| *** soniya is now known as soniya29 | 04:00 | |
| *** ysandeep|out is now known as ysandeep | 05:12 | |
| *** bhagyashris_ is now known as bhagyashris | 05:52 | |
| *** jpena|off is now known as jpena | 07:02 | |
| *** ysandeep is now known as ysandeep|lunch | 09:07 | |
| *** rlandy|out is now known as rlandy | 10:24 | |
| *** ysandeep|lunch is now known as ysandeep | 10:31 | |
| stephenfin | fungi: clarkb: I'm seeing failures to find pbr on a couple of older Python 3.5-based jobs. Have you folks seen this? https://zuul.opendev.org/t/openstack/build/4ea2ca9c30a64be8baa7b4eb83b2123f | 10:37 |
|---|---|---|
| frickler | stephenfin: iirc there was some issue with pypi around that time, can you recheck and check if it works now? | 10:48 |
| stephenfin | frickler: ack, will do (y) | 10:49 |
| opendevreview | Merged openstack/project-config master: Add ansible-collection-kolla to projects.yaml https://review.opendev.org/c/openstack/project-config/+/836763 | 10:57 |
| *** dviroel|afk is now known as dviroel\ | 11:13 | |
| *** dviroel\ is now known as dviroel | 11:13 | |
| *** ysandeep is now known as ysandeep|afk | 12:17 | |
| *** timburke_ is now known as timburke | 12:57 | |
| *** ysandeep|afk is now known as ysandeep | 13:06 | |
| *** dmellado_ is now known as dmellado | 13:31 | |
| fzzf[m] | fungiclarkb : hi,when building the diskimage have follow error.... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/LEyUdYFFPEyQXnpqdBrnULWi) | 14:08 |
| fzzf[m] | * fungiclarkb : hi,when building the diskimage have follow error.... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/DXnZxKQeEjDFziPtaRyCTELU) | 14:11 |
| *** dviroel is now known as dviroel|ruck | 14:14 | |
| clarkb | fzzf[m]: https://opendev.org/openstack/project-config/src/branch/master/nodepool/elements/bindep-fallback.txt the file is there. That repo is expected to be cached on the images we build so the image build can copy the file out of the repo cache. I think you must not be using the repo cache. YOu can eitherdrop the bindep fallback or cache the project-config repo | 14:47 |
| fungi | that fallback is also ancient and we should probably think about trying to ween projects off of it | 14:52 |
| fungi | it's a relic of the "legacy" zuul v2 era jobw | 14:52 |
| fungi | jobs | 14:52 |
| fzzf[m] | <clarkb> "fzzf: https://opendev.org/..." <- do you mean add project-config project on DIB_CUSTOM_PROJECTS_LIST_URL's file. else What is the specific operation about drop the bindep fallback? | 15:11 |
| clarkb | fzzf[m]: we use this element https://opendev.org/openstack/project-config/src/branch/master/nodepool/elements/openstack-repos | 15:14 |
| clarkb | if yo uare using a custom list of projects already then yes you need to update that list | 15:14 |
| fzzf[m] | <clarkb> "fzzf: we use this element https:..." <- yes,I had use DIB_CUSTOM_PROJECTS_LIST_URL: 'file:///etc/nodepool/repos.yaml', I'll add project-config to file. I'm curious how to drop the bindep fallback | 15:19 |
| clarkb | fzzf[m]: if you want to drop it then you'll need to modify the element that includse the file to stop including it | 15:19 |
| clarkb | gmann: fungi: https://storyboard.openstack.org/#!/story/2009973 not sure what the best option is here. Do any supported openstack projcets use devstack-gate anymore? | 15:20 |
| clarkb | there is a zuul role to run a devstack for testing. I might suggest that to them. Or mayb ewe need to fix d-g anyway for our CI? | 15:20 |
| fungi | probably a question for #openstack-qa, i'd be happy to finally retire the repo | 15:20 |
| clarkb | the d-g repo does have a warning in the readme saying it is deprecate and don't use it | 15:21 |
| fzzf[m] | clarkb: okay, I get it. thanks :D | 15:21 |
| fungi | clarkb: looks like frickler updated d-g 6mos ago for another dependency issue | 15:23 |
| clarkb | looks like d-g is expected to be used through wallaby | 15:23 |
| fungi | there were a dozen patches committed to d-g last year | 15:25 |
| fungi | and yeah, if memory serves, grenade lagged behind normal devstack switching over to the new ansible roles by several cycles | 15:26 |
| frickler | oh my, why is d-g on storyboard? but also IMHO nobody really should be using it any longer | 15:26 |
| fungi | maybe once stable/wallaby is the oldest maintained branch we can retire it completely | 15:26 |
| fungi | since it will no longer need grenade jobs | 15:26 |
| frickler | iiuc victoria will go into em next week? | 15:27 |
| fungi | yeah, but i wouldn't be surprised if projects are still running devstack/grenade jobs on em branches | 15:28 |
