| *** jamesmcarthur has joined #zuul | 00:22 | |
| *** jamesmcarthur has quit IRC | 00:46 | |
| *** jamesmcarthur has joined #zuul | 00:47 | |
| *** jamesmcarthur has quit IRC | 00:52 | |
| *** mattw4 has quit IRC | 00:56 | |
| *** jamesmcarthur has joined #zuul | 01:11 | |
| *** SpamapS has quit IRC | 01:16 | |
| *** SpamapS has joined #zuul | 01:16 | |
| *** sshnaidm|afk has quit IRC | 01:42 | |
| *** SpamapS has quit IRC | 01:52 | |
| *** SpamapS has joined #zuul | 01:53 | |
| *** rlandy|bbl is now known as rlandy | 02:12 | |
| *** jamesmcarthur has quit IRC | 02:14 | |
| *** jamesmcarthur has joined #zuul | 02:32 | |
| *** rlandy has quit IRC | 02:40 | |
| *** bhavikdbavishi has joined #zuul | 03:00 | |
| *** SpamapS has quit IRC | 03:08 | |
| *** SpamapS has joined #zuul | 03:09 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/nodepool master: static: enable using a single host with different user or port https://review.opendev.org/659209 | 03:21 |
|---|---|---|
| *** SpamapS has quit IRC | 03:30 | |
| *** SpamapS has joined #zuul | 03:32 | |
| *** SpamapS has quit IRC | 03:40 | |
| *** jamesmcarthur has quit IRC | 03:40 | |
| *** SpamapS has joined #zuul | 03:41 | |
| *** jamesmcarthur has joined #zuul | 03:58 | |
| *** jamesmcarthur has quit IRC | 04:03 | |
| *** hashar has joined #zuul | 04:19 | |
| *** mattw4 has joined #zuul | 04:24 | |
| SpamapS | tristanC: worth noting that you can add the same host twice, as long as it has different inventory_hostname's | 04:37 |
| SpamapS | It's a problem for 'hosts: all', but I actually avoid that in my playbooks in general, because I often have ancillary temporary hosts in the inventory for delegations. | 04:39 |
| *** mattw4 has quit IRC | 04:39 | |
| tristanC | SpamapS: good idea, thanks, one less patch then :-) | 04:56 |
| *** bhavikdbavishi1 has joined #zuul | 05:12 | |
| *** bhavikdbavishi has quit IRC | 05:14 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 05:14 | |
| *** hashar has quit IRC | 05:24 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul master: bubblewrap: bind mount /etc/subuid https://review.opendev.org/659218 | 05:35 |
| *** bjackman_ has joined #zuul | 05:44 | |
| *** spsurya has joined #zuul | 05:49 | |
| tristanC | clarkb: corvus: it seems like a new version of skopeo is installed on the executor and the jobs using push-to-intermediate-registry are now failing because of a missing subuid file, for example: http://zuul.openstack.org/build/f385f07558b44495815fa7e65cf37267 | 05:52 |
| tristanC | i think https://review.opendev.org/659218 should fix that, but i don't know if your ubuntu ship with that file on the host | 05:53 |
| tristanC | iiuc, the ppa is building skopeo continuously and it pulled a change that is not released yet in: https://launchpadlibrarian.net/420875087/skopeo_0.1.36-1~dev~ubuntu18.04~ppa16_0.1.36-1~dev~ubuntu18.04~ppa17.diff.gz | 05:54 |
| *** bjackman_ has quit IRC | 05:57 | |
| *** bjackman_ has joined #zuul | 06:09 | |
| *** threestrands has joined #zuul | 06:12 | |
| *** bjackman_ has quit IRC | 06:23 | |
| *** abraverm has quit IRC | 06:41 | |
| *** bhavikdbavishi1 has joined #zuul | 06:49 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: WIP: Fix scheduler breakage when using periodic jobs with artifacts https://review.opendev.org/659222 | 06:50 |
| *** bhavikdbavishi has quit IRC | 06:51 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 06:51 | |
| *** bhavikdbavishi1 has joined #zuul | 06:55 | |
| *** bhavikdbavishi has quit IRC | 06:56 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 06:56 | |
| *** bjackman_ has joined #zuul | 06:56 | |
| *** abraverm has joined #zuul | 07:15 | |
| *** themroc has joined #zuul | 07:26 | |
| *** markwork has joined #zuul | 07:31 | |
