| opendevreview | Chris Wedgwood proposed openstack/tempest master: [DNM] Retry object creation on 409 Conflict https://review.opendev.org/c/openstack/tempest/+/807909 | 00:30 |
|---|---|---|
| opendevreview | lkk proposed openstack/tempest master: Fix unused parameters in _create_subnet method https://review.opendev.org/c/openstack/tempest/+/807976 | 00:54 |
| opendevreview | Chris Wedgwood proposed openstack/tempest master: [DNM] Retry object creation on 409 Conflict https://review.opendev.org/c/openstack/tempest/+/807909 | 00:59 |
| opendevreview | Chris Wedgwood proposed openstack/tempest master: [DNM] Retry object creation on 409 Conflict https://review.opendev.org/c/openstack/tempest/+/807909 | 02:03 |
| opendevreview | Chris Wedgwood proposed openstack/tempest master: make tempest-tox-plugin-sanity-check non-voting https://review.opendev.org/c/openstack/tempest/+/807983 | 02:03 |
| opendevreview | lkk proposed openstack/tempest master: Fix unused parameters in _create_subnet method https://review.opendev.org/c/openstack/tempest/+/807976 | 03:17 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Fix tempest plugins sanity jobs by exclude failing plugins https://review.opendev.org/c/openstack/tempest/+/807986 | 03:36 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Fix tempest plugins sanity jobs by exclude failing plugins https://review.opendev.org/c/openstack/tempest/+/807986 | 04:12 |
| gmann | kopecmartin: ^^ fixing tempest gate: https://review.opendev.org/c/openstack/tempest/+/807986 | 04:13 |
| opendevreview | Chris Wedgwood proposed openstack/tempest master: [DNM] Retry object creation on 409 Conflict https://review.opendev.org/c/openstack/tempest/+/807909 | 04:55 |
| opendevreview | OpenStack Proposal Bot proposed openstack/devstack master: Updated from generate-devstack-plugins-list https://review.opendev.org/c/openstack/devstack/+/807989 | 06:10 |
| opendevreview | Merged openstack/tempest master: Fix tempest plugins sanity jobs by exclude failing plugins https://review.opendev.org/c/openstack/tempest/+/807986 | 06:24 |
| frickler | clarkb: from this it looks like dstat isn't working stable, either. maybe we should drop it completely for now? I don't think it is being looked at very often and likely more gate stability would be preferable https://zuul.opendev.org/t/openstack/build/849182ef795c4dd7b654efb80d503fda | 07:10 |
| opendevreview | Radosław Piliszek proposed openstack/devstack master: [WIP] Test running on Debian Bullseye https://review.opendev.org/c/openstack/devstack/+/789083 | 07:10 |
| yoctozepto | clarkb, frickler: yeah, we dropped dstat because it was buggy, not on a whim | 07:13 |
| yoctozepto | I guess we could apply a poor-man's solution which could collect some stats in background just dropping lines into a file? | 07:14 |
| opendevreview | Merged openstack/devstack master: Updated from generate-devstack-plugins-list https://review.opendev.org/c/openstack/devstack/+/807989 | 07:38 |
| *** jpena|off is now known as jpena | 07:41 | |
| opendevreview | Radosław Piliszek proposed openstack/devstack master: [WIP] Test running on Debian Bullseye https://review.opendev.org/c/openstack/devstack/+/789083 | 08:56 |
| opendevreview | Martin Kopec proposed openstack/tempest master: Use stable constraint in tox to release new tag for Xena https://review.opendev.org/c/openstack/tempest/+/808031 | 09:37 |
| opendevreview | Martin Kopec proposed openstack/tempest master: Add releasenote to tag the Tempest for Xena release https://review.opendev.org/c/openstack/tempest/+/808032 | 09:37 |
| opendevreview | Martin Kopec proposed openstack/tempest master: Use the master constraint in tox.ini again https://review.opendev.org/c/openstack/tempest/+/808033 | 09:38 |
| opendevreview | Martin Kopec proposed openstack/tempest master: Add release notes page for version 29.0.0, 28.1.0, 28.0.0 https://review.opendev.org/c/openstack/tempest/+/808037 | 09:51 |
