| *** pojadhav|afk is now known as pojadhav | 04:52 | |
| *** soniya29 is now known as soniya29|ruck | 05:09 | |
| *** jpena|off is now known as jpena | 07:14 | |
| opendevreview | Merged openstack/devstack stable/wallaby: Remove libvirt-python from upper-constraints https://review.opendev.org/c/openstack/devstack/+/805824 | 07:18 |
|---|---|---|
| rpittau | thanks gmann frickler :) | 07:42 |
| opendevreview | Dr. Jens Harbott proposed openstack/hacking master: Drop py36 and py37 support https://review.opendev.org/c/openstack/hacking/+/844570 | 07:49 |
| *** soniya is now known as soniya|ruck | 08:18 | |
| opendevreview | Rajat Dhasmana proposed openstack/tempest master: Add test to rebuild volume backed instance https://review.opendev.org/c/openstack/tempest/+/831018 | 09:01 |
| *** soniya|ruck is now known as soniya|ruck|lunch | 09:01 | |
| opendevreview | Dr. Jens Harbott proposed openstack/hacking master: Drop py36 and py37 support https://review.opendev.org/c/openstack/hacking/+/844570 | 09:19 |
| frickler | gmann: ^^ with that the releasenotes look fine, except that the "current" ones still include older releases, but I guess that that is acceptable. when this merges we could release 5.0.0 and add another section for that. and in the future get better at requiring releasenotes incorporated in important changes | 09:21 |
| *** soniya|ruck|lunch is now known as soniya|ruck | 10:46 | |
| whoami-rajat | dansmith, hey, I corrected the tempest test for rebuilding (there was another instance being created, not sure why other tests create it) but now I'm facing another issue with ssh which I'm not able to resolve, can you help me out with that? https://review.opendev.org/c/openstack/tempest/+/831018/16/tempest/api/compute/servers/test_server_actions.py#883 | 11:43 |
| gmann | frickler: yeah, +2 | 13:42 |
| dansmith | whoami-rajat: the upstream run is complaining about a secgroup, so maybe that's related? but yeah, I'll stack locally and see if I can figure it out | 13:44 |
| whoami-rajat | dansmith, ack, will take a look thanks | 14:12 |
| opendevreview | Gorka Eguileor proposed openstack/tempest master: Add LUKS v2 tests https://review.opendev.org/c/openstack/tempest/+/836065 | 14:21 |
| *** soniya|ruck is now known as soniya|ruck|dinner | 14:45 | |
| dansmith | whoami-rajat: I'm not sure why you changed the ssh code, because it seems related.. it must be trying the wrong IP now or something | 14:56 |
| dansmith | I assume you did that for some reason, but I'm not sure why -- it was sshing successfully before | 14:56 |
| whoami-rajat | dansmith, it was but there was another instance there somehow affecting this one (maybe as a failover) which i removed and after that ssh stopped working | 15:06 |
| whoami-rajat | I'm referring to this method https://paste.opendev.org/show/baMcaRPskZzPyE06Urt5/ | 15:07 |
| whoami-rajat | dansmith, after removing this, ssh stopped working ^ | 15:07 |
| dansmith | whoami-rajat: but that's probably affecting the whole class | 15:07 |
| dansmith | whoami-rajat: but why do you think that's affecting the volume rebuild? | 15:08 |
| whoami-rajat | dansmith, so when i put pdb in the test, i saw that the file was being created in one server and the other one was being rebuild so the file always existed because we're rebuilding the wrong one | 15:08 |
| whoami-rajat | dansmith, I'm not sure how any of that works but removing the other instance allowed rebuild to work correctly, (i tried with console, creating a file before rebuild and it wasn't there after the rebuild | 15:09 |
| dansmith | hmm, I don't see how that could be from the code, but okay | 15:09 |
| dansmith | we're calling rebuild on the same server object we're using for the ssh | 15:10 |
| dansmith | (it seems) | 15:10 |
| whoami-rajat | yes we are, that's why I've no idea how the other instance comes into play | 15:11 |
| dansmith | I wouldn't think removing that function would change the ssh behavior, but the other changes you made (validation resources, etc) would be more likely | 15:12 |
| dansmith | whoami-rajat: okay | 15:12 |
| whoami-rajat | in any case, we wouldn't need that secondary instance, we're just testing rebuild on one server | 15:12 |
| dansmith | whoami-rajat: so to be clear, you don't think the other instance is affecting the rebuild, but that we're somehow calling rebuild on the wrong server as a result, is that right? | 15:12 |