| *** hashar has joined #zuul | 07:41 | |
| *** corvus has quit IRC | 07:47 | |
| *** bhavikdbavishi1 has joined #zuul | 07:51 | |
| *** corvus has joined #zuul | 07:52 | |
| *** bhavikdbavishi has quit IRC | 07:52 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 07:52 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul master: web: remove SafeLoader left-over from ZuulJSONEncoder https://review.opendev.org/659026 | 07:59 |
| *** markwork has quit IRC | 08:01 | |
| *** themroc has quit IRC | 08:06 | |
| *** themroc has joined #zuul | 08:06 | |
| *** bhavikdbavishi has quit IRC | 08:13 | |
| *** lennyb has joined #zuul | 08:16 | |
| *** threestrands has quit IRC | 08:18 | |
| *** abraverm has quit IRC | 08:22 | |
| *** abraverm has joined #zuul | 08:23 | |
| corvus | tristanC: the pre-run idea sounds promising :) | 08:37 |
| corvus | tobiash: i'm in munich airport (transferring to barcelona) :) | 08:38 |
| corvus | i reviewed a lot of changes on the plane | 08:39 |
| corvus | tristanC: https://review.opendev.org/659218 looks good -- the file that references exists on our executors | 08:44 |
| corvus | so to fix our jobs we're going to need to restart executors with that | 08:45 |
| corvus | we can either manually apply the patch, or make the jobs non-voting to get that merged, or update the executor config to add that path, or force-merge it | 08:47 |
| *** pcaruana has joined #zuul | 08:51 | |
| *** bhavikdbavishi has joined #zuul | 08:51 | |
| corvus | i'm sleep deprived, so i think i will take the safest route and make the jobs non-voting | 08:52 |
| * AJaeger waves to corvus (munich is just one hour south from here ;) | 08:57 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: bubblewrap: bind mount /etc/subuid https://review.opendev.org/659218 | 08:57 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Make image build jobs voting again https://review.opendev.org/659242 | 08:57 |
| corvus | zuul-maint: ^ those are high priority :) thanks tristanC | 08:58 |
| tobiash | corvus: looks like you had a long flight ;) | 08:59 |
| corvus | tobiash: 10.5 hours... then i have a 2 hour flight... then a 1 hour flight... | 08:59 |
| corvus | AJaeger: o/ | 09:00 |
| corvus | cause in the world of star alliance, munich is on the way to barcelona... | 09:00 |
| * tobiash waves at corvus | 09:00 | |
| corvus | the food situation at MUC is *much* better than FRA | 09:00 |
| tobiash | :) | 09:01 |
| openstackgerrit | Mark Meyer proposed zuul/zuul master: Create a basic Bitbucket build status reporter https://review.opendev.org/658335 | 09:06 |
| openstackgerrit | Mark Meyer proposed zuul/zuul master: Create a basic Bitbucket event source https://review.opendev.org/658835 | 09:06 |
| *** jangutter has quit IRC | 09:06 | |
| fbo_ | Hi, we discovered that issue with dequeue-ref usage yesterday: https://review.opendev.org/659110/ project name is not take in account there. | 09:13 |
| corvus | fbo_: thanks, lgtm | 09:17 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: Prevent Zuul scheduler to crash at startup if gerrit down https://review.opendev.org/576192 | 09:18 |
| fbo_ | thanks :) | 09:20 |
| *** sshnaidm has joined #zuul | 09:21 | |
| corvus | clarkb: i think https://review.opendev.org/649900 will be of interest to you | 09:32 |
| *** markwork has joined #zuul | 09:38 | |
| markwork | Hey ho, James E Blair? Can I'd like to tack on the functional testing in part to the third patchset and then maybe add a fourth with any leftovers | 09:40 |
| openstackgerrit | Monty Taylor proposed zuul/zuul master: Remove unneeded extras definition https://review.opendev.org/659255 | 09:40 |
| corvus | markwork: that sounds great | 09:40 |
| *** bhavikdbavishi1 has joined #zuul | 09:40 | |