| *** dviroel|ruck is now known as dviroel|ruck|lunch | 15:28 | |
| fungi | this might be something to talk to qa and tc folks about (removal of d-g support for em) | 15:28 |
| frickler | but they could drop them as em no longer assumes full test coverage iiuc | 15:28 |
| fungi | exactly | 15:28 |
| fungi | just something we'd need to communicate | 15:28 |
| frickler | I mean, if people want to maintain and fix d-g instead, I won't object, but I don't expect much active contributions to happen there | 15:29 |
| gmann | +1 on removing the support for EM but we need to keep d-g until stable/xena is EM. | 15:33 |
| *** ysandeep is now known as ysandeep|out | 15:34 | |
| gmann | initially we thought of having stable/xena beinf EOL but considering EM is also good option | 15:36 |
| clarkb | I'm working on a naive patch that I'll push up and if it works it works | 15:39 |
| clarkb | I'll also leave a note on the story that they should use something other than d-g | 15:40 |
| clarkb | per the warning in the readme | 15:40 |
| clarkb | oh I see if wallaby has problems it is expected to switch. If xena has problems we fix it | 15:42 |
| gmann | yeah | 15:42 |
| clarkb | part of the problem here is this is the CI system that huawei set up for integration testing between openstack and other communities and now they don't really support the ci system either? | 15:42 |
| fungi | gmann: why until stable/xena is em if the last thing using it was grenade? or are we doing grenade for em branches? | 15:43 |
| clarkb | fungi: it may be there are non grenade jobs there too? | 15:43 |
| gmann | fungi: not for greande, but we depredated it officially with warning in stable/xena? | 15:43 |
| fungi | well, there shouldn't have been though | 15:43 |
| clarkb | I wonder about that given the gap between xena and wallaby | 15:43 |
| gmann | yeah, non grenade too | 15:43 |
| gmann | fungi: yeah but 3rd party CI | 15:43 |
| fungi | we warned people to stop using it, if they didn't stop then that's on them | 15:43 |
| opendevreview | Clark Boylan proposed openstack/devstack-gate master: Pin Ara install deps https://review.opendev.org/c/openstack/devstack-gate/+/836991 | 15:44 |
| fungi | but yes, i'm fine saying we're no longer maintaining it and handing the core group over to whoever volunteers | 15:44 |
| gmann | fungi: but we said we will support util syable/wallaby | 15:44 |
| gmann | stable/wallaby | 15:44 |
| gmann | so yes stable/wallaby to EM or EOL then we can remove | 15:44 |
| fungi | so then why do we need to wait for stable/xena to be em? | 15:45 |
| gmann | sorry, stable/wallaby to be EM | 15:45 |
| gmann | EM or EOL | 15:45 |
| clarkb | ah ok | 15:45 |
| gmann | do you want me to add this topic in TC PTG 'testing strategy' section ? | 15:45 |
| gmann | if EM is ok and we do not need to wait for EOL? | 15:45 |
| fungi | probably worth at least bringing up. we've basically been asking projects to stop using devstack-gate for several years already so it would be nice to no longer need to maintain it | 15:46 |
| gmann | so in stable/xena we deprecated saying we accept fix for stable/wallaby only so yes stable/wallaby gone and we can remove | 15:46 |
| opendevreview | Clark Boylan proposed openstack/devstack-gate master: Pin Ara install deps https://review.opendev.org/c/openstack/devstack-gate/+/836991 | 15:46 |
| gmann | and their 3rd party CI too | 15:46 |
| clarkb | fungi: agreed. not ethis is a third party individual out of that thing that mrhillsman set up at huawei | 15:46 |
| fungi | openlab | 15:47 |
| clarkb | I expect that they hav ebeen somewhat abandoned by the people helping them so we may also want to see if we can help them migrat eor whatever | 15:47 |
| clarkb | ya that was it | 15:47 |
| clarkb | I'll write up a response to the issue indicating I've done a naive fix attempt and that it may or may not get merged and that users really need to use something else like the zuul-jobs ensure-devstack role | 15:48 |
| gmann | clarkb: fungi frickler added in https://etherpad.opendev.org/p/tc-zed-ptg#L283 | 15:49 |
| clarkb | thanks | 15:49 |
| clarkb | ok message posted to the issue | 15:52 |
| frickler | I think reworking those test jobs they are using may be the best solution. they need to test against recent openstack anyway | 15:56 |
| frickler | I'll see whether some of my colleagues can help with that | 15:57 |
| clarkb | ++ I left a note about the ensure-devstack role as a possibility | 15:58 |
| *** jpena is now known as jpena|off | 16:05 | |
