| opendevreview | Merged openstack/devstack master: Fix rdo_release for unmaintained branches https://review.opendev.org/c/openstack/devstack/+/922473 | 00:18 |
|---|---|---|
| opendevreview | Merged openstack/devstack master: add ubuntu noble (24.04) support https://review.opendev.org/c/openstack/devstack/+/922427 | 00:19 |
| opendevreview | Merged openstack/tempest master: Add image dependencies test for image <-> instance snapshot https://review.opendev.org/c/openstack/tempest/+/914198 | 00:30 |
| opendevreview | yatin proposed openstack/devstack stable/2024.1: Fix rdo_release for unmaintained branches https://review.opendev.org/c/openstack/devstack/+/922572 | 04:30 |
| opendevreview | yatin proposed openstack/devstack stable/2023.2: Fix rdo_release for unmaintained branches https://review.opendev.org/c/openstack/devstack/+/922573 | 04:31 |
| opendevreview | yatin proposed openstack/devstack stable/2023.1: Fix rdo_release for unmaintained branches https://review.opendev.org/c/openstack/devstack/+/922574 | 04:35 |
| whoami-rajat | hey #openstack-qa , can i have reviews on the following devstack patches? | 05:20 |
| whoami-rajat | 1. https://review.opendev.org/c/openstack/devstack/+/909841/3 | 05:20 |
| whoami-rajat | 2. https://review.opendev.org/c/openstack/devstack/+/910473/4 | 05:20 |
| opendevreview | Amit Uniyal proposed openstack/whitebox-tempest-plugin master: Rebuild with uefi image property https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/922577 | 07:11 |
| ykarel | kopecmartin, gmann please also have a look at the backports https://review.opendev.org/q/I967de13094ff6df46737a22d4e1758f9900dfbc9 | 10:16 |
| ykarel | thx in advance | 10:22 |
| opendevreview | Merged openstack/devstack master: stackrc: Remove USE_PYTHON3 https://review.opendev.org/c/openstack/devstack/+/920658 | 10:52 |
| opendevreview | Rodolfo Alonso proposed openstack/devstack master: [Neutron] Add a new Neutron service: neutron-ovn-maintenance-worker https://review.opendev.org/c/openstack/devstack/+/922611 | 11:11 |
| sean-k-mooney | kopecmartin: gmann tempest-full timeed out on https://review.opendev.org/c/openstack/tempest/+/922108 | 13:00 |
| sean-k-mooney | im debating whtat to do | 13:00 |
| sean-k-mooney | either i can extend the timout on tempest-full-py3 or the parent devstack-tempest | 13:00 |
| sean-k-mooney | or i can move the optimisaton in the follwo up patch to devstack-tempest or also apply the to tempest-full to move this forwoard | 13:01 |
| sean-k-mooney | kopecmartin: gmann do ye have a perference in how i move this forward | 13:02 |
| sean-k-mooney | do we want to be concervitive and just bump the timeouts or be eger and apply the optimistaions to a wider set of jobs from the get go | 13:03 |
| *** haleyb is now known as haleyb|out | 13:16 | |
| kopecmartin | sean-k-mooney: we saw that the optimization speeded up the job deployment, i think we can enable it in more jobs, i'd find it more logical than just increasing timeout (when we have a way how to decrease the deployment time) | 13:20 |
| sean-k-mooney | the timeout i was expecting to have merged a few weeks ago to unblock the nova gate | 13:22 |
| sean-k-mooney | form my perspective we have basically been blocked from merging anything for the beter part of a month because of the timeouts we were seeing | 13:22 |
| sean-k-mooney | but ya that was never intended as a fix | 13:22 |
| sean-k-mooney | what ill do is ill propose a new patch to apply the optimisation to devstack-tempest directly and we can see how that goes | 13:23 |
| kopecmartin | what if we force-pushed the patch(es)? frickler? i mean, if we have a (partial) fix for a timeout that we have a problem to merge because of timeouts, we're in a virtuous circle | 13:26 |