| markwork | corvus: Cool, so for me I'll mark the first patchset as done and am waiting for the final review on the 2nd, and will continue with the event source today. | 09:41 |
| *** bhavikdbavishi has quit IRC | 09:41 | |
| corvus | markwork: given that, you can consider my -1 on that first patch temporary, otherwise it looks fine. we'll probably just land the whole series once everything is in place | 09:42 |
| corvus | markwork: and i suspect you can copy a lot of the functional test stuff from the pagure patch (and later, once both bitbucket and pagure merge, we'll probably swing back around and refactor that to reduce the duplication) | 09:44 |
| *** bhavikdbavishi1 has quit IRC | 09:45 | |
| *** sshnaidm is now known as sshnaidm|afk | 09:49 | |
| *** bhavikdbavishi has joined #zuul | 09:52 | |
| markwork | corvus: that sounds like a good deal :) | 09:54 |
| markwork | I got a RETRY_LIMIT from zuul-build-image... all the other checks look good. Something broken in the build infra? | 09:56 |
| *** hashar has quit IRC | 09:57 | |
| corvus | markwork: yeah, we have a patch working its way through to fix it | 09:57 |
| corvus | https://review.opendev.org/659218 | 09:57 |
| *** bhavikdbavishi has quit IRC | 10:17 | |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Fix missing check if logger is None https://review.opendev.org/659262 | 10:26 |
| *** yolanda_ has joined #zuul | 10:39 | |
| *** bhavikdbavishi has joined #zuul | 10:41 | |
| *** jangutter has joined #zuul | 11:14 | |
| *** abraverm has quit IRC | 11:17 | |
| *** hashar has joined #zuul | 11:30 | |
| tristanC | corvus: binding the subuid map may not be enough, it seems like you can't use that ppa version of skopeo (git master) in bubblewrap | 11:50 |
| *** bhavikdbavishi has quit IRC | 11:50 | |
| tristanC | corvus: a quick test show that it's going to fail with: "ERRO[0000] error opening "allow" to /proc/31/setgroups: write /proc/31/setgroups: operation not permitted" | 11:50 |
| *** panda is now known as panda|launch | 12:01 | |
| *** spsurya has quit IRC | 12:08 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Annotate node request processing with event id https://review.opendev.org/658889 | 12:17 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: WIP: Annotate builds with event id https://review.opendev.org/658895 | 12:18 |
| *** bhavikdbavishi has joined #zuul | 12:19 | |
| *** jamesmcarthur has joined #zuul | 12:19 | |
| *** dkehn has quit IRC | 12:25 | |
| pabelanger | wow, that is a large queue of patches in gate for zuul :D | 12:25 |
| pabelanger | merge all the things | 12:26 |
| *** jamesmcarthur has quit IRC | 12:29 | |
| *** rlandy has joined #zuul | 12:36 | |
| tristanC | corvus: the proper fix for skopeo failing would be to add all those capabilities to the bwrap command: https://github.com/containers/skopeo/blob/master/cmd/skopeo/unshare_linux.go#L9 | 12:36 |
| tristanC | corvus: or rollback to skopeo-0.1.35 | 12:37 |
| tristanC | here is a quick reproducer: http://paste.openstack.org/show/751417/ | 12:42 |
| tristanC | i think we should at least ensure the user running bwrap is not root if we set those caps | 12:45 |
| *** jamesmcarthur has joined #zuul | 12:50 | |
| *** hashar has quit IRC | 12:59 | |
| *** panda|launch is now known as panda | 13:05 | |
| *** bjackman_ has quit IRC | 13:06 | |
| *** bhavikdbavishi has quit IRC | 13:10 | |
| *** themr0c has joined #zuul | 13:20 | |
| *** themroc has quit IRC | 13:22 | |
| dmsimard | FYI ansible 2.8 is still on track for release tomorrow | 13:27 |
| tobiash | cool | 13:31 |
| mordred | tristanC: I think I'd vote for adding the capabilities and ensuring the user isn't root | 13:53 |
