| *** akekane_ is now known as abhishekk | 07:32 | |
| *** pojadhav is now known as pojadhav|afk | 08:06 | |
| *** pojadhav|afk is now known as pojadhav | 08:36 | |
| *** jpena|off is now known as jpena | 10:10 | |
| *** pojadhav is now known as pojadhav|afk | 10:19 | |
| *** pojadhav|afk is now known as pojadhav | 11:13 | |
| opendevreview | Merged openstack/devstack master: Remove unnecessary unset for project-scoped token in glance https://review.opendev.org/c/openstack/devstack/+/817074 | 11:27 |
|---|---|---|
| *** pojadhav is now known as pojadhav|brb | 11:37 | |
| *** pojadhav|brb is now known as pojadhav | 12:08 | |
| *** pojadhav is now known as pojadhav|brb | 13:38 | |
| kopecmartin | gmann: do you think we can start with this? https://bugs.launchpad.net/tempest/+bug/1911044 - dropping unittest2 | 14:10 |
| kopecmartin | gmann: btw, when the openstack-health is not up to date, is there other way how to check results of a test marked as unstable? | 14:16 |
| *** pojadhav|brb is now known as pojadhav | 14:38 | |
| opendevreview | Artom Lifshitz proposed openstack/whitebox-tempest-plugin master: Stop silently overriding clients with admin clients https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/804780 | 14:47 |
| opendevreview | Artom Lifshitz proposed openstack/whitebox-tempest-plugin master: Remove create_test_server() helper override https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/805300 | 14:47 |
| artom | jparker, ^^ wanna take a look at those when you get a chance? | 14:48 |
| jparker | artom: sure thing | 14:49 |
| artom | Finishing up the admin clients override clean up | 14:49 |
| artom | Hrmm, I'll need to clean up the commit message though, it's no longer relevant | 14:50 |
| artom | In fact, I'll probably just squash them together | 14:51 |
| jparker | artom: you mean in regards to the unit test and the emulator thread test? | 14:51 |
| artom | Yeah | 14:51 |
| artom | With the extensive refactoring you did to remove NSM, there's little need to keep these separate | 14:51 |
| *** pojadhav is now known as pojadhav|brb | 14:58 | |
| opendevreview | Artom Lifshitz proposed openstack/whitebox-tempest-plugin master: Stop silently overriding clients with admin clients https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/805300 | 15:05 |
| artom | jparker, ^^ there, maybe wait for CI to confirm I didn't screw it up first :) | 15:05 |
| * jparker looking now | 15:09 | |
| opendevreview | Merged openstack/tempest master: Replace deprecated assertDictContainsSubset https://review.opendev.org/c/openstack/tempest/+/819751 | 15:32 |
| opendevreview | Alfredo Moralejo proposed openstack/devstack master: Support CentOS Stream 9 https://review.opendev.org/c/openstack/devstack/+/800903 | 15:47 |
| gmann | kopecmartin: on unittest, let me finish pinning tempest in train testing (after fixing the neutron issues) and after that we can iterate on this if stestr version etc are all ok for master supported branches https://review.opendev.org/c/openstack/devstack/+/816549 | 15:54 |
| gmann | kopecmartin: on checking test status for unstable, no I do not know any other way then o-h. | 15:58 |
| jparker | artom: I'm trying to understand why the zuul is failing for the commit, does create_test_server default to os_primary and then when os_admin is called by get_server_xml it can't see the name? I would still think os_admin would get that information as well no? | 15:58 |
| artom | jparker, yeah, it's weird. I suppose it's a good thing, forces us to be clean and explicit about clients and resource cleanup | 15:59 |
| *** pojadhav|brb is now known as pojadhav | 16:29 | |
| *** gmann is now known as gmann_afk | 17:34 | |
| *** pojadhav is now known as pojadhav|out | 17:46 | |
| *** jbadiapa is now known as jbadiapa|off | 17:46 | |
| *** jpena is now known as jpena|off | 18:18 | |
| artom | jparker, huh, so in my "stop overriding with admin clients" patch I see the following error for an instance built from an image gotten from copy_default_image: Invalid input received: Inv | 18:41 |
| artom | alid image identifier or unable to access requested image. | 18:41 |
| artom | You had that as well, right? But for a different thing... | 18:41 |
| *** gmann_afk is now known as gmann | 18:53 | |
| opendevreview | Artom Lifshitz proposed openstack/tempest master: WIP: Don't override images client in compute admin tests https://review.opendev.org/c/openstack/tempest/+/820062 | 19:09 |
| opendevreview | Artom Lifshitz proposed openstack/whitebox-tempest-plugin master: Stop silently overriding clients with admin clients https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/805300 | 19:19 |
