| *** mattw4 has joined #zuul | 00:01 | |
| *** rlandy|rover has quit IRC | 00:09 | |
| *** jamesmcarthur has joined #zuul | 00:09 | |
| *** mattw4 has quit IRC | 00:17 | |
| *** jamesmcarthur has quit IRC | 00:26 | |
| *** rfolco has joined #zuul | 00:29 | |
| *** jamesmcarthur has joined #zuul | 00:31 | |
| *** jamesmcarthur has quit IRC | 00:35 | |
| *** rf0lc0 has joined #zuul | 00:35 | |
| *** rfolco has quit IRC | 00:36 | |
| *** spsurya has joined #zuul | 01:05 | |
| *** rfolco has joined #zuul | 01:16 | |
| *** tributarian has quit IRC | 01:17 | |
| *** rf0lc0 has quit IRC | 01:18 | |
| *** igordc has quit IRC | 01:25 | |
| *** jamesmcarthur has joined #zuul | 02:14 | |
| *** rfolco has quit IRC | 02:21 | |
| *** jamesmcarthur has quit IRC | 02:37 | |
| *** altlogbot_1 has quit IRC | 02:44 | |
| *** altlogbot_0 has joined #zuul | 02:46 | |
| *** bhavikdbavishi has joined #zuul | 03:18 | |
| *** bhavikdbavishi1 has joined #zuul | 03:21 | |
| *** bhavikdbavishi has quit IRC | 03:23 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 03:23 | |
| *** jamesmcarthur has joined #zuul | 04:24 | |
| *** smyers has quit IRC | 04:25 | |
| *** raukadah is now known as chkumar|ruck | 04:38 | |
| *** ianychoi_ has quit IRC | 04:44 | |
| *** jamesmcarthur has quit IRC | 05:03 | |
| *** bjackman has joined #zuul | 05:05 | |
| *** smyers has joined #zuul | 05:29 | |
| *** jamesmcarthur has joined #zuul | 05:30 | |
| *** panda has quit IRC | 05:47 | |
| *** panda has joined #zuul | 05:48 | |
| openstackgerrit | Merged zuul/zuul master: web: refactor the errorsIds into the build action https://review.opendev.org/675350 | 05:52 |
|---|---|---|
| *** jamesmcarthur has quit IRC | 06:03 | |
| openstackgerrit | Merged zuul/zuul master: Correctly identify failed tasks https://review.opendev.org/675488 | 06:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-base-jobs master: Update ansible and ansible-lint for tests https://review.opendev.org/675530 | 06:29 |
| *** jamesmcarthur has joined #zuul | 07:00 | |
| *** tosky has joined #zuul | 07:23 | |
| *** jamesmcarthur has quit IRC | 07:34 | |
| *** Diabelko has joined #zuul | 08:04 | |
| openstackgerrit | Merged zuul/zuul-jobs master: Don't install centos repos on RHEL https://review.opendev.org/674572 | 08:31 |
| *** jamesmcarthur has joined #zuul | 08:32 | |
| *** AshBullock has joined #zuul | 08:57 | |
| *** jamesmcarthur has quit IRC | 09:05 | |
| *** AshBullock has quit IRC | 09:28 | |
| *** bjackman has quit IRC | 09:36 | |
| *** bjackman has joined #zuul | 09:38 | |
| *** jamesmcarthur has joined #zuul | 09:41 | |
| *** bjackman has quit IRC | 10:07 | |
| *** bjackman has joined #zuul | 10:10 | |
| *** jamesmcarthur has quit IRC | 10:15 | |
| *** bhavikdbavishi has quit IRC | 10:37 | |
| *** rfolco has joined #zuul | 11:11 | |
| *** jamesmcarthur has joined #zuul | 11:11 | |
| mordred | tristanC: love the buildset page - on the link-to-buildset page, I'm getting errors in the dashboard | 11:27 |
| *** jamesmcarthur has quit IRC | 11:27 | |
| *** jamesmcarthur has joined #zuul | 11:32 | |
| *** jamesmcarthur has quit IRC | 11:36 | |
| *** jamesmcarthur has joined #zuul | 11:41 | |
| tristanC | mordred: perhaps zuul web needs to be restarted with https://review.opendev.org/674759? | 11:51 |
| mordred | tristanC: oh - yeah - that would probably do it | 11:52 |
| tristanC | mordred: well, the build page could also silently skip the row when the buildset uuid is missing | 11:53 |
| mordred | tristanC: yeah- I was just about to ask that - but then I was thinking that would be cruft code in a very short amount of time | 11:53 |
| mordred | but is probably more technically correct for handling the CD case gracefully | 11:53 |
| *** rlandy has joined #zuul | 11:59 | |
| *** rlandy is now known as rlandy|rover | 11:59 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul master: web: link the buildset page from the build https://review.opendev.org/675493 | 11:59 |
