709414 | 08:24 |
| *** jcapitao has quit IRC | 08:29 | |
| *** jcapitao has joined #zuul | 08:29 | |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Enforce sql connections for scheduler and web https://review.opendev.org/630472 | 09:20 |
| *** bolg has joined #zuul | 09:21 | |
| *** yolanda__ has joined #zuul | 09:37 | |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Scheduler test app factory https://review.opendev.org/708812 | 09:39 |
| *** yolanda has quit IRC | 09:40 | |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Make test setup_config more pure https://review.opendev.org/709084 | 09:42 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Scheduler test app factory https://review.opendev.org/708812 | 09:42 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Add spec for scale out scheduler https://review.opendev.org/621479 | 09:50 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Enforce sql connections for scheduler and web https://review.opendev.org/630472 | 09:51 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: WIP: Store unparsed branch config in Zookeeper https://review.opendev.org/705716 | 10:52 |
| *** avass has joined #zuul | 11:09 | |
| tobiash | zuul-maint: did anyone also encounter occasional ssh remote host identification changed errors in builds? | 11:23 |
| tobiash | I did some digging in the logs and it seems that nodepool stole the fip from a build node with a running job | 11:23 |
| tobiash | I wonder if wait_for_server in openstacksdk (this does fip attachment) has some race | 11:24 |
| tobiash | mordred: maybe you have a hunch | 11:25 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: WIP: Store unparsed branch config in Zookeeper https://review.opendev.org/705716 | 11:37 |
| tobiash | I wonder if it has something to do with https://opendev.org/opendev/system-config/src/branch/master/playbooks/templates/clouds/nodepool_clouds.yaml.j2#L18 | 11:39 |
| tobiash | we have the same settings | 11:39 |
| *** jcapitao is now known as jcapitao_lunch | 12:01 | |
| tobiash | mordred: reading the neutron api docs, am I right that it's basically impossible to attach an unused fip race free to an instance? | 12:19 |
| tobiash | https://docs.openstack.org/api-ref/network/v2/?expanded=update-floating-ip-detail | 12:19 |
| *** nborg1 has joined #zuul | 12:19 | |
| nborg1 | mnaser: Sorry for the late response, but no, I did not see any warning bell for the tenant when the new project did not appear in the project list. A reboot of the scheduler and web "solved" it. | 12:21 |
| nborg1 | Today we have another problem. gerrit tags do not seem to match any jobs. Should they not automatically match jobs on the branch where the tag is pushed? | 12:22 |
| nborg1 | We cannot release :o | 12:22 |
| avass | Our release jobs seems to not run. It looks like no job variants matches the change. debug log looks like this for every job in the projectconfig: http://paste.openstack.org/show/789927/ | 12:24 |
| avass | Have I missed something? | 12:24 |
| AJaeger | avass: where is the release job configuration? | 12:29 |
| AJaeger | avass: where is the release job configuration (project stanza) ? is it in the projectconfig? | 12:29 |
| avass | yeah, in a different file under zuul.d/ in the same project | 12:30 |
| avass | uuh what I mean is job configuration and project configuration is in the same project in different files under zuul.d/ | 12:31 |
| AJaeger | You have a Jobt that is not “branch aware”. Typically these are jobs that are triggered by tag based events. | 12:31 |
| AJaeger | So, release pipeline has not branch assocatiated. | 12:31 |