| whoami-rajat | dansmith, i introduced those changes to make ssh work (taken reference from other tests) but you can try and revert back, it still won't work, i can push another PS showing that | 15:13 |
| dansmith | adding that function back in does not fix ssh for me locally, FWIW, so I think it must be related to the other changes in 15..16 | 15:13 |
| whoami-rajat | dansmith, yes, we're creating file in server 1 and calling rebuild on server 2 and verifying on server 1 | 15:13 |
| dansmith | okay | 15:13 |
| dansmith | okay I may have fixed it | 15:16 |
| whoami-rajat | nice, what was it? | 15:18 |
| dansmith | I'm not telling | 15:18 |
| whoami-rajat | :D | 15:19 |
| dansmith | lemme do some more poking first | 15:19 |
| whoami-rajat | as long as the test gets fixed, i don't mind knowing the trick | 15:19 |
| whoami-rajat | ack | 15:19 |
| dansmith | hah | 15:20 |
| * whoami-rajat will be away for 1-2 hours | 15:20 | |
| whoami-rajat | but irc will be active | 15:20 |
| dansmith | whoami-rajat: I'll comment on the patch | 15:20 |
| whoami-rajat | dansmith, Great, thanks for looking into it :) | 15:20 |
| dansmith | np | 15:21 |
| opendevreview | Dan Smith proposed openstack/tempest master: Add test to rebuild volume backed instance https://review.opendev.org/c/openstack/tempest/+/831018 | 15:43 |
| dansmith | whoami-rajat: ^ | 15:43 |
| *** soniya|ruck|dinner is now known as soniya|ruck | 15:49 | |
| opendevreview | Dan Smith proposed openstack/tempest master: Add test to rebuild volume backed instance https://review.opendev.org/c/openstack/tempest/+/831018 | 15:50 |
| dansmith | gmann: kopecmartin seems like "StorPool" 3rd party CI is broken? complaining about missing infra nodepools and such' | 15:51 |
| gmann | dansmith: yeah, it has been like that since starting and not sure who own it | 16:30 |
| dansmith | gmann: ah okay I hadn't noticed it complaining about this type of thing before, but I know it fails a lot | 16:31 |
| gmann | yeah, I raised it many times but not sure if kopecmartin knows about iut | 16:31 |
| gmann | it | 16:31 |
| gmann | knows i mean owner of this CI or so | 16:31 |
| dansmith | ah okay, sorry for the noise then | 16:31 |
| dansmith | it's been a while since I worked directly on a tempest patch I guess | 16:32 |
| sean-k-mooney | would people object if i decoupled the useage of the controller inventory host name from being """ when: inventory_hostname == 'controller'"" to say """when: inventory_hostname == groups[controllers][0]""" | 16:39 |
| sean-k-mooney | i have not looked into exactly what that woudl require | 16:40 |
| sean-k-mooney | but currenlty if i want to reuse some of the roles from devstack my inventory host name must be contoller exactly or it wont work | 16:40 |
| sean-k-mooney | my actual hostname can be anything | 16:40 |
| sean-k-mooney | but in the invetory it mush be controller and then you can use ansible_host to map it to your actual hostname | 16:41 |
| sean-k-mooney | actully i feel like that shoudl not need to be done the way im using the role but i get odd issue when i try to use other names that go away if i call it contoler i might have missed somehing but that not a thing to debug on a friday | 16:49 |
| sean-k-mooney | oh its probably this https://github.com/SeanMooney/ansible_role_devstack/blob/16dce64c4b1ee9fc7f6d50eb85f9aafb407ae68e/ansible/roles/devstack_compute/defaults/main.yml#L32-L46= | 16:51 |
| sean-k-mooney | im not consitnetly accessing the contoler via the group | 16:52 |
| sean-k-mooney | if i fix that is houdl be able to rename the vms and generate the local config properly | 16:53 |
| sean-k-mooney | that would actully expalin why this worked fine on the contollers and only failed for the subnodes/computes | 16:57 |
| *** jpena is now known as jpena|off | 17:11 | |
| whoami-rajat | dansmith, wow, the rebuild test passed, I remember seeing one instance containing the file and other one showing rebuild status but that could be unrelated to the failure, anyway thanks for fixing the test | 17:41 |
| dansmith | whoami-rajat: yep, no problem :) | 17:41 |
| whoami-rajat | :) | 17:42 |
| *** soniya|ruck is now known as soniya|ruck|out | 18:05 | |
| opendevreview | James Parker proposed openstack/whitebox-tempest-plugin master: [WIP] Attach and detach an SR-IOV port https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/845382 | 19:50 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!