| opendevreview | Merged openstack/tempest master: Move test_dhcp_port_status_active from neutron-tempest-plugin https://review.opendev.org/c/openstack/tempest/+/824440 | 01:24 |
|---|---|---|
| *** yadnesh|away is now known as yadnesh | 04:12 | |
| opendevreview | Ghanshyam proposed openstack/tempest master: Use project_reader in server test https://review.opendev.org/c/openstack/tempest/+/871210 | 04:26 |
| opendevreview | Amit Uniyal proposed openstack/tempest master: WIP: Adds test for resize server swap to 0 https://review.opendev.org/c/openstack/tempest/+/858885 | 07:16 |
| opendevreview | Amit Uniyal proposed openstack/tempest master: WIP: Adds test for resize server swap to 0 https://review.opendev.org/c/openstack/tempest/+/858885 | 07:21 |
| *** jpena|off is now known as jpena | 08:44 | |
| plibeau4 | hello we have issue with rally-tox-cover on https://review.opendev.org/c/openstack/rally-openstack/+/871586/2 do you have idea to fix it? | 09:25 |
| lpiwowar | gmann: I have answered your question and thank you for approving the fix! https://review.opendev.org/c/openstack/tempest/+/869440/5//COMMIT_MSG | 09:35 |
| opendevreview | Amit Uniyal proposed openstack/tempest master: WIP: Adds test for resize server swap to 0 https://review.opendev.org/c/openstack/tempest/+/858885 | 09:45 |
| *** yadnesh is now known as yadnesh|away | 13:15 | |
| sean-k-mooney | just an FYI i have some fixes for gnocchi installation via devstack both locally and in ci https://review.opendev.org/q/topic:gnocchi-devstack-fixes currently the openstack/telemetry-tempest-plugin patch is blocked by the rust/python cryptograpy issue on train but once that is resolved it woudl be nice to move those forward | 13:34 |
| fungi | early heads up that debian 12 (bookworm) and ubuntu 23.04 (lunar) are going to disallow `sudo pip install` and anything you install with pip will need to go into venvs/virtualenvs | 15:40 |
| fungi | https://salsa.debian.org/cpython-team/python3/-/blob/master/debian/README.venv | 15:41 |
| fungi | https://discuss.python.org/t/10302/69 | 15:41 |
| fungi | this is going to mean devstack refactors to rely entirely on venv installation before openstack moves officially to ubuntu 24.04 lts, but ideally much sooner since anyone on the next versions of debian and ubuntu are going to start hitting this in only a few months | 15:42 |
| sean-k-mooney | fungi: well that means we need to revive/merge the patch that installs everything with devstack in a venv | 15:42 |
| fungi | yes, precisely | 15:42 |
| sean-k-mooney | for what its worth the latest version of that patch was working fine for me | 15:43 |
| sean-k-mooney | but i have not used it in a few months | 15:43 |
| sean-k-mooney | https://review.opendev.org/c/openstack/devstack/+/558930 for context of other who have not seen that | 15:44 |
| fungi | yep, i was thinking specifically of clarkb's implementation | 15:45 |
| clarkb | others should feel welcome to take that work and update it or abandon it and start over etc | 15:45 |
| fungi | glad to hear it was still usable fairly recently, since it's been sitting there for so long | 15:45 |
| fungi | we're coming up on its 5-year anniversary | 15:45 |
| clarkb | I expect my ability to drive something like this to merging these days is much more limited | 15:45 |
| sean-k-mooney | well its been 9 months since i ran it | 15:45 |
| sean-k-mooney | but i was able to use it to install on an m1 macbook pro using debian bookworm | 15:46 |
| clarkb | sean-k-mooney: thats good to hear | 15:46 |
| sean-k-mooney | i have since nuked that linux install actully not macbook pro it was a macbook air | 15:47 |
| sean-k-mooney | as far as i am aware the kvm issues i hit have also been fixed since then | 15:47 |
| sean-k-mooney | on x86 you should not have issues i just pushed some draft comment i had pending but i have lost all the context on this | 15:48 |
| sean-k-mooney | fungi: so i take it this would be a goal to compelte in the B cycle | 15:48 |
| sean-k-mooney | the work and start/resume in devstack before then as the patch is opt in | 15:48 |
| sean-k-mooney | but we would want to have ci jobs testing with this in 2023.2 | 15:49 |
| clarkb | sean-k-mooney: ya I think my suggestion would be to start it as soon as someone is able since you can switch to venvs before it becomes necessary | 15:49 |
| opendevreview | Sofia Enriquez proposed openstack/devstack-plugin-nfs master: Runs all encrypted scenario tests https://review.opendev.org/c/openstack/devstack-plugin-nfs/+/742905 | 15:49 |
| sean-k-mooney | incidently the gnocchi devstack fixes i have in https://review.opendev.org/q/topic:gnocchi-devstack-fixes | 15:50 |
| sean-k-mooney | are actuly due to sudo pip install and the git changes that happend recently | 15:50 |
| sean-k-mooney | i would still prefer to merge those before the venv work but it would not have been requied if we were installing in a venv | 15:51 |
| fungi | sean-k-mooney: i guess if it touches a bunch of separate plugins in projects or something, then the community goals framework is applicable. if it just needs updating in devstack itself, then just get it done, doesn't need coordination with other teams | 16:24 |
| clarkb | fungi: when I last poked at it seriously devstack and grenade were definitely the core bits of it, but it would also break projects (at the time I was trying to make ironic play nice with it) | 16:25 |
| clarkb | fungi: maye what we need is less of a community wide goal and more of a community wide "sorry this is happening and we'll just need to accept it" | 16:25 |
| fungi | and also, like clarkb says, the sooner the better. debian 12 and ubuntu 23.04 are only a few months away and users are going to start using them and expecting devstack to work | 16:25 |
