| opendevreview | Takashi Kajinami proposed openstack/tempest master: Remove deprecated alias of test decorators (1/2) https://review.opendev.org/c/openstack/tempest/+/908552 | 02:47 |
|---|---|---|
| opendevreview | Takashi Kajinami proposed openstack/tempest master: Remove deprecated alias of test decorators (2/2) https://review.opendev.org/c/openstack/tempest/+/914565 | 02:47 |
| opendevreview | Takashi Kajinami proposed openstack/tempest master: Replace deprecated os_adm https://review.opendev.org/c/openstack/tempest/+/914568 | 02:59 |
| opendevreview | Takashi Kajinami proposed openstack/tempest-stress master: Get rid of six https://review.opendev.org/c/openstack/tempest-stress/+/914579 | 06:38 |
| opendevreview | Takashi Kajinami proposed openstack/tempest-stress master: Split requirements for documentation build https://review.opendev.org/c/openstack/tempest-stress/+/914580 | 06:41 |
| opendevreview | Takashi Kajinami proposed openstack/tempest-stress master: Split requirements for documentation build https://review.opendev.org/c/openstack/tempest-stress/+/914580 | 06:43 |
| opendevreview | Maxim Sava proposed openstack/tempest master: Add image dependency test coverage https://review.opendev.org/c/openstack/tempest/+/914612 | 08:28 |
| opendevreview | Maxim Sava proposed openstack/tempest master: Add image dependency test coverage https://review.opendev.org/c/openstack/tempest/+/914198 | 08:52 |
| kopecmartin | gmann: PTG slot, Monday and Wednesday, both 1 hour long , 14-15 UTC, wdyt? | 10:48 |
| opendevreview | Maxim Sava proposed openstack/tempest master: Add image dependency test coverage https://review.opendev.org/c/openstack/tempest/+/914198 | 11:53 |
| rosmaita | has anyone noticed that grenade-skip-level has never passed on stable/2024.1 ? https://zuul.opendev.org/t/openstack/builds?job_name=grenade-skip-level&branch=stable%2F2024.1&skip=0 | 13:20 |
| rosmaita | error is /opt/stack/old/neutron/devstack/plugin.sh: line 23: /opt/stack/old/devstack/lib/neutron_plugins/ovs_source: No such file or directory | 13:20 |
| frickler | rosmaita: yes I noticed that on https://review.opendev.org/c/openstack/grenade/+/914114 but didn't look closer yet | 13:29 |
| rosmaita | frickler: i can't figure it out, we were able to merge some cinder patches to stable/2024.1 but those didn't run grenade-skip-level | 13:30 |
| rosmaita | example: https://review.opendev.org/c/openstack/cinder/+/913816 | 13:30 |
| rosmaita | not sure what accounts for that | 13:31 |
| frickler | that was before grenade was branched, which only happened this Monday I think. before that the job didn't exist yet | 13:32 |
| rosmaita | dammit | 13:38 |
| opendevreview | Maxim Sava proposed openstack/tempest master: Add image dependency test coverage https://review.opendev.org/c/openstack/tempest/+/914198 | 14:57 |
| clarkb | the old branch in this case should be stable/2023.1 right? since we're doing a 2023.1 to 2024.1 skip level upgrade | 15:16 |
| clarkb | https://opendev.org/openstack/devstack/src/branch/stable/2023.1/lib/neutron_plugins/ovs_source does exist so thats weird | 15:17 |
| clarkb | I think the job may be misconfigured it appears to be using wallaby on the old side? https://zuul.opendev.org/t/openstack/build/2b3d02550958455fb2f2099eb1e1cde5/log/job-output.txt#1605 | 15:22 |
| frickler | https://zuul.opendev.org/t/openstack/build/2b3d02550958455fb2f2099eb1e1cde5/log/zuul-info/inventory.yaml#287 | 15:25 |
| frickler | the inheritance below also looks weird | 15:25 |
| frickler | seems those matchers also might need to get updated after every release | 15:26 |
| clarkb | or just remove them.... | 15:26 |
| clarkb | I'm not sure why we can't manage the branch specific config on the branch directly. | 15:26 |
| frickler | I was guessing it might be something with forward testing, but no idea really | 15:27 |
| clarkb | https://opendev.org/openstack/grenade/src/branch/stable/zed/.zuul.yaml#L381-L407 | 15:27 |
| clarkb | but ya that chooses wallaby which is also wrong for zed... | 15:28 |
| clarkb | reading gmann's comment there I'm not sure that was necessary | 15:28 |
| clarkb | on a SLURP release you add the skip level job. On non slurp you remove it | 15:29 |
| clarkb | then let zuul automatically associate branches normally without overrides and that is all you should need to do to ensure the job runs on the correct changes and with the correct branches | 15:29 |
| frickler | I think that that was branched from then-master and ignored afterwards instead of getting properly updated | 15:29 |
| clarkb | basically that comment says it was only wanted for yoga. So on yoga (or master that becomes yoga later at the time) you add the job definition I linked and put the job in the list of jobs to run. Thats all you do | 15:31 |