| AJaeger | That's why we have all the publish-to-pypi project stanza in project-config and not in repo. | 12:31 |
| AJaeger | avass: in-repo works only if you have only a master branch | 12:32 |
| avass | ah | 12:32 |
| AJaeger | avass: release pipeline is special as it's not branch aware... | 12:32 |
| AJaeger | https://docs.openstack.org/infra/manual/creators.html#central-config-exceptions | 12:33 |
| AJaeger | Not sure what the Zuul docs say to it... | 12:33 |
| avass | I'm not sure it really mentions it, but I guess it sort of makes sense. | 12:33 |
| avass | What we want to do is do a release on the branch where a commit is tagged. If that makes sense | 12:35 |
| nborg1 | Does that mean that we cannot have a parent in the project either? For instance how we should compile? | 12:37 |
| AJaeger | avass: sure, that is what we do in OpenDev as well - and put the template in your project-config | 12:37 |
| AJaeger | nborg1: AFAIK you need to put it elsewhere - in a repo without branches. | 12:39 |
| AJaeger | avass, nborg1, best discuss with rest of team later - and would be good if you put up a brief example so that we all know what we talk about. | 12:39 |
| nborg1 | AJaeger: - project: { name: <project_name> release: { jobs: [ compile ] }}, where we want to define the compile job in the untrusted-project, since it changes over time. | 12:43 |
| AJaeger | nborg1: please discuss with rest of team later. My first reaction: Won't work, but I'm not the ultimate expert here | 12:44 |
| AJaeger | bbl | 12:44 |
| tobiash | zuul-maint: re fip-reuse: removing floating-ip from clouds.yaml cache didn't help | 12:51 |
| *** rlandy has joined #zuul | 12:56 | |
| *** rfolco has joined #zuul | 13:04 | |
| *** jamesmcarthur has joined #zuul | 13:13 | |
| *** jcapitao_lunch is now known as jcapitao | 13:19 | |
| *** jamesmcarthur has quit IRC | 13:32 | |
| *** jamesmcarthur has joined #zuul | 13:45 | |
| Shrews | tobiash: "stole" an in-use fip? i've never heard of that. not even sure how that's possible | 13:54 |
| Shrews | but, everything around fip is so weird anyway | 13:55 |
| *** saneax has quit IRC | 14:07 | |
| *** Goneri has joined #zuul | 14:19 | |
| *** jamesmcarthur has quit IRC | 14:22 | |
| mnaser | what's the difference between zuul's waiting and queued status | 14:35 |
| mnaser | seems like things in retry are in waiting status | 14:36 |
| pabelanger | queued is when the change has been added to the queue, waiting is for nodepool to fully the node request to start running the job | 14:37 |
| fungi | avass: sounds like you might want something like https://review.opendev.org/578557 | 14:37 |
| pabelanger | I usually see more 'waiting' when at capacity in nodepool | 14:37 |
| mnaser | can a change go from queued => waiting at capacity? | 14:38 |
| pabelanger | or multinode jobs, cannot fulfill the noderequest | 14:38 |
| fungi | avass: the commit message explains why "release on the branch where a commit is tagged. If that makes sense | 14:38 |
| tobiash | mnaser: waiting is waiting for a parent job | 14:38 |
| mnaser | tobiash: in this case there is no parent job though | 14:38 |
| fungi | avass: er, the commit message explains why "release on the branch where a commit is tagged" is a vague assertion | 14:38 |
| tobiash | that sounds like a bug | 14:38 |
| fungi | s/assertion/condition/ | 14:39 |
| mnaser | ex: opendev change 709461,1 | 14:39 |
| mnaser | openstack-ansible-deploy-aio_lxc-debian-buster (2. attempt) | 14:39 |
| mnaser | is current 'waiting' | 14:39 |
| pabelanger | it is waiting for nodepool | 14:39 |
| mnaser | also the fact that pabelanger and tobiash seem to have different thoughts on the value maybe means we should document it :p | 14:40 |