| kopecmartin | gmann: I pushed patches for tempest 29.0.0 release ^^ also made a few to openstack/releases | 09:57 |
| kopecmartin | so tempest is ready , once we merge all patches we need to | 09:57 |
| kopecmartin | what are we gonna do with patrole? gates are still broken - most likely some of the patrole tests are too, are we gonna release a new release with them? | 09:58 |
| yoctozepto | kopecmartin: what's so broken in patrole? | 10:27 |
| opendevreview | Merged openstack/tempest master: create volume type with repeated name https://review.opendev.org/c/openstack/tempest/+/805124 | 10:38 |
| kopecmartin | yoctozepto: some of the tests fail on gates, only on some of them, I'm not sure if they are not supposed to be executed on the environments they failed or they need to be fixed , both are possible | 10:46 |
| yoctozepto | kopecmartin: ack, so we lack knowledge, ok | 10:49 |
| kopecmartin | yep :/ | 10:51 |
| *** jpena is now known as jpena|lunch | 11:32 | |
| *** pojadhav is now known as pojadhav|brb | 11:44 | |
| *** jpena|lunch is now known as jpena | 12:25 | |
| *** pojadhav|brb is now known as pojadhav | 12:34 | |
| opendevreview | Rodolfo Alonso proposed openstack/tempest master: [OVN] Remove OVN metadata ports from network port list https://review.opendev.org/c/openstack/tempest/+/808081 | 14:00 |
| gmann | kopecmartin: ack about tempest, will reivew | 14:05 |
| gmann | kopecmartin: let me give try today on patrole | 14:06 |
| opendevreview | Radosław Piliszek proposed openstack/devstack master: [WIP] Test running on Debian Bullseye https://review.opendev.org/c/openstack/devstack/+/789083 | 14:14 |
| opendevreview | Balazs Gibizer proposed openstack/devstack master: DNM: reproduce strange placement overcapacity https://review.opendev.org/c/openstack/devstack/+/808084 | 14:23 |
| gmann | slaweq: FYI, though it is not ready yet. I am working in your use case of neutron-tempest-plugin scenario slow test run in serial https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/807809 and depends-on | 14:31 |
| gmann | tox env defining in neutron-tempest-plugin would work as it end up recreating the devstack created tempest virtual env and remove the installed neutron-tempest-plugins | 14:32 |
| gmann | I am trying with run-tempest var but some challenges in that too | 14:32 |
| slaweq | gmann: thx, but I'm not sure if that will solve my issue tbh | 14:33 |
| slaweq | because I wanted to do something to not create new ci jobs for those tests | 14:33 |
| gmann | slaweq: so you want to run scenario slow as serial and rest all in parallel in all scenario jobs right? | 14:34 |
| slaweq | but I wanted more to like mark some our tests as "advanced" (with decorators.attr()) and then run in one job first all !advanced tests and then in serial all advanced tests | 14:34 |
| gmann | slaweq: yeah, any metadata can be used for that like slow/advance or so | 14:35 |
| slaweq | gmann: yes, but I don't want to create new job to run those "slow/advanced" tests | 14:35 |
| gmann | yeah, it will be in existing one. like one worker running all advanced tests and rest all worker !advance but in same job | 14:36 |
| slaweq | ahh, ok | 14:36 |
| slaweq | that would be great | 14:36 |
| gmann | asme way tempest-full does currently | 14:36 |
| slaweq | gmann++ that would be exactly what we need | 14:36 |
| gmann | but if I can make it work. but good to know requirements are clear let's see what solution work | 14:37 |
| slaweq | gmann: thx | 14:38 |
| slaweq | I was also looking at grouping tests in stestr https://stestr.readthedocs.io/en/latest/MANUAL.html#grouping-tests | 14:38 |
| slaweq | that probably would also work for us with specific regex so advanced tests would be run by the same worker one by one | 14:39 |
| slaweq | but tempest run command don't supports that now | 14:39 |