| tristanC | mordred: but if it was working without, maybe skopeo needs an option/fix to not create namespace for opendev use case (iiuc, remote docker api shouldn't need client cap) | 13:55 |
| mordred | tristanC: yeah - that too | 13:55 |
| mordred | tristanC: maybe we file a bug with skopeo (and that's a nice reproducer there) - but add the caps to zuul while we wait for it - and we can drop them again later once skopeo is updated. or we can keep them if they reject the bug :) | 13:57 |
| tristanC | mordred: couldn't we rollback the skopeo version installed on the executor nodes? | 13:58 |
| mordred | maybe that's an easier temporary fix | 14:00 |
| tristanC | mordred: iiuc, SETFCAP may be dangerous, e.g. breakout vector, or creating file that can't be removed... | 14:00 |
| tristanC | the version before skopeo_0.1.36-1~dev~ubuntu18.04~ppa16_0.1.36-1~dev~ubuntu18.04~ppa17 shouldn't be affected | 14:01 |
| mordred | that's such a nice version number | 14:01 |
| mordred | tristanC: so 0.1.36-1~dev~ubuntu16.04.2~ppa17 is the bad one, and you're saying kopeo - 0.1.36-1~dev~ubuntu16.04.2~ppa16 should be good? | 14:05 |
| mordred | I shouldn't ask questions about the actual package version not getting updated should I? | 14:05 |
| tristanC | mordred: iiuc yes, ppa16 shouldn't be affected | 14:05 |
| tristanC | mordred: there are no version because it seems like that ppa is continuously packaged, the offending commit isn't tagged (or packaged in fedora/centos) yet | 14:06 |
| mordred | ah - makes sense | 14:08 |
| mordred | tristanC: https://review.opendev.org/659296 Pin skopeo to unbreak skopeo+bubblewrap | 14:08 |
| mordred | tristanC: how does that look then? | 14:08 |
| tristanC | mordred: i can't tell how long previous version are available, isn't there a rollback/downgrade command to restore the previous installed version | 14:10 |
| mordred | tristanC: in this case, since it's a PPA, the previous versions are there indefinitely | 14:10 |
| mordred | tristanC: you can look at the 'superseeded' packages: | 14:11 |
| mordred | https://launchpad.net/~projectatomic/+archive/ubuntu/ppa/+packages?field.name_filter=skopeo&field.status_filter=&field.series_filter=xenial | 14:11 |
| tristanC | mordred: hum, ppa16 was for bionic, in xenial that number seems affected | 14:14 |
| tristanC | you need ppa15, at least according to: https://launchpadlibrarian.net/420904271/skopeo_0.1.36-1~dev~ubuntu16.04.2~ppa14_0.1.36-1~dev~ubuntu16.04.2~ppa15.diff.gz | 14:15 |
| mordred | oh good | 14:15 |
| tristanC | i meant, ppa14, bug was introduced in ppa15 | 14:16 |
| mordred | doh | 14:16 |
| *** jamesmcarthur has quit IRC | 14:25 | |
| clarkb | mordred: corvus it is my understanding that 649900 will return opendev's zuul to the state before zaro's gerrit change which we specificially had zaro change in gerrit because we didn't like that old behavior | 14:47 |
| clarkb | in particular users have to remove their approval and reapply it for things to trigger when they recomment | 14:47 |
| clarkb | rather than doing what they mean and treating it as another approval if they haven't removed their vote | 14:48 |
| clarkb | (I've left a -1 comment on the change basically saying we wanted this behavior, got khai to change gerrit, and the old behavior is confusing) | 14:50 |
| *** gchenuet has joined #zuul | 15:12 | |
| gchenuet | Hi guys! I noticed that our `/data` partition in executor are growing fast as Zuul keep all Ansible run (code + logs). Is it the correct or a bug ? Should I need to create a crontask ? | 15:14 |
| clarkb | gchenuet: by default zuul should clean that up, there is a debugging flag you can set to keep that data though. Is it possible you have that flag set? | 15:14 |
| gchenuet | clarkb: yes! I have enable the debug flag for a job. Thanks for the reminder. | 15:19 |