| clarkb | for when the devstack changes land | 16:25 |
| sean-k-mooney | it can be done mostlyin devstack | 16:25 |
| sean-k-mooney | then we just need to opt into it in the ci jobs by setting the flag | 16:26 |
| sean-k-mooney | and eventually the default can be flipped in devstack and or the jobs | 16:26 |
| clarkb | sean-k-mooney: I think I'd make it the default | 16:26 |
| clarkb | once it is working | 16:26 |
| sean-k-mooney | yep me to | 16:26 |
| fungi | yeah, i wouldn't bother with goals/govenance, just make sure it's widely announced | 16:26 |
| clarkb | and force people to opt out rather than in | 16:26 |
| sean-k-mooney | and deprecate teh old way then remvoe it it C | 16:27 |
| sean-k-mooney | fungi: i would suggest just doing a PSA about it on the mailing list and ensure we talk about it in the PTG where approriate | 16:27 |
| fungi | yeah, i'll put something together for the ml, just wanted to give folks a heads up in here as soon as i noticed it was going forward | 16:29 |
| dansmith | gmann: kopecmartin: the test timeouts lately are kinda weird | 17:01 |
| dansmith | three minutes between subsequent test results in the section getting run when the timeout happens | 17:01 |
| *** jpena is now known as jpena|off | 17:35 | |
| opendevreview | Vadym Markov proposed openstack/devstack stable/wallaby: Use proper sed separator for paths https://review.opendev.org/c/openstack/devstack/+/865841 | 18:15 |
| gmann | dansmith: kopecmartin : I am thinking if this is adding something in timeout https://review.opendev.org/c/openstack/tempest/+/821732 | 18:28 |
| gmann | but this add serial run tag at test class level and supposed to run all serial test at the end | 18:28 |
| gmann | but it can add something to job timeout as running serial tests at the end add up some time to total test run | 18:30 |
| dansmith | ah, are those running serial? they have concurrency=4, but yeah, the timing is like they're running one at a time | 18:30 |
| gmann | two tests for now run serial and at the end of all test run irrespective of concurrency | 18:31 |
| dansmith | oh is it just two? | 18:31 |
| gmann | oh no, its is more than two as serial tag in base class of aggregate tests | 18:32 |
| dansmith | okay, because I see two run successfully before it times out in the run I'm looking at | 18:32 |
| gmann | this is taking 187 sec https://zuul.opendev.org/t/openstack/build/9b79f54d3f6c4e05ae6619fdac7ad95a/log/job-output.txt#25738 | 18:36 |
| gmann | which is last test before timeout and it did not go till serial tests run itself | 18:37 |
| gmann | and this does lot of things so not sure which operation adding more time now https://github.com/openstack/tempest/blob/7c9389a00072dbcf7b4772d156f505cabeba4252/tempest/scenario/test_minimum_basic.py#L169 | 18:37 |
| gmann | and this is the test I have seen timeout in multinode job also. | 18:38 |
| gmann | I feel we should mark it as slow test (to run in tempest-slow job) | 18:38 |
| gmann | and if we still see timeout then increase the job time because of those serial test run addition | 18:38 |
| gmann | dansmith: kopecmartin: what you think? | 18:39 |
| gmann | another example of this test taking 258 sec - https://zuul.openstack.org/build/084729c5a1fe46e091a17a50441e99d5/log/job-output.txt#34173 | 18:40 |
| dansmith | gmann: sure, but there's clearly another test that is running for almost 4 minutes *after* that test | 18:40 |
| gmann | humm. yeah that is good point | 18:42 |
| gmann | let's mark it slow and see of we can find another slow tests. | 18:43 |
| gmann | though we do not know what added time in what operation in this release or it is due to ubuntu 22.04 | 18:44 |
| dansmith | yeah, maybe the long one will run and then we'll timeout and see what it is :) | 18:44 |
| gmann | in this, this test fail (took 258 sec) but job reached to the end and running serial aggregate tests https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_084/872566/1/check/tempest-multinode-full-py3/084729c/controller/logs/stackviz/index.html#/stdin/timeline?test=tempest.serial_tests.scenario.test_aggregates_basic_ops.TestAggregatesBasicOps.test_aggregate_basic_ops | 18:48 |
| dansmith | yeah, that's long | 19:02 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Mark test_minimum_basic_instance_hard_reboot_after_vol_snap_deletion as slow test https://review.opendev.org/c/openstack/tempest/+/872691 | 19:27 |
| gmann | dansmith: kopecmartin ^^ also opened bug to track the timeout https://bugs.launchpad.net/tempest/+bug/2004780 | 19:34 |
| kopecmartin | gmann: thanks, i already +2 and +w it as it's Friday evening for me already and i'm not sure when i'll get back to the computer .. | 19:34 |
| dansmith | cool | 19:35 |
| gmann | kopecmartin: thanks. have a nice weekend | 19:35 |
| gmann | I am planning to shutdown my PC today EOD otherwise its every weekend end up working story :) | 19:36 |
| dansmith | gmann: ++ | 19:36 |
| kopecmartin | have a nice weekend everyone | 19:40 |
| kopecmartin | gmann: yeah , do that | 19:40 |
| kopecmartin | i have all work notifications silenced during weekends (and everyday after 6pm) | 19:41 |
| gmann | ++ | 19:41 |
| dansmith | aw man, chunked get made it to the gate, but is now failing there | 20:17 |
| gmann | humm | 21:06 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Mark test_minimum_basic_instance_hard_reboot_after_vol_snap_deletion as slow test https://review.opendev.org/c/openstack/tempest/+/872691 | 23:40 |
| gmann | kopecmartin: ^^ if you are still around. found one more slow tests and need approval | 23:42 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!