| clarkb | none of the other branches would run the job ebcause they don't list the job to run | 15:31 |
| clarkb | then when yoga is branched and master/zed are a thing you can decide if you want to remove the skip level job and that doesn't require modifications to the job definitions only where you indicate the job should run | 15:32 |
| clarkb | then when 2023.1 branches you update the job definition on 2023.1/master to set the old branch to yoga. Add possibly add the job to your list of jobs to run | 15:33 |
| frickler | just speculating, maybe this was mixed up because of the explicit matchers needed in tempest and o-z-j templates | 15:34 |
| opendevreview | Ashley Rodriguez proposed openstack/devstack-plugin-ceph stable/2023.1: Fix stable CI https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/909055 | 16:43 |
| gmann | rosmaita: clarkb: frickler: it came from stable/zed branch. let me cleanup this job to be specific branch to run instead of negative regex where not to run which we need to update every time. | 16:59 |
| gmann | will be pushing the changes soon after my breakfast | 16:59 |
| rosmaita | gmann: sounds good, thanks! | 17:16 |
| opendevreview | Ghanshyam proposed openstack/grenade master: Remove grenade-skip-level job from non SLURP master https://review.opendev.org/c/openstack/grenade/+/914685 | 17:33 |
| opendevreview | Ghanshyam proposed openstack/grenade stable/2023.1: Fix grenade-skip-level job to run on respective stable branch https://review.opendev.org/c/openstack/grenade/+/914687 | 17:45 |
| opendevreview | Ghanshyam proposed openstack/grenade stable/zed: Remove grenade-skip-level job from stable/zed https://review.opendev.org/c/openstack/grenade/+/914690 | 17:54 |
| opendevreview | Ghanshyam proposed openstack/grenade stable/2024.1: Update TOX_CONSTRAINTS_FILE for stable/2024.1 https://review.opendev.org/c/openstack/grenade/+/914114 | 17:57 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Update integrated template for grenade-skip-level job https://review.opendev.org/c/openstack/tempest/+/914695 | 18:19 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Update integrated template for grenade-skip-level job https://review.opendev.org/c/openstack/tempest/+/914695 | 18:19 |
| gmann | rosmaita: clarkb : frickler: kopecmartin: these should fix the grenade-skip-level job https://review.opendev.org/q/topic:%22qa-2024-1-release%22+label:+Review-Priority%3E1 | 18:20 |
| gmann | please NOTE, setup for stable/2024.1 and master in QA is not yet completed so there might be a few more things you will notice where jobs are not testing correct upgrade or branch. | 18:20 |
| rosmaita | gmann: ack, and thanks for the patches | 18:21 |
| clarkb | gmann: I -1'd one of those changes because it keeps the branches: list and I'm like 99% positive that we can just remove it | 18:37 |
| clarkb | but please let me know if there is a specific reason for keeping it | 18:38 |
| gmann | checking | 18:38 |
| gmann | clarkb: reason I added branch variant is if anyone using grenade-skip-level in the job template and in NON SLURP master release we are removing the job definition then it will find the job from those stable branch and start testing wrongly | 18:43 |
| gmann | that is what is happening in current case | 18:43 |
| gmann | because this job is little special that it need to run on SLURP and not on NON SLURP, I wanted to make each job definition on stable branch explicit because they all has different setting in their own version of job | 18:44 |
| clarkb | then you shouldn't define the job in grenade | 18:45 |
| gmann | its grenade common job | 18:45 |
| clarkb | because every time you go to create a new branch you're just going to recreate this problem | 18:45 |
| clarkb | and in 2 years someon will have to debug it again | 18:45 |
| gmann | with current approach it will happen. I am going with approach 1. no job definition on NON SLURP release 2. SLURP release will have their setting and job restricted to run on their gate only | 18:46 |
| clarkb | thats fine, as long as you delete the branches: specifier for the job on all branches | 18:46 |
| clarkb | and then rely on implicit branch matching and the job not being defined for branches if you dont' want it to run on that branch | 18:47 |
| gmann | then this problem will happen. Currently this problem is happening because we have negative regex in branches and we have to keep them updated in every release | 18:47 |
| clarkb | ya the positive match is better. I just want to avoid branches: entirely because we have consistent issues like this over time | 18:48 |