| clarkb | gchenuet: I think you may have to cleanup the data that was kept by the debug state manually, but once unset any new jobs should clean up after themselves | 15:19 |
| *** rfolco|ruck is now known as rfolco | 15:22 | |
| *** panda is now known as panda|ruck | 15:26 | |
| gchenuet | clarkb: ok. thanks! | 15:32 |
| *** themr0c has quit IRC | 15:52 | |
| *** hashar has joined #zuul | 15:53 | |
| openstackgerrit | Clark Boylan proposed zuul/zuul master: Handle artifacts on non change types. https://review.opendev.org/659329 | 15:54 |
| clarkb | tobiash: ^ if you are still around your review on that would be helpful | 15:54 |
| clarkb | (you wrote the earlier change that caused the issue so I'm guessing you understand the problem space a bit better than I do) | 15:54 |
| tobiash | clarkb: I wrote a similar fix today, I'll look in a bit | 16:08 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: Pagure driver - https://pagure.io/pagure/ https://review.opendev.org/604404 | 16:11 |
| tobiash | clarkb: we also had another run_handler exception problem today: https://review.opendev.org/659262 | 16:15 |
| tobiash | mordred: ^ | 16:15 |
| tobiash | clarkb: lgtm, but we definitely should test this | 16:18 |
| clarkb | ya I wanted to make sure I wasn't missing something important first | 16:18 |
| clarkb | I can work on tests in a bit | 16:18 |
| tobiash | clarkb: do you need assistance with the test? | 16:18 |
| clarkb | maybe? I was going to look at the test you wrote for the earlier change and udpate it to process tag events | 16:19 |
| clarkb | tobiash: I do need to step away from the computer for a bit and eat braekfasty | 16:20 |
| clarkb | so if you would like to add a test more quickly than I can feel free | 16:20 |
| tobiash | I have to grab dinner first as well ;) | 16:21 |
| clarkb | actually it might be easy to make this test so I'll try a quick stab at it | 16:25 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: Pagure driver - https://pagure.io/pagure/ https://review.opendev.org/604404 | 16:25 |
| *** hashar has quit IRC | 16:31 | |
| *** hashar has joined #zuul | 16:32 | |
| *** hashar has quit IRC | 16:52 | |
| *** flaper87 has joined #zuul | 17:10 | |
| *** gchenuet has quit IRC | 17:13 | |
| flaper87 | clarkb: o/ if you're around and have like 5mins to help me out, I'd super appreciate it. If you wonder why I'm pinging you directly is because someone told me that you're very good at debugging zuul deployments :D | 17:20 |
| flaper87 | I'm running the docker-compose example, slightly modified, and I'm seeing this in the scheduler: https://paste.fedoraproject.org/paste/BQlF9VexjUvrkWWnAbKtgw | 17:20 |
| flaper87 | I'm trying to point zuul to github instead of using the local gerrit (That's the slightly modified part) | 17:21 |
| flaper87 | I honestly have no idea where to look :D | 17:21 |
| nickx-intel | can you please paste your ../zuul/.gitreview flaper87 ? | 17:22 |
| nickx-intel | I don't know if/how zuul works with github but whatever config you have in there is likely relevant | 17:23 |
| pabelanger | flaper87: if you are using github, it looks like your merger might not be able to communicate with it. I'd check merger-debug.log for exception | 17:23 |
| pabelanger | you might have wrong credentials | 17:23 |
| flaper87 | pabelanger: that is entirely possible. I created the webhook token but I'm not sure if it's correct. :/ | 17:23 |
| flaper87 | pabelanger: lemme look | 17:24 |
| flaper87 | pabelanger: where can I find those logs in the docker-compose example? | 17:24 |
| clarkb | flaper87: docker logs $containername | 17:25 |
| clarkb | flaper87: and docker ps -a will give you the containers it is running | 17:25 |
| flaper87 | yeah, I meant in what container | 17:25 |