| sean-k-mooney | the main thing i want to avoid is wasting more ci cycles tryign to recheck gridend this in. | 13:32 |
| sean-k-mooney | i belive devstack-tempest is defiend in the devstack repo so ill push a seperate patch to applie the optimisation directly to that shortly | 13:34 |
| opendevreview | sean mooney proposed openstack/devstack master: enable openstack-cli-server and other perfromace tunings https://review.opendev.org/c/openstack/devstack/+/922630 | 14:32 |
| opendevreview | sean mooney proposed openstack/devstack master: enable openstack-cli-server and other perfromace tunings https://review.opendev.org/c/openstack/devstack/+/922630 | 14:33 |
| sean-k-mooney | kopecmartin: ok ^ does basicaly the same thing for all devstack jobs so lets see if the devstack gate is happy and we can consider how to proceed. | 14:34 |
| frickler | hmm, using 0.6.1 for that rebuild test may explain why it is failing so often. maybe using a second copy of 0.6.2 would be a better idea | 14:39 |
| frickler | kopecmartin: sure we can also force-merge things, I wasn't aware of the situation being that bad. didn't sound like that to me in last week's tc meeting | 14:40 |
| sean-k-mooney | a second copy of 0 | 14:42 |
| sean-k-mooney | 0.6.2 | 14:42 |
| sean-k-mooney | shoudl be fine imo | 14:43 |
| sean-k-mooney | im not sure why we use 0.6.1 | 14:43 |
| sean-k-mooney | as the alt image we jsut need 2 glance images | 14:43 |
| sean-k-mooney | but the content could be the same | 14:43 |
| sean-k-mooney | perhaps use a wholedisk and uec image | 14:44 |
| sean-k-mooney | of 0.6.2 | 14:44 |
| opendevreview | Merged openstack/devstack stable/2024.1: Fix rdo_release for unmaintained branches https://review.opendev.org/c/openstack/devstack/+/922572 | 15:10 |
| opendevreview | Merged openstack/devstack stable/2023.2: Fix rdo_release for unmaintained branches https://review.opendev.org/c/openstack/devstack/+/922573 | 15:10 |
| opendevreview | Merged openstack/devstack stable/2023.1: Fix rdo_release for unmaintained branches https://review.opendev.org/c/openstack/devstack/+/922574 | 15:39 |
| sean-k-mooney | frickler: just read your comment im not sure whtat the policy would be for dropping bulseye and oepnEuler22.03 i would have to check the testing runtims i guess | 17:33 |
| sean-k-mooney | is that peopel pefernece or should i just disable the cache turning on those jobs | 17:33 |
| frickler | sean-k-mooney: we talked about it some weeks ago and kopecmartin said something like "leave it running while it is working", but maybe wait for feedback from him or gmann | 17:35 |
| sean-k-mooney | sure ill respond on gerrit but im fine with either approch | 17:37 |
| clarkb | sort of related: I've been working on centos 8 stream cleanup today and have realized that a number of projcets are still running fips jobs based on centos 8 stream that are failing and also mean I can't remove the base jobs | 17:38 |
| gmann | sean-k-mooney: frickler: I am ok to try it. tempest-full-py3 working but it timeout many time and many time it was 'recheck'ed which hide the issues. especially I am interested to enable ZSWAP in tempest-full-py3 | 17:38 |
| clarkb | do we know if there has been any effort to move fips to 9 stream or something else? | 17:38 |
| gmann | clarkb: last time I checked on this, there was some talk to try fips testing on rocky but I have not seen any effort on that | 17:39 |
| sean-k-mooney | clarkb: i tought all the fips jobs where ment to be on c9s now | 17:39 |
| sean-k-mooney | i have not really been inovled in that however | 17:39 |
| clarkb | sean-k-mooney: there may be c9s jobs too, but there is still a lot of c8s if so | 17:41 |
| gmann | I think Ade lee was waiting for Ubuntu Jammy to have it enable and we talked that in parallel it is worth to try on Rocky also | 17:41 |