| pabelanger | likely something going on with a nodepool provider | 14:40 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Use scheduler factory consistently in tests https://review.opendev.org/709542 | 14:40 |
| fungi | i thought "waiting" state was when a build is waiting on a build of another job specified as a dependency (either directly or via provides/requires) | 14:41 |
| fungi | possible we reuse it for multiple conditions? | 14:41 |
| pabelanger | maybe | 14:42 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: WIP: Store unparsed branch config in Zookeeper https://review.opendev.org/705716 | 14:42 |
| mnaser | ok so the UI | 14:43 |
| mnaser | if job.queued = false { 'waiting' } else { 'queued' } | 14:43 |
| mnaser | so if its not queued, it shows queued .. otherwise its waiting | 14:45 |
| mnaser | wat | 14:45 |
| *** avass has quit IRC | 14:49 | |
| corvus | mnaser: i just woke up, but i'm pretty sure that says if it's not queued it shows waiting otherwise it's queued | 14:54 |
| corvus | 'queued' is supposed to mean that the request for notes is in nodepool's queue. 'waiting' is supposed to mean that it is waiting to make that request (usually because a parent job has not completed, a requirement-providing job has not completed, or it needs a semaphore) | 14:55 |
| corvus | if none of the 3 seem relevant here, then i agree with tobiash, that's sounding like a bug | 14:56 |
| tobiash | mnaser, pabelanger: this is the change that introduced waiting in the ui: https://review.opendev.org/660878 | 14:58 |
| tobiash | could be that that state is not correctly set in the retry case (however this should only be a ui issue) | 14:59 |
| tobiash | except if it's not started at all, then you might hit an edge case somewhere regarding retries | 15:00 |
| corvus | tobiash: yeah, but why would it get reset back to false? | 15:00 |
| corvus | i think there is a nodepool request for this job | 15:00 |
| tobiash | corvus: I don't know, maybe that hits a different code path? | 15:00 |
| tobiash | corvus: but retried jobs get their node requests via findJobsToRequest as well right? | 15:01 |
| corvus | tobiash: i believe so | 15:02 |
| tobiash | then I don't understand that yet | 15:02 |
| corvus | another job on that change just hit a retry and is now 'queued' | 15:02 |
| tobiash | I also observed such oddities in our system, but hadn't yet time to look deeper (assuming it's just a ui issue) | 15:03 |
| mnaser | in this case 689629,19 is also all in waiting, with no retries | 15:03 |
| tobiash | corvus: maybe a side effect as we store it in the job, not in the buildset | 15:06 |
| corvus | tobiash: in that case, i'd expect the opposite to be true (it would show queued instead of waiting) | 15:06 |
| tobiash | yes, weird | 15:07 |
| corvus | mnaser: the jobs in that change all seem to have nodepool requests too. and 3 of them just went from waiting to running | 15:09 |
| corvus | 2 tenant reconfigurations happened during the wait for those jobs | 15:12 |
| corvus | one of which was due to a change landing on openstack/openstack-ansible-tests @ master | 15:13 |
| mnaser | so maybe it dequeued them..? | 15:19 |
| mnaser | those are jobs which are defined there | 15:19 |
| tobiash | oh, those re-freeze the job | 15:19 |
| tobiash | so that makes sense that they are reset to waiting | 15:19 |
| tobiash | jobs are re-frozen but the node request already exists -> waiting state | 15:20 |
| tobiash | should be easy to fix then | 15:20 |
| corvus | yeah, that's what i'm thinking | 15:20 |
| corvus | tobiash: we can probably fix that in findjobstorequest -- just force the queued state in the loop above where it checks if there's already a request | 15:22 |