| clarkb | and zuul really isn't intended to be used that way when you have branched configs | 18:48 |
| gmann | clarkb: if I remove the branch variant in stable/2023.1 and remove job definition in master then if anyone run job on master, will not zuul find it from stable/2023.1? | 18:48 |
| clarkb | I think the job won't be found at all and it will be skipped | 18:49 |
| gmann | That is happening in stable/2024.1 case right - https://review.opendev.org/c/openstack/grenade/+/914114 | 18:49 |
| clarkb | this is what happened to some repo after the 2024.1 branching recently | 18:49 |
| gmann | clarkb: ^^ but it is picking the job from stable/zed | 18:49 |
| clarkb | gmann: yes because your branches specifier explicitly tells it to | 18:50 |
| clarkb | because stable/2024.1 matches the branches specifier | 18:50 |
| clarkb | if you delete the branches specifier then there is no explicit matching that can override the implicit behavior | 18:50 |
| gmann | clarkb: because it has negative regex means it is for stable.2024.1 ? | 18:50 |
| gmann | i see, I think i get your point now. I though zuul will look into all branches and pick if no branch variant | 18:51 |
| gmann | *thought | 18:51 |
| clarkb | regex: ^master|stable/(stein|train|ussuri|victoria|wallaby|xena|zed)$\nnegate: true is the rule | 18:51 |
| gmann | yeah | 18:51 |
| clarkb | stable/2024.1 does not match that regex then we negate it which evaluiates to true | 18:51 |
| clarkb | that means apply this job definition to stable/2024.1 | 18:51 |
| clarkb | without a branches specifier only stable/zed will match that definition | 18:52 |
| gmann | so if no branch variant then it will not pick job from stable branch right? because that is consider to be for stable/2024.1 only as no branch variant? | 18:52 |
| gmann | clarkb: cool, I think that was the confusion | 18:52 |
| gmann | let me update the patch | 18:52 |
| clarkb | gmann: the whole idea behind zuul's branched job definition behavior is that the current branch's job definitions applies to changes for that branch implicitly | 18:53 |
| clarkb | when you add an explicit rule you remove that implicit behavior and replace it by whatever you provide | 18:53 |
| clarkb | you should only do this in rare circumstances on branched job configs | 18:53 |
| clarkb | it is more common and useful when you define jobs in unbranched repos (like openstack-zuul-jobs) | 18:54 |
| gmann | clarkb: ok. it will make it more easy now. | 18:54 |
| clarkb | beacuse there is a single branch with all the jobs in it in that case and you may want to apply one version to a specific branch or another | 18:54 |
| clarkb | The main issues arise when you make a branches rule on master today that does what you want. Then some time in the future you create new branches which accidentally match (or not) the job definition for the old job and creates unexpected job behaviors | 18:55 |
| gmann | clarkb: one more question, if no branches rule on master job definition also that means also this job is for master only and not for stable branches ? | 18:58 |
| gmann | or we need to be explicit branches rule as 'master' if we want to keep it for master only | 18:58 |
| clarkb | it should only be for master | 18:59 |
| clarkb | this was the exact thing someone ran into with the 2024.1 branching | 18:59 |
| clarkb | devstack didn't have stable/2024.1 yet so no devstack based jobs would run against stable/2024.1 on that project (I forget which one) | 18:59 |
| clarkb | and it is why I suggested devstack needs to branch first before anything else in the future | 18:59 |
| gmann | yeah, that was kuryr-kubernetes job | 19:00 |
| opendevreview | Ghanshyam proposed openstack/grenade stable/2023.1: Fix grenade-skip-level job to run on respective stable branch https://review.opendev.org/c/openstack/grenade/+/914687 | 19:01 |
| gmann | clarkb: ^^ done | 19:02 |
| gmann | I will make note of these in QA release wiki for future ref | 19:02 |
| opendevreview | Merged openstack/tempest master: Remove deprecated alias of test decorators (1/2) https://review.opendev.org/c/openstack/tempest/+/908552 | 20:33 |
| opendevreview | Martin Kopec proposed openstack/devstack stable/2024.1: Update branches for stable/2024.1 https://review.opendev.org/c/openstack/devstack/+/914717 | 21:53 |
| opendevreview | Martin Kopec proposed openstack/devstack stable/2024.1: Update DEVSTACK_SERIES to 2024.2 https://review.opendev.org/c/openstack/devstack/+/914718 | 21:55 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Add Tempest gate job for stable/2024.1 https://review.opendev.org/c/openstack/tempest/+/914719 | 21:57 |