| gmann | slaweq: yeah, till in that we face chalange to pass it in existing tox env. if we need to update tox env then we can pass more args in that. | 14:40 |
| opendevreview | Merged openstack/os-testr master: Use py3 as the default runtime for tox https://review.opendev.org/c/openstack/os-testr/+/793226 | 14:41 |
| opendevreview | Merged openstack/os-testr master: Ussuri+ is python3 only and update python to python3 https://review.opendev.org/c/openstack/os-testr/+/789966 | 14:45 |
| gmann | slaweq: ah, this link i want to show you https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/807800 | 14:49 |
| gmann | pasted other patch link which test same arg for different use case | 14:50 |
| slaweq | gmann: ahh, in this one I see exactly how it can works as we need :) | 14:51 |
| slaweq | thx a lot | 14:51 |
| slaweq | I commented there | 14:51 |
| slaweq | gmann++ | 14:52 |
| gmann | slaweq: thanks | 14:54 |
| gmann | kopecmartin: waiting for this to pass check pipeline and then I will hit +A on all three https://review.opendev.org/c/openstack/tempest/+/808033 | 14:55 |
| kopecmartin | gmann: ack, thanks | 14:58 |
| opendevreview | Chris Wedgwood proposed openstack/tempest master: object_storage: retry object creation on Conflict https://review.opendev.org/c/openstack/tempest/+/807909 | 15:06 |
| clarkb | yoctozepto: the issue is pcp is broken too (and imo far to complicated for our needs). | 15:19 |
| yoctozepto | clarkb: ++ for far too complicated | 15:19 |
| clarkb | frickler: ya maybe dropping both until a reliable replacement can be found is reasonable. Or just accept ~3 failed jobs a day due to pcp packaging being broken | 15:19 |
| *** abhishekk is now known as abhishekk|afk | 15:45 | |
| lajoskatona | gmann: thanks, it's passed now (https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/803752 ) | 16:08 |
| gmann | lbragstad: perfect, glad to know that. | 16:10 |
| gmann | lbragstad: please ignore, | 16:10 |
| gmann | its for lajoskatona :) | 16:11 |
| lbragstad | gmann :) | 16:11 |
| gmann | its 'l' and tab which finds you everywhere :) | 16:11 |
| clarkb | https://zuul.opendev.org/t/openstack/build/337a940d847f41feb4d4030fb7d3e351 is another failure due to pcp in the gate. It is definitely happening often enough we notice :( | 16:34 |
| clarkb | yoctozepto: fwiw I mentioned this on the pcp revert change but collectl might be a reasonable alternative? | 16:34 |
| *** jpena is now known as jpena|off | 16:48 | |
| opendevreview | Rodolfo Alonso proposed openstack/tempest master: [OVN] Remove OVN metadata ports from network port list https://review.opendev.org/c/openstack/tempest/+/808081 | 16:51 |
| *** poojajadhav is now known as pojadhav | 16:56 | |
| *** pojadhav is now known as pojadhav|out | 17:14 | |
| *** frenzy_friday is now known as anbanerj|ruck | 17:21 | |
| yoctozepto | clarkb: never used it | 17:27 |
| clarkb | I haven't either but it appears to be available on debuntu, suse, and epel | 17:34 |
| *** akekane__ is now known as abhishekk | 17:39 | |
| clarkb | also looks like there are tools to import collectl data into pcp (for viewing I guess?) | 17:48 |
| yoctozepto | hmm, epel is worrying me a bit but otherwise it seems reasonable; maybe let's track this issue in launchpad and discuss our options there; could you report it? | 18:10 |
| clarkb | yoctozepto: you want it as a devstack issue or an issue against the pcp package? | 18:24 |
| anticw | tempest-full-test-account-py3 and tempest-full-test-account-no-admin-py3 (both non-voting) look to fail 100% of the time at present, any reason they shouldn't be disabled | 18:54 |
| anticw | before you suggesting fixing them, idkh and this is as much about taking load of the builders which already take an age | 18:54 |
| clarkb | often a job is non voting while it is being developed or in response to a new error that someone intends to fix. If that isn't the case here then moving the job to experimental might make mroe sense or removing it entirely | 18:57 |