| tobiash | yepp | 15:23 |
| tobiash | then we make sure there is no similar issue :) | 15:23 |
| corvus | okay, i have written that one line. i will also write a test, which will take a bit. :) | 15:24 |
| mnaser | i seem to always wake corvus to some sort of weird zuul issue :) | 15:25 |
| corvus | the only thing i like better than a weird zuul issue is zuul not having any issues. | 15:25 |
| mnaser | :P | 15:29 |
| mnaser | i wonder if allowed-projects to pipeline makes sense | 15:37 |
| mnaser | i'm looking at the `release-approval` pipeline and im seeing changes like openstack/openstack-ansible changes in there | 15:37 |
| mnaser | and i can only assume there's an overhead of queue and dequeue | 15:37 |
| mordred | fungi, corvus: ugh. the gitdb issue openstack hit is hitting zuul-jobs py27 jobs | 15:47 |
| fungi | of course it is | 15:47 |
| mordred | https://zuul.opendev.org/t/zuul/build/0cf976a0c22a422d8756e4519fefd259/console | 15:47 |
| fungi | mnaser: zuul momentarily enqueues items into every pipeline to check whether they're applicable | 15:48 |
| fungi | it's more of an implementation detail, possible we could come up with a different way to reflect that in the status api (or hide it in the web dashboard) | 15:48 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Ensure job.queued is correct after reconfig https://review.opendev.org/709559 | 15:49 |
| mordred | fungi: I think maybe our mirrors just haven't caught up to the new gitdb releases? | 15:49 |
| mordred | ERROR: Could not find a version that satisfies the requirement gitdb>=4.0.1 (from gitdb2>=2.0.0->GitPython<2.1.12,>=2.1.8->-r /home/zuul/src/opendev.org/zuul/zuul-jobs/test-requirements.txt (line 5)) (from versions: 0.5.0, 0.5.1, 0.5.2, 0.5.3, 0.5.4, 0.6.0, 0.6.1, 0.6.2, 0.6.3, 0.6.4) | 15:49 |
| mordred | gitdb>=4.0.1 *does* exist on pypi | 15:50 |
| corvus | mnaser, pabelanger, tobiash: Ensure job.queued is correct after reconfig https://review.opendev.org/709559 | 15:50 |
| corvus | wow, our mirrors have cool 404 pages | 15:55 |
| corvus | mordred: http://mirror.bhs1.ovh.openstack.org/pypi/simple/gitdb/ does have 4.0.1 now | 15:56 |
| corvus | but maybe it didn't when the job ran? | 15:56 |
| corvus | the mirrors are proxies, now, right? | 15:56 |
| corvus | seems weird for it to fail then | 15:56 |
| mordred | corvus: maybe we just got unlucky with cache ttl of the index page? | 15:57 |
| mordred | corvus: f29 has been archived: http://mirror.sjc1.vexxhost.openstack.org/fedora/releases/29/README | 15:59 |
| mordred | corvus: so our fedora 29 docker jobs are also broken | 15:59 |
| fungi | mordred: as corvus mentioned, we don't mirror pypi any longer, so we could have a stale cache i suppose. this is under python 3.x right? | 16:02 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Fix unittests for python2 and fedora https://review.opendev.org/709302 | 16:02 |
| clarkb | fungi: indexes have a ttl of like 5-10 minutes. The packages themselves would be cached for a day (our max cache period) iirc | 16:03 |
| mordred | fungi: it was under 2.7 | 16:03 |
| mordred | clarkb: yeah - I'm guessing we just got REALLY unlucky with that index | 16:03 |
| clarkb | I think we've also simply seen pypi fail for $internet reasons | 16:03 |
| mordred | yeah | 16:03 |
| fungi | http://mirror.sjc1.vexxhost.openstack.org/pypi/simple/gitdb2/ shows 4.0.2 sdist and a python3-only wheel | 16:03 |
| *** jcapitao has quit IRC | 16:07 | |
| *** rishabhhpe has joined #zuul | 16:07 | |
| *** jcapitao has joined #zuul | 16:08 | |