| *** bhavikdbavishi has joined #zuul | 12:15 | |
| openstackgerrit | Merged zuul/zuul master: web: add buildset page https://review.opendev.org/630079 | 12:15 |
| *** bhavikdbavishi has quit IRC | 12:19 | |
| *** jamesmcarthur has quit IRC | 12:23 | |
| *** bhavikdbavishi has joined #zuul | 12:38 | |
| *** jamesmcarthur has joined #zuul | 12:51 | |
| *** bjackman has quit IRC | 13:02 | |
| *** panda has quit IRC | 13:14 | |
| *** panda has joined #zuul | 13:15 | |
| *** jeliu_ has joined #zuul | 13:23 | |
| *** jamesmcarthur has quit IRC | 13:39 | |
| *** jamesmcarthur has joined #zuul | 13:41 | |
| *** jamesmcarthur has quit IRC | 13:48 | |
| *** gtema_ has joined #zuul | 13:48 | |
| *** spsurya has quit IRC | 13:54 | |
| pabelanger | corvus: it seems to have worked, I see zuul-manifest.json https://logs.zuul.ansible.com/70/170/f347ed2b5e80e72a40473809d5b5309e5998d9fa/check/ansible-network-vyos-appliance/6e863c2/ | 13:54 |
| *** chkumar|ruck is now known as raukadah | 13:57 | |
| corvus | pabelanger: that looks good, ready to merge https://github.com/ansible/project-config/pull/172 now? | 14:13 |
| pabelanger | yup! | 14:14 |
| corvus | pabelanger: next we'll need to run this script: http://paste.openstack.org/show/755695/ | 14:25 |
| pabelanger | sure | 14:26 |
| corvus | pabelanger: if you set up a clouds.yaml for vexxhost, you can run that script with "python3 update.py vexxhost logs" if 'logs' is the name of the container (i can't remember) | 14:26 |
| pabelanger | k, installing dependencies now | 14:27 |
| corvus | pabelanger: it'll output 2 lines of info, if you can pastebin them, i'd appreciate it | 14:30 |
| pabelanger | done: http://paste.openstack.org/show/755697/ | 14:32 |
| pabelanger | guess I should look to create new containers in other regions too | 14:33 |
| corvus | pabelanger: that's weird -- i expected to see "'X-Container-Meta-Access-Control-Allow-Origin': '*'" in the second line, but i don't | 14:34 |
| corvus | pabelanger: can you run it again and paste the output, just to eliminate any weird caching? | 14:34 |
| pabelanger | sure: http://paste.openstack.org/show/755698/ | 14:35 |
| corvus | okay, it must have been caching somewhere, because it's in both lines there. so we should be good. i'll just test it out now... | 14:37 |
| pabelanger | k, I have to run into town for an hour to run an errand, but happy to help once back | 14:37 |
| corvus | pabelanger: neat, i guess that worked because now we've managed to put 2 access-control headers on there. i guess 1 is from swift and the other is from the proxy. so i think we should remove the add-header line from the proxy | 14:45 |
| corvus | curl -H 'Origin: example.com' https://logs.zuul.ansible.com/86/60086/b2208a233da84f13fdce193df42405d64a572bfa/third-party-check/ansible-test-network-integration-vyos-python37/31c168c/job-output.json --output /dev/null -D - | 14:46 |
| corvus | will demonstrate ^ | 14:46 |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 14:48 |
| corvus | pabelanger: https://github.com/ansible-network/windmill-config/pull/488 | 15:07 |
| *** panda has quit IRC | 15:19 | |
| clarkb | corvus: what ended up being the cors issue in ovh? | 15:19 |
| clarkb | I missed how that was fixed | 15:20 |
| *** panda has joined #zuul | 15:20 | |
| corvus | clarkb: my error in testing. it turns out that swift doesn't return the cors header unless an Origin header is sent. i believe that's the case for radosgw in vexxhost too. | 15:21 |
| clarkb | got it | 15:22 |
| corvus | of course, i'm really confused about why the radosgw docs say they don't support cors (maybe they mean they don't support anything other than * ?) | 15:22 |
| corvus | and, of course, rax does always send the header... so... | 15:22 |
| clarkb | corvus: ya maybe they mean the fine grained request types that s3 supports? | 15:22 |
| corvus | yeah. because [real] swift *does* support that | 15:22 |