| clarkb | gmann: ya I remember ubuntu was a bit of an adventure as fips support is a followup thing and not available at release | 17:42 |
| gmann | centos jobs are very unstable and I do not see any effort to make it stable | 17:42 |
| clarkb | in any case the c8s based jobs cannot function anymore so we need to reove or convert them | 17:42 |
| sean-k-mooney | gmann: ya so one of the issues with testing in general is that upstream centos dont offically supprot fips in that they do not conisder it part fo there testing cirtia or have it in there automated ci | 17:42 |
| gmann | yeah, we do not have alternate except someone try it on Rocky and see | 17:42 |
| sean-k-mooney | i personally would remove the c8s jobs | 17:43 |
| sean-k-mooney | and either do best effort testign with c9s or jammy when that is aviable | 17:43 |
| gmann | sean-k-mooney: but centos jobs iteself are not stable and that is why we moved them as advance/unstable in testing runtime | 17:43 |
| gmann | ++, I am onboard on removing plan | 17:44 |
| clarkb | sean-k-mooney: ya to be clear I'm going to remove the test nodes and the job fail 100% of the time already today | 17:44 |
| sean-k-mooney | ya my point was i dont think c8s coverage is really giving us much considerign it requires a non defualt python verison to be used amoung other things | 17:44 |
| clarkb | so the questino is mostly around is there an alternative I can point people to when we break them harder than its already broken today (and sounds like no unfortunately) | 17:44 |
| sean-k-mooney | clarkb: all the repos are going to go away if they have not already right | 17:44 |
| sean-k-mooney | for c8s | 17:45 |
| clarkb | they already went away which is why 100% of jobs running on c8s fail today | 17:45 |
| clarkb | upstream deleted all the pacakges and then our mirrors replicated that | 17:45 |
| sean-k-mooney | ack | 17:53 |
| sean-k-mooney | am ill check novas repos and see if we have any refernce to c8s | 17:53 |
| sean-k-mooney | if so we can proactivly delete them if you dont have paches up already | 17:53 |
| sean-k-mooney | we had a policy of not runing centos in check/geat for nova but there might be an experimntal job | 17:54 |
| sean-k-mooney | looks like we move 2 years ago to c9s https://github.com/openstack/nova/commit/9a7028f02cbdfe034b98f524c1bc8c497f14e423 | 17:55 |
| clarkb | I haven't pushed changes to projects. Just prepping stuff in the infrastructure side under topic:drop-centos-8-stream | 17:57 |
| clarkb | oh ok there is a c9s fips job so I guess the short term answer is to update to c9s for fips | 17:57 |
| clarkb | https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/922649 you can see here that glance for example uses a c8s fips job | 17:58 |
| clarkb | (it is the source of the -1 from zuul on the change and that is glance master not an old stable branch) | 17:58 |
| sean-k-mooney | ya short term solution is move fips to c9s anywhere it currenly c8s | 18:04 |
| sean-k-mooney | i have no idea how stable that actlly is because we never run it | 18:05 |
| sean-k-mooney | i think neutron uses c9s more oftent then we do | 18:05 |
| sean-k-mooney | looks like they do also for fips https://github.com/openstack/neutron/blob/master/zuul.d/base.yaml#L163-L174 and https://github.com/openstack/neutron/blob/master/zuul.d/base.yaml#L248 | 18:07 |
| opendevreview | Ashish Gupta proposed openstack/tempest master: [WIP]Add test to verify hostname allows FQDN https://review.opendev.org/c/openstack/tempest/+/922342 | 18:20 |
| opendevreview | Carlos Eduardo proposed openstack/devstack-plugin-ceph master: [manila] Enforce cephfsnative as the default driver https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/922678 | 21:51 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!