| nborg1 | fungi: https://review.opendev.org/#/c/578557/17 is close to what we are looking for. Ideally, we would like the configuration from the tagged commit. We cannot use only branches where the tag is on the head of branches though (mentioned in a comment), as tagging is our only way to send meta-data to the jobs. We need to be able to rerun jobs with different meta-data, such as which firmware to build for. But | 16:11 |
| nborg1 | building from any branch containing the tag should be good enough. | 16:11 |
| fungi | nborg1: cool, feedback on that change would be great in that case | 16:12 |
| rishabhhpe | Hi All, My job is not getting executed for third party CI .. its erroring out in ansible .. can anyone here please suggest what i am missing in configuration -: http://paste.openstack.org/show/789940/ | 16:14 |
| rishabhhpe | i had pasted the error log above | 16:14 |
| Shrews | rishabhhpe: /bin/sh: 1: auto: not found i think you need ansible 2.8 if using 'auto' for the python interpreter | 16:18 |
| *** jcapitao is now known as jcapitao_afk | 16:18 | |
| rishabhhpe | shrews: i am not passing anything for ansible it is itself taking ansible 2.7 | 16:18 |
| clarkb | Shrews: it is nodepool that should set that | 16:19 |
| clarkb | iirc | 16:19 |
| Shrews | yeah, but i thought zuul handled that somewhere, too | 16:21 |
| Shrews | i seem to recall a bug fix for it... | 16:21 |
| rishabhhpe | clarkb: Shrews: is there anything missed configuration wise ? | 16:22 |
| clarkb | Shrews: ya I think its supposed to have an ansible version check that accomodates the value set by nodepool | 16:22 |
| clarkb | (and that is if unset then do auto for >=2.8 and python2 otherwise? | 16:22 |
| *** mattw4 has joined #zuul | 16:23 | |
| *** jcapitao_afk has quit IRC | 16:23 | |
| *** bolg has quit IRC | 16:24 | |
| *** bolg has joined #zuul | 16:24 | |
| Shrews | if rishabhhpe is running the latest zuul, that's probably a mis-direction error. hard to parse that paste :/ | 16:24 |
| *** jcapitao_afk has joined #zuul | 16:25 | |
| *** bolg has quit IRC | 16:25 | |
| Shrews | https://opendev.org/zuul/zuul/src/branch/master/zuul/executor/server.py#L1437-L1459 | 16:26 |
| Shrews | rishabhhpe: does your version of zuul have that bit of code? ^^ | 16:26 |
| Shrews | that's been in place for a while though, so probably not the issue | 16:28 |
| *** jcapitao_afk is now known as jcapitao | 16:32 | |
| *** Defolos has quit IRC | 16:37 | |
| mordred | fungi: failed again: https://zuul.opendev.org/t/zuul/build/0c040fcde1204c3b9b3004da2ca2c645 :( | 16:38 |
| *** jamesmcarthur has joined #zuul | 16:38 | |
| mordred | corvus: ^^ | 16:41 |
| mordred | now I'm really confused | 16:41 |
| corvus | mordred: and that's the same region/mirror | 16:41 |
| fungi | mordred: maybe gitdb2>4 is now python3-only? | 16:42 |
| mordred | oh. hrm. lemme see if I can reproduce that | 16:42 |
| mordred | corvus: do you know off the top of your head why zuul-jobs declares an explicit depend on GitPython? | 16:44 |
| corvus | no, i'll check the bottom of my head | 16:44 |
| mordred | fungi: yes - the same error happens against pypi | 16:44 |
| clarkb | the reason is the data-require-python | 16:44 |
| clarkb | its >=python3.4 | 16:44 |
| clarkb | see page source of http://mirror.bhs1.ovh.openstack.org/pypi/simple/gitdb/ | 16:45 |
| corvus | wow that would be a cool thing to put in the visible part of the page | 16:45 |
| mordred | awesome. | 16:45 |
| mordred | well - at least that means it's not going to kill us in zuul proper | 16:45 |
| clarkb | its visible in the non simple index too | 16:45 |
| clarkb | on the left side panel | 16:46 |
| mordred | clarkb, corvus, fungi: doesn't look like we need GitPython for zuul-jobs unittests- so maybe it's in there as a pin for the zuul install? | 16:47 |