| mordred | corvus: or maybe they mean they don't support server-side enforcement | 15:27 |
| *** gtema_ has quit IRC | 15:43 | |
| *** jamesmcarthur has joined #zuul | 15:54 | |
| openstackgerrit | Merged zuul/zuul master: Refactor task result detection https://review.opendev.org/675489 | 16:12 |
| *** armstrongs has joined #zuul | 16:16 | |
| pabelanger | corvus: approved | 16:18 |
| corvus | pabelanger: cool, lemme know when it's deployed and we can test again | 16:18 |
| pabelanger | github events are slow today, hopefully not long | 16:19 |
| SpamapS | everybody turning on CICD today | 16:19 |
| pabelanger | in our case, i just think ansible/ansible is seeing more activity with ansible 2.9 window starting to close | 16:20 |
| corvus | pabelanger: if this works, i think i have enough info to send an email to the list letting swift users know what to do to take advantage of the new log stuff | 16:21 |
| *** armstrongs has quit IRC | 16:21 | |
| corvus | SpamapS: when you have a minute i'd like to work with you on seeing if any changes are needed for the s3 log storage you're using | 16:22 |
| corvus | SpamapS: (briefly -- there's a bunch of new in-browser log features i'm adding to zuul, we're just about ready to start rolling them out in opendev, but they require a cors header setting on the log storage; i don't expect it to be a problem -- s3 should support what we need) | 16:23 |
| corvus | SpamapS: http://zuul.opendev.org/t/zuul/build/daff0fbd153b4b969103fc7ffc172661 if you want to take a peek -- see the 'logs' and 'console' tabs at the top | 16:24 |
| *** mattw4 has joined #zuul | 16:28 | |
| SpamapS | corvus: I love the feature, and yeah, I think if a CORS header is all that's needed that's a very simple change. | 16:35 |
| corvus | SpamapS: i'll write up the process for testing and updating swift, and that'll mostly apply for s3 except if we need to reconfigure the bucket, that'll obviously be different. so maybe i'll ping you next week after i send out that email | 16:38 |
| SpamapS | I'm pretty sure it's per object but I could be wrong. | 16:39 |
| SpamapS | have not looked closely | 16:39 |
| pabelanger | corvus: okay, change should be promoted now | 16:39 |
| corvus | SpamapS: i *thought* i saw a bucket option, but i dunno. | 16:39 |
| SpamapS | It's possible there's both | 16:40 |
| corvus | yeah, that's actually the case on rackspace | 16:40 |
| corvus | pabelanger: cool, https://dashboard.zuul.ansible.com/t/ansible/build/31c168cc83b84c0b9a64d3b869ab7064 looks like it's working again, so i think you're all ready | 16:41 |
| corvus | pabelanger: we'll flip the switch for opendev on monday, and if there aren't any problems, make a release next week and then you should be good to do the same | 16:41 |
| pabelanger | ++ | 16:42 |
| corvus | SpamapS: are you using the 'download-artifact' role at all? | 16:42 |
| *** armstrongs has joined #zuul | 16:47 | |
| *** irclogbot_1 has quit IRC | 16:47 | |
| armstrongs | Does the zuul master branch have the new embedded logging and console tabs yet? If so is there any specific config i need to get them populated. I see it on the upstream instance and i was keen to use it too on my instance :) | 16:47 |
| corvus | armstrongs: we were just discussing that -- yes there are some things that need to be done, but i haven't written up the process yet. i will send out an email next week describing it. | 16:48 |
| armstrongs | thanks sorry for being impatient it just looks awesome | 16:48 |
| corvus | armstrongs: what storage backend are you using? static log server or swift? | 16:49 |
| corvus | or something else | 16:49 |
| armstrongs | just using a local file-system | 16:49 |
| corvus | armstrongs: ok -- serving via apache? | 16:49 |
| armstrongs | so 2 ha logging servers with mounted nfs | 16:49 |
| armstrongs | yeah | 16:49 |
| armstrongs | apache | 16:49 |
| corvus | cool, shouldn't be a problem; i'll make sure to include a section on that :) | 16:50 |