| anticw | i looked at all the jobs that passed other checks and all the merged of late and it's failed consistently for all of them | 19:03 |
| anticw | i didn't look say a month ago if it was working | 19:03 |
| anticw | ok, i looked a month ago ... both were failing then | 19:04 |
| anticw | going back to july 13 i can't find any instances where either of them worked | 19:06 |
| *** slaweq1 is now known as slaweq | 19:19 | |
| opendevreview | Archit Modi proposed openstack/whitebox-tempest-plugin master: Add Rbd direct download test https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/786979 | 19:22 |
| opendevreview | Archit Modi proposed openstack/whitebox-tempest-plugin master: Add Rbd direct download test https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/786979 | 19:25 |
| clarkb | anticw: I would say propose the chagne to remove the job then and see if anyone objects | 20:00 |
| *** dviroel is now known as dviroel|afk | 20:00 | |
| opendevreview | Merged openstack/tempest master: Use stable constraint in tox to release new tag for Xena https://review.opendev.org/c/openstack/tempest/+/808031 | 20:16 |
| opendevreview | Merged openstack/tempest master: Add releasenote to tag the Tempest for Xena release https://review.opendev.org/c/openstack/tempest/+/808032 | 20:16 |
| opendevreview | Clark Boylan proposed openstack/devstack master: Try to use collectl instead of dstat https://review.opendev.org/c/openstack/devstack/+/808134 | 20:37 |
| clarkb | yoctozepto: ^ bug filed and is noted in that commit message. Note the chagne itself is super hacky and to do this right I think we want an lib/collectl and treat it as a separate service from dstat. Then disable dstat by default and enable collectl where we were using dstat. But I figured having something that showed us what it would look like first was good before we invest in | 20:38 |
| clarkb | it as others may not like it | 20:38 |
| opendevreview | Archit Modi proposed openstack/whitebox-tempest-plugin master: Add Rbd direct download test https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/786979 | 20:52 |
| opendevreview | Clark Boylan proposed openstack/devstack master: Try to use collectl instead of dstat https://review.opendev.org/c/openstack/devstack/+/808134 | 21:09 |
| gmann | anticw: yeah, they are pre-provisional cred jobs. need to debug the failure and probably make them as voting jobs | 21:24 |
| gmann | anticw: clarkb we cannot remove them as we need to run them for pre-provision creds. we should fix them instead and then make them voting. | 21:25 |
| gmann | I will check those sometime next week | 21:25 |
| clarkb | ya I mean the ideal is always that someone fix them but if they have been running for months non voting and never passing... | 21:52 |
| opendevreview | Clark Boylan proposed openstack/devstack master: Try to use collectl instead of dstat https://review.opendev.org/c/openstack/devstack/+/808134 | 21:55 |
| opendevreview | Clark Boylan proposed openstack/devstack master: Try to use collectl instead of dstat https://review.opendev.org/c/openstack/devstack/+/808134 | 22:41 |
| clarkb | yoctozepto: https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_65e/808134/4/check/devstack/65eb7c0/controller/logs/collectl/ finally. You can use a local install of collectl (I did it in a container) to see the what the raw.gz contains | 23:09 |
| clarkb | yoctozepto: do `collectl -p path_to.raw.gz` and it plays it back | 23:10 |
| clarkb | I think this may be sufficient if we want to update that change to add support for a collcetl service properly and then enable it instead of dstat | 23:10 |
| anticw | gmann: ah ok, thanks, they were working sporadically some time ago | 23:26 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!