| iurygregory | does anyone know if there are any plans in Zed to have CI running in ubuntu 22.04? =) | 16:08 |
| clarkb | iurygregory: I'm not even sure if anyone hsa started working on test nodes for 22.04. That said help with that and setting up mirrors etc would be grately appreciated | 16:09 |
| clarkb | oh wait frickler started looking at it that was what the debootstrap stuff was about | 16:10 |
| clarkb | so ya some movement | 16:10 |
| frickler | clarkb: iurygregory: I started testing dib with it .. right | 16:10 |
| iurygregory | clarkb, ack | 16:10 |
| frickler | https://review.opendev.org/c/openstack/diskimage-builder/+/836228 | 16:10 |
| iurygregory | frickler, gotcha, but do you think people will try to move the whole CI during zed? | 16:11 |
| clarkb | frickler: sorry too many thing sthis week in all different directions. Thank you for picking that up | 16:11 |
| frickler | certainly not the whole CI, but getting devstack running on it and projects adding n-v jobs for that would seem feasible | 16:11 |
| clarkb | iurygregory: note that the TC generally sets that direction | 16:11 |
| iurygregory | clarkb, right =) | 16:12 |
| clarkb | via the supported distro release stuff which I think tends to lag a cycle behind ubuntu | 16:12 |
| clarkb | so jammy releases in 2 weeks, during zed we spin up the tooling and infrastructure, then in aardvark switch over. But that is based on history not specific discussion I'v eheard this time around | 16:12 |
| *** dviroel|ruck|lunch is now known as dviroel|ruck | 16:13 | |
| iurygregory | tks! | 16:14 |
| fungi | iurygregory: the policy in openstack so far is that the test platform for the cycle is whatever LTS version was available at the start of the cycle (so prior cycle rc deadline or release day-ish) | 16:21 |
| fungi | if 22.04 has been officially ga in march then it might have been considered | 16:22 |
| iurygregory | fungi, gotcha! | 16:23 |
| fungi | ("available" officially, not in prerelease/frozen) | 16:23 |
| frickler | well it is called 22.04 for a reason ;) | 16:43 |
| fungi | yes, that was my point ;) | 16:44 |
| clarkb | gmann: frickler fungi https://review.opendev.org/c/openstack/devstack-gate/+/836991 passes CI and does seem to run d-g in CI: https://zuul.opendev.org/t/openstack/build/54729bf3fa7c42e281eb45a667389831/log/job-output.txt#1560 | 18:23 |
| clarkb | we might want to approve that to avoid any issues with old branches? Though not sure if we want to check if old branche sare affected first | 18:24 |
| gmann | clarkb: thanks, at least legacy-tempest-neutron-full-stable which is stable/wallaby is passing on it. | 18:26 |
| gmann | clarkb: fungi frickler I am +2 as stable/wallaby is working and if it create issue on older stable let people report that so that we will get to know who is using it. | 18:29 |
| opendevreview | Clark Boylan proposed openstack/devstack-gate master: DO NOT MERGE testing if d-g is happy https://review.opendev.org/c/openstack/devstack-gate/+/837008 | 18:36 |
| clarkb | decided to gather that baseline data really quick | 18:36 |
| *** dasm|off is now known as dasm | 18:42 | |
| clarkb | it looks like things actually work in d-g in our ci | 19:05 |
| clarkb | I think the reason is that we must have a newer pip which is dep solving a valid set for us | 19:05 |
| clarkb | anyway I don't think the pins will hurt anything the ara stuff is a small thing | 19:08 |
| frickler | looking at the log they seem to be running d-g on focal, which is even less supported https://logs.openlabtesting.org/logs/64/1364/0d5a60e53d0f24febc83fa5a7e88f0824314ca37/check/terraform-provider-openstack-acceptance-test-foo/fc1d504/job-output.txt.gz | 19:12 |
| clarkb | oh intersting | 19:13 |
| frickler | likely some setup being done a couple of years ago and never touched afterwards, except the default node type getting updated | 19:14 |
| frickler | hmm, no, was explicitly set to use focal https://github.com/terraform-provider-openstack/terraform-provider-openstack/blame/main/.zuul.yaml#L21 | 19:19 |
| frickler | anyways, I'll have a go trying to update that config, maybe tomorrow or at least soonish | 19:19 |
| *** rlandy is now known as rlandy|mtg | 19:53 | |
| *** dviroel|ruck is now known as dviroel|afk | 20:39 | |
| opendevreview | Merged openstack/devstack-gate master: Pin Ara install deps https://review.opendev.org/c/openstack/devstack-gate/+/836991 | 20:44 |
| *** rlandy|mtg is now known as rlandy | 21:17 | |
| *** dasm is now known as dasm|off | 22:44 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!