| armstrongs | great news thanks again | 16:50 |
| armstrongs | and great work too looks ace | 16:50 |
| corvus | thanks :) | 16:50 |
| *** armstrongs has quit IRC | 16:51 | |
| *** irclogbot_2 has joined #zuul | 16:51 | |
| *** jamesmcarthur has quit IRC | 16:59 | |
| *** igordc has joined #zuul | 17:17 | |
| *** jamesmcarthur has joined #zuul | 17:21 | |
| *** bhavikdbavishi has quit IRC | 17:40 | |
| *** jamesmcarthur has quit IRC | 17:50 | |
| SpamapS | corvus:no | 17:51 |
| SpamapS | corvus:though I expect we'll start using it soon | 17:51 |
| corvus | SpamapS: cool, wait till after monday :) | 17:53 |
| * SpamapS git stashes | 17:57 | |
| *** odyssey4me has left #zuul | 19:45 | |
| openstackgerrit | Jeff Liu proposed zuul/zuul-jobs master: Add skip-registry-check parameter to deploy openshift role https://review.opendev.org/675688 | 19:45 |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 19:47 |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 19:49 |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 20:39 |
| *** EmilienM is now known as EmilienM|pto | 21:01 | |
| *** igordc has quit IRC | 21:01 | |
| *** shachar has quit IRC | 21:01 | |
| *** shachar has joined #zuul | 21:01 | |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 21:19 |
| openstackgerrit | Jeremy Stanley proposed zuul/zuul-jobs master: Update documentation for Debian/Ubuntu mirror_info https://review.opendev.org/675703 | 21:21 |
| *** jeliu_ has quit IRC | 21:21 | |
| fungi | corvus: ^ it wasn't clear to me where the actual implementation corresponding to that documentation lives. was it just a sketch? | 21:22 |
| fungi | that's my attempt to more closely conform it to reality | 21:22 |
| corvus | fungi: yeah, it's documentation only at this point; the next step is to implement an alternative mirror config role that implements it | 21:23 |
| corvus | or, really, probably one new role for each kind of mirror | 21:24 |
| corvus | then get people using those and retire the current configure-mirrors role | 21:24 |
| fungi | okay, cool, well if that schema looks sane then hopefully it can serve as a usable model | 21:24 |
| corvus | so this is exactly the right time to make this doc update :) | 21:24 |
| fungi | what was there wasn't as flexible and was missing some bits i expect we'd need | 21:24 |
| fungi | it's also not totally clear to me that debian and ubuntu need to be separate vars... the schema should be identical because they both use the same toolchain and configuration format, so only the values would differ | 21:25 |
| fungi | you can sort of see from the example comparison there how they differ for debian/buster and ubuntu/bionic | 21:26 |
| corvus | fungi: i expect a role that installed these would need to look up the value to use based on the os it's running on; so if a site had ubuntu mirrors but not debian, then we'd need to reflect that | 21:26 |
| fungi | makes sense, mainly wondering how to cut down on the obvious amounts of duplication there, but it's not critical | 21:27 |
| corvus | (i wonder if we should also include information about releases, so that we could differ the values based on that....) | 21:27 |
| fungi | yeah, my main concern with abstracting it too much is that not everybody is going to mirror all the same repositories and suites, and how they name the paths for some of them can influence suite names too | 21:28 |
| corvus | fungi: oh i see buster in the suite... i wonder if it would be sufficient for the role to iterate through the entries and only include the ones for the current os? | 21:29 |
| fungi | and also they might do things like combine debian and debian-security into a single repository (which is doable since the suite names for them intentionally don't overlap), so could end up with just the suite names differing at the same url | 21:29 |
| fungi | which is more akin to how ubuntu arranges theirs | 21:30 |