| opendevreview | Martin Kopec proposed openstack/devstack stable/2024.1: Cap max microversions for stable/2024.1 https://review.opendev.org/c/openstack/devstack/+/914723 | 22:14 |
| opendevreview | Martin Kopec proposed openstack/devstack stable/2024.1: Stop installing Tempest at system wide for stable branch https://review.opendev.org/c/openstack/devstack/+/914724 | 22:17 |
| kopecmartin | gmann: I pushed devstack release patches ^^ | 22:31 |
| kopecmartin | let's sync, what's left? | 22:31 |
| kopecmartin | i saw you grenade patches | 22:31 |
| gmann | kopecmartin: thanks | 22:31 |
| gmann | kopecmartin: i just fixed the grenade skip level job but rest other setting is still needs to be done | 22:32 |
| gmann | checking devstack changes | 22:32 |
| kopecmartin | oh, right, .. ok, i'm on it then | 22:33 |
| gmann | thanks | 22:33 |
| gmann | kopecmartin: this needs to be pushed to master branch, rest all lgtm https://review.opendev.org/c/openstack/devstack/+/914718/1 | 22:36 |
| kopecmartin | ups .. i wasn't paying attention | 22:37 |
| opendevreview | Martin Kopec proposed openstack/devstack master: Update DEVSTACK_SERIES to 2024.2 https://review.opendev.org/c/openstack/devstack/+/914725 | 22:38 |
| opendevreview | Martin Kopec proposed openstack/devstack stable/2024.1: Cap max microversions for stable/2024.1 https://review.opendev.org/c/openstack/devstack/+/914723 | 22:39 |
| opendevreview | Martin Kopec proposed openstack/devstack stable/2024.1: Stop installing Tempest at system wide for stable branch https://review.opendev.org/c/openstack/devstack/+/914724 | 22:39 |
| gmann | thanks re-applied +W | 22:40 |
| opendevreview | Martin Kopec proposed openstack/grenade master: Update the master grenade setting for 2024.1 to master upgrade https://review.opendev.org/c/openstack/grenade/+/914727 | 22:45 |
| opendevreview | Martin Kopec proposed openstack/grenade stable/2024.1: Update grenade setting in stable/2024.1 https://review.opendev.org/c/openstack/grenade/+/914728 | 22:48 |
| kopecmartin | seems all is done .. unless we need to release some of the projects with independent release model | 22:51 |
| kopecmartin | it's been 2 years we released bashate | 22:53 |
| kopecmartin | the same with devstack-tools | 22:55 |
| gmann | kopecmartin: I think no change merged in bachate which need release? | 22:56 |
| gmann | 1 comment in grenade setting https://review.opendev.org/c/openstack/grenade/+/914728/comment/155647b3_8ad1023e/ | 22:56 |
| kopecmartin | yeah, nothing relevant was added in bashate | 22:57 |
| * kopecmartin looking | 22:57 | |
| opendevreview | Martin Kopec proposed openstack/grenade stable/2024.1: Update grenade setting in stable/2024.1 https://review.opendev.org/c/openstack/grenade/+/914728 | 22:58 |
| gmann | I think extension cap on stable/2024.1 devstack is also needed to be done. it might be late for you so I can push that later | 22:58 |
| opendevreview | Ghanshyam proposed openstack/grenade master: Upgrade skiplevel jobs to check the right from_branches https://review.opendev.org/c/openstack/grenade/+/913892 | 23:04 |
| kopecmartin | gmann: ack, thank you .. i always struggle finding out what new extensions were added anyway :D | 23:05 |
| gmann | frickler: can you remove -2 from this https://review.opendev.org/c/openstack/grenade/+/913892 | 23:05 |
| gmann | kopecmartin: this one also though it need frickler to remove his -2 https://review.opendev.org/c/openstack/grenade/+/913892 | 23:05 |
| opendevreview | Merged openstack/tempest master: Update integrated template for grenade-skip-level job https://review.opendev.org/c/openstack/tempest/+/914695 | 23:06 |
| opendevreview | Merged openstack/grenade master: Remove grenade-skip-level job from non SLURP master https://review.opendev.org/c/openstack/grenade/+/914685 | 23:06 |
| opendevreview | Merged openstack/grenade stable/2023.1: Fix grenade-skip-level job to run on respective stable branch https://review.opendev.org/c/openstack/grenade/+/914687 | 23:06 |
| opendevreview | Merged openstack/grenade stable/zed: Remove grenade-skip-level job from stable/zed https://review.opendev.org/c/openstack/grenade/+/914690 | 23:06 |
| opendevreview | Ghanshyam proposed openstack/devstack stable/2024.1: Cap stable/2024.1 network, swift, volume api_extensions for tempest https://review.opendev.org/c/openstack/devstack/+/914732 | 23:33 |
| gmann | kopecmartin: ^^ extension cap. added neutron folks also to double check | 23:34 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Create releasenotes page for tempest 37.0.0, 38.0.0 https://review.opendev.org/c/openstack/tempest/+/914733 | 23:40 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!