| flaper87 | :) | 17:25 |
| flaper87 | I don't see a merger container | 17:25 |
| pabelanger | do we run a dedicated merger for docker-compose | 17:25 |
| pabelanger | you might want to look in zuul-exeuctor too | 17:25 |
| flaper87 | I see a bunch of failures saying the executor couldn't talk to the schedule. | 17:27 |
| flaper87 | lemme check if I messed something up | 17:27 |
| flaper87 | oh, yeha, I see the GH error now | 17:28 |
| pabelanger | cool | 17:29 |
| flaper87 | pabelanger: atal: could not read Username for 'https://github.com': No such device or address' | 17:29 |
| * flaper87 digs into this | 17:30 | |
| pabelanger | flaper87: did you setup your zuul.conf like: https://zuul-ci.org/docs/zuul/admin/github_setup.html | 17:30 |
| pabelanger | that's for using a github app | 17:30 |
| flaper87 | pabelanger: I didn't use apps | 17:30 |
| flaper87 | do I have to use GH apps? | 17:30 |
| flaper87 | I just created a token | 17:31 |
| clarkb | tobiash: there is a bug in my chagne (caught by tests so that is good) I have a working test for the tag case but the existing test fails | 17:31 |
| tobiash | cool | 17:31 |
| flaper87 | pabelanger: token is the issue | 17:34 |
| flaper87 | thanks | 17:34 |
| pabelanger | cool | 17:34 |
| pabelanger | flaper87: I think there is more coverage on github apps, but non github apps should work | 17:34 |
| clarkb | tobiash: ok I think I see the issue, it is the addition of the branch info | 17:35 |
| tobiash | ah ok | 17:36 |
| openstackgerrit | Clark Boylan proposed zuul/zuul master: Handle artifacts on non change types. https://review.opendev.org/659329 | 17:38 |
| clarkb | tobiash: ^ there that should pass tests now and actually test the change. I've put a TODO in there about how things get weird in a particular spot (basically we assume changes for artifacts in the test suite) | 17:39 |
| *** mattw4 has joined #zuul | 17:40 | |
| clarkb | where did that .keep file come from? | 17:42 |
| clarkb | why does that file keep causing problems for us? | 17:44 |
| clarkb | I have ahd to untangle it ~3 times now | 17:44 |
| openstackgerrit | Clark Boylan proposed zuul/zuul master: Handle artifacts on non change types. https://review.opendev.org/659329 | 17:45 |
| tobiash | clarkb: the yarn thing deletes it | 17:45 |
| clarkb | latest patchset adds it back in again | 17:45 |
| *** mattw4 has quit IRC | 17:45 | |
| clarkb | should be ready for review | 17:46 |
| tobiash | the keep file is there so the target dir exists in git, but yarn deletes the content of that dir during build | 17:46 |
| tobiash | lgtm | 17:46 |
| clarkb | can we maybe add the favico.ico instead as that should be fairly static and yarn will put it back again? | 17:46 |
| pabelanger | is it same to recheck zuul failures or is docker job still failing? | 17:46 |
| tobiash | clarkb: maybe we can on demand the dir instead before running yarn | 17:47 |
| clarkb | pabelanger: the tls'd mirrors is still a wip | 17:47 |
| clarkb | pabelanger: if you'd like to help review changes I can dig them up for you | 17:47 |
| clarkb | https://review.opendev.org/#/c/658281/ and its children | 17:48 |
| pabelanger | clarkb: sorry, I mean the current zuul-upload-image failures, is that related to tls mirrors? | 17:48 |
| clarkb | pabelanger: I believe the issues we have been having all stem back to trying to use shared NAT for ipv4 to talk to dockerhub via the buildset registry | 17:49 |
| tobiash | pabelanger: I at least saw no successful build yet | 17:49 |
| clarkb | but i haven't looked again today to see if there are new problems | 17:49 |
| pabelanger | http://logs.openstack.org/49/649949/1/gate/zuul-upload-image/a8b9625/job-output.txt.gz#_2019-05-15_13_06_50_880419 | 17:49 |