| corvus | fungi: put another way -- maybe we should make sure that example has 'stretch' in it too :) | 21:30 |
| fungi | ahh, yeah, i'll toss it and maybe xenial in there for clarity | 21:30 |
| *** rlandy|rover has quit IRC | 21:31 | |
| corvus | fungi: would it make sense for suite to be a list? | 21:32 |
| openstackgerrit | Jeremy Stanley proposed zuul/zuul-jobs master: Update documentation for Debian/Ubuntu mirror_info https://review.opendev.org/675703 | 21:32 |
| fungi | lemme think about that a sec | 21:33 |
| fungi | yeah, i think that ought to work and could simplify things too | 21:34 |
| fungi | though it does make it harder for a given node to know which ones to use, right? | 21:34 |
| fungi | it might be that we need these to actually be named for their releases, so not ubuntu but ubuntu-xenial or ubuntu-bionic for the name of the mirror | 21:35 |
| corvus | fungi: if any of the suites for an entry match, then use that list entry? | 21:35 |
| fungi | like, as a substring match? | 21:36 |
| corvus | fungi: yeah | 21:36 |
| corvus | does that work? | 21:36 |
| fungi | it's possible that for something like the nodesource ubuntu package mirror they end up differentiating by base url to the package repository and using a common suite name, though probably for all real-world cases it would likely work | 21:38 |
| fungi | i think if there are common things to mirror which are apt package repos but don't follow that scheme, we could just have a different mirror scheme | 21:38 |
| fungi | i'll try respinning with suite as a list | 21:39 |
| corvus | cool | 21:39 |
| openstackgerrit | Jeremy Stanley proposed zuul/zuul-jobs master: Update documentation for Debian/Ubuntu mirror_info https://review.opendev.org/675703 | 21:43 |
| fungi | there ya go | 21:43 |
| fungi | of course doing it that way, an entry in the mirror no longer corresponds to a single line in sources.list but rather to multiple lines via matrix expansion on the suites list | 21:44 |
| fungi | i suppose type could be changed to types as well so a single entry could expand to both deb and deb-src lines (though odds of people actually making use of deb-src repos in ci jobs is slim unless they're rebuilding debian packages from debian package sources) | 21:45 |
| fungi | may make more sense to have type get a default value of deb so it doesn't have to be specified unless deb-src is really desired | 21:46 |
| *** panda has quit IRC | 22:30 | |
| *** panda has joined #zuul | 22:32 | |
| daniel2 | So it appears that qemu isn't included in the nodepool builder docker image | 22:41 |
| daniel2 | 2019-08-09 22:41:58.568 | + diskimage_builder/lib/disk-image-create:main:262 : echo 'qcow2 output format specified but qemu-img executable not found.' | 22:42 |
| clarkb | daniel2: I think https://opendev.org/zuul/nodepool/src/branch/master/bindep.txt needs to be modified to getthose deps | 22:56 |
| daniel2 | But how does this help in docker? | 22:57 |
| daniel2 | I would guess this would be built into the docker image itself. | 22:58 |
| clarkb | the dockerfile installs deps using that file's listing | 22:58 |
| daniel2 | https://opendev.org/zuul/nodepool/src/branch/master/Dockerfile This is all I found | 23:01 |
| daniel2 | this seems to have both though | 23:02 |
| daniel2 | is there a seperate one for builder? | 23:02 |
| clarkb | that one file buildsall the images aiui | 23:02 |
| pabelanger | we should have nodepool-builder docker some how parent to diskimage-builder docker image, otherwise we need to duplicate dependencies | 23:07 |
| pabelanger | or have nodepool-builder use bindep from dib to install them | 23:08 |
| clarkb | does dib have a dockerfile? | 23:09 |
| pabelanger | not sure, we might need to build it | 23:09 |
| *** tosky has quit IRC | 23:32 | |
| *** weshay is now known as weshay_pto | 23:37 | |
| *** mattw4 has quit IRC | 23:46 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!