| corvus | mordred: yeah, a74ff558167653f0010b1a7e002ad56f6e84db1c says it's a pin because gitpython is py3-only | 16:48 |
| corvus | and that's for the zuul install | 16:48 |
| mordred | corvus: so putting a python_version > "3" in there would not be incorrect | 16:49 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Fix unittests for python2 and fedora https://review.opendev.org/709302 | 16:49 |
| mordred | corvus: although - since we >3 zuul in this patch, we should be able to jisut remove gitpython from the file | 16:49 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Fix unittests for python2 and fedora https://review.opendev.org/709302 | 16:51 |
| openstackgerrit | Merged zuul/zuul master: Make most test cases work on MacOS https://review.opendev.org/707585 | 16:52 |
| corvus | mordred: yeah, i think that's the way to go | 16:58 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Add pause-buildset-registry role https://review.opendev.org/709256 | 17:17 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Fix cleanup of symlink fixtures https://review.opendev.org/709306 | 17:17 |
| *** jamesmcarthur has quit IRC | 17:17 | |
| mordred | fungi, clarkb: https://review.opendev.org/#/c/709302 is now green - feel like a quick +A? | 17:17 |
| mordred | Shrews: ^^ | 17:17 |
| *** jamesmcarthur has joined #zuul | 17:18 | |
| openstackgerrit | James E. Blair proposed zuul/nodepool master: k8s: allow port forwarding https://review.opendev.org/709259 | 17:19 |
| rishabhhpe | shrews: and clarkb: this is my zuul version zuul (3.10.2.dev53) | 17:19 |
| corvus | Shrews, tristanC: ^ can you take a look at 709259? i'm going to update the zuul change that uses it | 17:20 |
| rishabhhpe | what check i can inject which will help me in debugging ? | 17:20 |
| *** jamesmcarthur_ has joined #zuul | 17:20 | |
| *** jamesmcarthur has quit IRC | 17:21 | |
| clarkb | rishabhhpe: you could try setting the python version explicitly on the nodepool configuration | 17:21 |
| rishabhhpe | well i had did it from my end .. initially it was running with python2 and python3 both but then i set only python3 and removed python from nodepool while building the image but still lands in same error | 17:22 |
| rishabhhpe | clarkb: removed python2 | 17:22 |
| *** jcapitao has quit IRC | 17:25 | |
| clarkb | rishabhhpe: I think you can only set a single version not both | 17:26 |
| clarkb | I would try setting it to python2 or python3 and then zuul should use that version aiui (rather than setting auto) | 17:26 |
| clarkb | let me find docs | 17:26 |
| clarkb | rishabhhpe: https://zuul-ci.org/docs/nodepool/configuration.html#attr-diskimages.python-path set that value | 17:27 |
| rishabhhpe | clarkb: yes i removed python2 and used only python3 to be initialised the spawned VM for nova .. and let me check the doc u shared | 17:29 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Remove enum34 from test-requirements https://review.opendev.org/709575 | 17:33 |
| mordred | corvus, AJaeger: follow up to the previous one ^^ I think we can just remove the enum34 | 17:34 |
| *** evrardjp has quit IRC | 17:34 | |
| *** evrardjp has joined #zuul | 17:35 | |
| *** jamesmcarthur_ has quit IRC | 17:35 | |
| *** jamesmcarthur_ has joined #zuul | 17:38 | |
| AJaeger | mordred: cool - let's wait for tests and then I'll +2A | 17:44 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Stream output from kubectl pods https://review.opendev.org/709261 | 17:46 |
| openstackgerrit | Merged zuul/zuul-jobs master: Fix unittests for python2 and fedora https://review.opendev.org/709302 | 17:49 |