| jparker | artom: ack yea I believe it was for the rescue volume back commit I'm working on, it looks like you may have solved it though with your latest ps | 19:29 |
| artom | jparker, hopefully so, though it needs a change in tempest that I suspect will explode many more things | 19:42 |
| artom | ... that will hopefully be easy to fix with just s/image_client/admin_image_client/ | 19:43 |
| opendevreview | Merged openstack/tempest master: Define default OS_TEST_TIMEOUT for every tempest tox env https://review.opendev.org/c/openstack/tempest/+/703315 | 20:14 |
| jparker | artom: unless the connection method is floating the create_test_server will no longer return active state correct since the wait_until isn't explicitly passed? | 20:31 |
| jparker | artom: I just tried running the commit locally and its failing early b/c it can't find a host since its in build state still | 20:31 |
| artom | jparker, yeah, I wonder if we can just make wait_until='ACTIVE' the default | 20:54 |
| artom | In tempest itself - seems like the most common use of it | 20:54 |
| jparker | artom: yea I would think it would be the most common state required | 20:56 |
| artom | Yeah, a quick grep told me there's only like 2 or cases of wait_until something other than ACTIVE | 20:56 |
| artom | Dammit, there's the following logic in the common/compute.py implementation: | 21:01 |
| artom | if CONF.validation.connect_method == 'floating': | 21:01 |
| artom | if wait_until is None: | 21:01 |
| artom | wait_until = 'ACTIVE' | 21:01 |
| artom | IOW, it relies on the default being None | 21:01 |
| artom | So we can't change it in tempest directly. Guess we'll just do it in whitebox ourselves... | 21:01 |
| opendevreview | Ade Lee proposed openstack/tempest master: WIP - replace paramiko with libssh https://review.opendev.org/c/openstack/tempest/+/806274 | 21:02 |
| jparker | artom: but if wait_until defaults to ACTIVE couldn't it just be removed? | 21:05 |
| artom | jparker, yeah, I'd have liked to default it to ACTIVE in tempest | 21:05 |
| jparker | artom: so couldnt the logic be removed then? | 21:06 |
| artom | Oh, I see what you're saying | 21:06 |
| artom | I think you're right, the only case it would still be needed for is if a test passed explicitly passed wait_until=None (instead of relying on the default) but passed validatable=True | 21:09 |
| artom | And relied on that logic | 21:09 |
| artom | Which would be... very weird | 21:09 |
| jparker | I guess the other concern would be added time | 21:12 |
| jparker | artom: if there are a large number of instances where wait_until just is not passed and now they have to wait for the instance to reach ACTIVE | 21:13 |
| artom | jparker, also a good point | 21:14 |
| artom | I guess in the spirit of being explicit and not relying on potentially hard to find defaults, we should just add it everywhere in whitebox... | 21:14 |
| artom | There aren't that many places... | 21:15 |
| jparker | yea its fine | 21:15 |
| jparker | artom: I can work on it some tomorrow and do local tests to make sure everything is working | 21:17 |
| jparker | artom: when I added it locally to one of the tests it started passing again | 21:17 |
| artom | jparker, well, I have something in progress locally, lemme at least just finish that up and push it | 21:18 |
| jparker | artom: ack sounds good | 21:18 |
| jparker | artom: need to go get the kids from school I can sync up with you tomorrow | 21:18 |
| gmann | kopecmartin: this is ready https://review.opendev.org/c/openstack/tempest/+/801920 | 21:19 |
| opendevreview | Artom Lifshitz proposed openstack/tempest master: WIP: Don't override images client in compute admin tests https://review.opendev.org/c/openstack/tempest/+/820062 | 21:24 |
| opendevreview | Artom Lifshitz proposed openstack/whitebox-tempest-plugin master: Stop silently overriding clients with admin clients https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/805300 | 21:38 |
| opendevreview | Artom Lifshitz proposed openstack/tempest master: Don't override images client in compute admin tests https://review.opendev.org/c/openstack/tempest/+/820062 | 21:38 |
| opendevreview | Artom Lifshitz proposed openstack/whitebox-tempest-plugin master: Stop silently overriding clients with admin clients https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/805300 | 23:54 |
| opendevreview | Artom Lifshitz proposed openstack/whitebox-tempest-plugin master: Don't show server twice needlessly https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/820081 | 23:54 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!