| pabelanger | open /etc/subuid: no such file or directory | 17:49 |
| pabelanger | checking backscroll now | 17:49 |
| tobiash | clarkb: there was/is a skopeo update that now requires privileges | 17:49 |
| clarkb | tobiash: oh fun | 17:49 |
| pabelanger | okay | 17:50 |
| clarkb | I jumped right into fixing the pipelines issue I didn't realize there was another issue | 17:50 |
| pabelanger | I see https://review.opendev.org/659296 | 17:50 |
| pabelanger | that is to pin | 17:50 |
| pabelanger | https://review.opendev.org/659218/ looks to also be related | 17:50 |
| pabelanger | but would need a release of zuul | 17:50 |
| clarkb | so fix process is rollback skopeo, merge zuul side fix, release that, then we can update skopeo again? | 17:51 |
| tobiash | pabelanger: tristanC pointed out that https://review.opendev.org/659218 won't fix it unfortunately | 17:51 |
| *** hashar has joined #zuul | 17:52 | |
| *** hashar is now known as hasharAway | 17:52 | |
| tobiash | related backscroll: http://eavesdrop.openstack.org/irclogs/%23zuul/%23zuul.2019-05-15.log.html#t2019-05-15T12:36:30 | 17:53 |
| tobiash | why do we run skopeo on the executor? | 17:54 |
| pabelanger | tobiash: ah, I see that now | 17:54 |
| pabelanger | tobiash: not sure, maybe to reduce the cost of using nodepool? | 17:54 |
| pabelanger | I have a recheck, +3 on 659296 now | 17:57 |
| tobiash | clarkb: did i understand correctly that the ipv6 problems required to run skopeo on the executor combined with socat magic? | 18:03 |
| pabelanger | okay, looks like some fallout to puppet breaking something in system-config, so zuul will be blocked for a bit, unless we flip job non-voting | 18:03 |
| clarkb | tobiash: yes though that was a different ipv6 problem than the current one | 18:03 |
| *** rlandy is now known as rlandy|brb | 18:12 | |
| *** ianychoi has quit IRC | 18:42 | |
| *** hasharAway is now known as hashar | 18:45 | |
| *** rlandy|brb is now known as rlandy | 18:45 | |
| *** mattw4 has joined #zuul | 19:09 | |
| *** mattw4 has quit IRC | 19:13 | |
| *** dkehn has joined #zuul | 19:19 | |
| openstackgerrit | Mark Meyer proposed zuul/zuul master: Create a basic Bitbucket event source https://review.opendev.org/658835 | 19:56 |
| *** markwork has quit IRC | 19:59 | |
| *** pcaruana has quit IRC | 20:14 | |
| *** rlandy has quit IRC | 20:15 | |
| *** rlandy has joined #zuul | 20:16 | |
| *** panda|ruck has quit IRC | 21:10 | |
| *** panda has joined #zuul | 21:14 | |
| *** hashar has quit IRC | 21:43 | |
| *** jlk has quit IRC | 21:51 | |
| *** jlk has joined #zuul | 21:53 | |
| *** mattw4 has joined #zuul | 21:56 | |
| *** mattw4 has quit IRC | 22:03 | |
| *** rlandy is now known as rlandy|bbl | 22:18 | |
| nickx-intel | I have a question ..... as I'm looking at zuul-jobs, this syntax seems pretty arcane | 22:28 |
| nickx-intel | .. zuul:rolevar:: zuul_output_dir | 22:28 |
| nickx-intel | :default: {{ ansible_user_dir }}/zuul-output | 22:28 |
| nickx-intel | I'm gathering that {{ zuul_output_dir }} is actually the syntax | 22:29 |
| clarkb | nickx-intel: those are python sphinx rst annotations. They make the rendered output prettier | 22:31 |
| clarkb | https://zuul-ci.org/docs/zuul/developer/docs.html#zuul-sphinx-directives is where they are documented | 22:32 |
| clarkb | now to find a link to the rendered docs for zuul_output_dir | 22:32 |
| clarkb | https://zuul-ci.org/docs/zuul-jobs/log-roles.html#rolevar-ensure-output-dirs.zuul_output_dir there | 22:33 |
| nickx-intel | hmmm .... thanks clarkb .... I'm just swimming in this right now | 22:38 |
| *** mattw4 has joined #zuul | 23:01 | |
| *** mattw4 has quit IRC | 23:21 | |
| *** panda has quit IRC | 23:30 | |
| *** panda has joined #zuul | 23:31 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!