| openstackgerrit | Merged zuul/zuul-jobs master: Add pause-buildset-registry role https://review.opendev.org/709256 | 17:49 |
| mordred | AJaeger: sweet, thanks | 17:50 |
| openstackgerrit | yatin proposed zuul/zuul-jobs master: Use CentOS8 dependencies repo to install openvswitch https://review.opendev.org/708402 | 18:08 |
| *** jamesmcarthur_ has quit IRC | 18:14 | |
| *** jamesmcarthur has joined #zuul | 18:15 | |
| mordred | AJaeger: while you're in the reviewing mood, https://review.opendev.org/#/c/709306/ is a cleanup patch | 18:15 |
| AJaeger | ;) | 18:16 |
| openstackgerrit | Merged zuul/zuul-jobs master: Remove enum34 from test-requirements https://review.opendev.org/709575 | 18:19 |
| *** jamesmcarthur has quit IRC | 18:20 | |
| AJaeger | ianw pushed a change to use contraints for javascript installs, could you review that as well, please? https://review.opendev.org/709414 . It would have been great to had that earlier today - but I haven't thought this completely through yet and like input | 18:24 |
| *** jamesmcarthur has joined #zuul | 18:26 | |
| *** jamesmcarthur has joined #zuul | 18:26 | |
| openstackgerrit | Merged zuul/zuul-jobs master: Remove deprecated upload-afs role https://review.opendev.org/705373 | 18:27 |
| openstackgerrit | David Shrewsbury proposed zuul/nodepool master: Use Fedora 30 for openshift functional job https://review.opendev.org/709581 | 18:29 |
| *** igordc has joined #zuul | 18:36 | |
| openstackgerrit | James E. Blair proposed zuul/nodepool master: k8s: allow port forwarding https://review.opendev.org/709259 | 18:37 |
| *** chandankumar is now known as raukadah | 18:38 | |
| rishabhhpe | clarkb: after adding the python-path to python3 i able to pass the step for error "auto" but i landed in this error now -:http://paste.openstack.org/show/789949/ | 18:38 |
| clarkb | rishabhhpe: bubblewrap (the bwrap command, is a dependency of zuul) | 18:39 |
| rishabhhpe | well for this i already did changes in image-builder .. | 18:40 |
| openstackgerrit | Merged zuul/zuul-jobs master: Use CentOS8 dependencies repo to install openvswitch https://review.opendev.org/708402 | 18:40 |
| rishabhhpe | but still its not working .. is there any way to troubleshoot this ? | 18:41 |
| clarkb | rishabhhpe: it is required where you run the zuul-executor | 18:41 |
| clarkb | it isn't needed in your test images | 18:41 |
| rishabhhpe | clarkb: so do i need to install it in my zuul-executor box ? | 18:42 |
| clarkb | rishabhhpe: yes | 18:43 |
| rishabhhpe | ok let me try this | 18:43 |
| rishabhhpe | apart from this do i need to restart any service or build new image ? or current setup will only work after installing this ? | 18:43 |
| clarkb | you shouldn't need to restart anything, but if it still fails with that error then restart the zuul-executor service | 18:44 |
| rishabhhpe | Ok | 18:45 |
| *** dpawlik has quit IRC | 18:49 | |
| *** rishabhhpe has quit IRC | 19:19 | |
| openstackgerrit | Merged zuul/nodepool master: Use Fedora 30 for openshift functional job https://review.opendev.org/709581 | 19:21 |
| openstackgerrit | Merged zuul/nodepool master: k8s: allow port forwarding https://review.opendev.org/709259 | 19:25 |
| openstackgerrit | Jeremy Stanley proposed zuul/zuul-jobs master: Precreate AFS target directory in upload https://review.opendev.org/709594 | 19:31 |
| openstackgerrit | Jeremy Stanley proposed zuul/zuul-jobs master: Precreate AFS target directory in upload https://review.opendev.org/709594 | 19:40 |
| *** jamesmcarthur has quit IRC | 19:44 | |
| *** dpawlik has joined #zuul | 19:45 | |
| *** igordc has quit IRC | 19:58 | |
| openstackgerrit | Merged zuul/zuul-jobs master: Precreate AFS target directory in upload https://review.opendev.org/709594 | 20:00 |
| *** dpawlik has quit IRC | 20:14 | |
| mnaser | appreciate eyes on https://review.opendev.org/#/c/709424/ | 20:45 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Stream output from kubectl pods https://review.opendev.org/709261 | 20:48 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Add destructor for SshAgent https://review.opendev.org/709609 | 20:48 |
| corvus | mnaser, tobiash: \o/ re 709424 | 20:57 |
| mnaser | corvus: tested that one locally too already :) | 20:58 |
| clarkb | I'm trying to sort out our smmap/gitdb/GitPython trouble locally and noticed ERROR: cherrypy 18.5.0 has requirement cheroot>=8.2.1, but you'll have cheroot 8.0.0 which is incompatible. is that a known issue? | 20:59 |
| clarkb | I think upstream must've fixed the cheroot problem then wanted to run newer cheroot and dropping the 8.0.0 requirement will fix things for zuul? | 20:59 |
| mordred | clarkb: it seems like a reasonable hypothesis? | 20:59 |
| *** jamesmcarthur has joined #zuul | 21:23 | |
| *** igordc has joined #zuul | 21:30 | |
| *** Defolos has joined #zuul | 21:31 | |
| clarkb | re cheroot, tobiash's issue hasn't been fixed upstream so we either want to pin cherrypy to something that is known to work with 8.0.0 or just keep using it this way since it seems to work :) | 21:37 |
| *** jamesmcarthur has quit IRC | 21:39 | |
| *** Goneri has quit IRC | 21:40 | |
| tristanC | corvus: i don't understand your comment https://review.opendev.org/#/c/709261/2/zuul/executor/server.py@351 . it seems like this is going to leak kubectl process if we don't get a port | 21:41 |
| corvus | tristanC: oh, i didn't understand your original comment. | 21:43 |
| corvus | tristanC: i thought you were trying to bring attention to the case where subprocess.popen failed. i did not notice that self.fwd had not been assigned yet. | 21:45 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Stream output from kubectl pods https://review.opendev.org/709261 | 21:46 |
| corvus | tristanC: ^ thanks | 21:46 |
| tristanC | corvus: alright | 21:48 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Stream output from kubectl pods https://review.opendev.org/709261 | 21:52 |
| corvus | missing newline | 21:52 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Add destructor for SshAgent https://review.opendev.org/709609 | 21:52 |
| *** jamesmcarthur has joined #zuul | 21:55 | |
| clarkb | reading up on why cherrypy updated the cheroot dep I think 8.0.0 is actually fine. The code change in cherrpy was precipitated by a change in cheroot that was in cheroot==7.0.0 but when they updated the cherrypy side they set the dep requirement to whatever was current cheroot | 22:28 |
| clarkb | I think we just live with that error for now while we wait on a fix for tobiash's issue | 22:29 |
| *** Goneri has joined #zuul | 22:32 | |
| *** mattw4 has quit IRC | 22:36 | |
| *** jamesmcarthur has quit IRC | 22:38 | |
| *** Goneri has quit IRC | 22:40 | |
| *** jamesmcarthur has joined #zuul | 22:48 | |
| *** jamesmcarthur has quit IRC | 23:02 | |
| openstackgerrit | Merged zuul/gcp-authdaemon master: Initial commit https://review.opendev.org/707438 | 23:06 |
| *** mattw4 has joined #zuul | 23:18 | |
| *** mattw4 has quit IRC | 23:29 | |
| *** mattw4 has joined #zuul | 23:34 | |
| *** tosky has quit IRC | 23:43 | |
| *** jamesmcarthur has joined #zuul | 23:45 | |
| *** Defolos has quit IRC | 23:48 | |
| *** jamesmcarthur has quit IRC | 23:49 | |
| *** jamesmcarthur has joined #zuul | 23:50 | |
| *** jamesmcarthur has quit IRC | 23:55 | |
| *** igordc has quit IRC | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!