| *** edmondsw has joined #openstack-placement | 00:13 | |
| edmondsw | trying to stack devstack and the wait_for_service is failing for placement | 00:16 |
|---|---|---|
| edmondsw | I'm getting a 503 for http://localhost/placement even though I see no errors with journalctl -xea -u devstack@placement-api.service and systemctl status shows it's active | 00:16 |
| edmondsw | somewhere else I should look? | 00:17 |
| edmondsw | I did notice that /var/log/apache2/horizon_error log says "AH02454: uwsgi: attempt to connect to Unix domain socket /var/run/uwsgi/nova-placement-api.socket (uwsgi-uds-nova-placement-api) failed" repeatedly | 00:23 |
| edmondsw | there is no such thing in /var/run/uwsgi, though... it has placement-api.socket, not nova-placement-api.socket | 00:23 |
| edmondsw | I had devstack installed on this system earlier, but today unstacked and cleaned, blew away /opt/stack, and then cloned again to start over... so I'm wondering if something didn't cleaned up in that process | 00:33 |
| *** mriedem has quit IRC | 01:23 | |
| *** e0ne has joined #openstack-placement | 07:13 | |
| *** openstackgerrit has quit IRC | 07:29 | |
| *** helenafm has joined #openstack-placement | 08:24 | |
| *** e0ne has quit IRC | 08:31 | |
| *** tssurya has joined #openstack-placement | 08:59 | |
| *** dtantsur|afk is now known as dtantsur | 09:19 | |
| *** e0ne has joined #openstack-placement | 10:08 | |
| *** cdent has joined #openstack-placement | 10:20 | |
| gibi | cdent: I've added some analysis to https://review.openstack.org/#/c/624055/4 about the notification test failure. no solution yet | 10:21 |
| cdent | thanks | 10:21 |
| cdent | gibi: yeah, this looks like yest another thing we can blame on eventlet :) | 10:22 |
| gibi | cdent: I'm happy to blame eventlet for everything :) | 10:23 |
| cdent | ✔ | 10:23 |
| gibi | I'm continue looking at this problem but if you have ideas what to try then I'm happy to try it | 10:25 |
| cdent | I don't have any immediate ideas other than the usual eventlet hack: stick a time.sleep(0) in there somewhere. | 10:26 |
| cdent | the "where" is the question | 10:26 |
| gibi | cdent: thanks. I can try putting it here and there to see if it helps | 10:28 |
| * cdent crosses fingers | 10:29 | |
| gibi | sometimes I see failin nova.tests.functional.test_servers.ServerMovingTests.test_live_migrate_rollback_cleans_dest_node_allocations and nova.tests.functional.test_servers.ServerLiveMigrateForceAndAbortWithNestedResourcesRequest.test_live_migrate_force_complete as well | 10:34 |
| gibi | both live migration tests | 10:34 |
| gibi | so I assume it is the same root cause | 10:34 |
| gibi | at least the same timeout is visible | 10:37 |
| gibi | in the logs of those tests | 10:37 |
| cdent | gibi, stephenfin : easy cleanup https://review.openstack.org/#/c/624460/ | 11:23 |
| *** e0ne has quit IRC | 11:30 | |
| *** ttsiouts has joined #openstack-placement | 11:47 | |
| *** yikun_ has quit IRC | 11:51 | |
| gibi | cdent: +A | 11:52 |
| *** dtantsur is now known as dtantsur|brb | 11:56 | |
| *** ttsiouts has quit IRC | 12:07 | |
| *** ttsiouts has joined #openstack-placement | 12:07 | |
| *** ttsiouts has quit IRC | 12:12 | |
| *** openstackgerrit has joined #openstack-placement | 12:12 | |
| openstackgerrit | Chris Dent proposed openstack/os-resource-classes master: Sync STANDARDS with placement/nova rc_fields https://review.openstack.org/624370 | 12:12 |
| openstackgerrit | Chris Dent proposed openstack/os-resource-classes master: Tune up documentation to be more useful https://review.openstack.org/624384 | 12:13 |
| *** ttsiouts has joined #openstack-placement | 12:54 | |
| openstackgerrit | Brin Zhang proposed openstack/nova-specs master: Specifying az when restore shelved server https://review.openstack.org/624689 | 12:56 |
| openstackgerrit | Brin Zhang proposed openstack/nova-specs master: Specifying az when restore shelved server https://review.openstack.org/624689 | 12:58 |
| *** dtantsur|brb is now known as dtantsur | 13:20 | |
| openstackgerrit | Merged openstack/placement master: Remove dead code in objects/resource_provider.py https://review.openstack.org/624460 | 13:25 |
| *** dtantsur is now known as dtantsur|brb | 13:56 | |
| gibi | cdent: https://github.com/eventlet/eventlet/issues/508 this seems relevant as the livemigration code in nova uses ThreadPoolExecutor | 14:09 |
| gibi | cdent: going to change that to GreenThreadPoolExecutor | 14:10 |
| gibi | cdent: as the bug report suggests | 14:10 |
| cdent | ha! good sleuthing. knew it was eventlet ;) | 14:10 |
| cdent | coreycb is everywhere | 14:10 |
| *** mriedem has joined #openstack-placement | 14:13 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: Document API error codes https://review.openstack.org/624705 | 14:22 |
| cdent | gibi: that ^ is in response to a bug you created a while ago. It's at least a start | 14:23 |
| gibi | cdent: ack, reading | 14:32 |
| *** irclogbot_2 has quit IRC | 14:32 | |
| gibi | cdent: it seems GreenThreadPoolExecutor helps | 14:32 |
| cdent | \o/ | 14:32 |
| cdent | if you want to commit those change on my patch, that would be great | 14:33 |
| cdent | or before it | 14:33 |
| gibi | cdent: I run a full test and then I will add the change into your patch | 14:34 |
| gibi | cdent: if it become big then we can split | 14:34 |
| cdent | ✔ | 14:34 |
| *** e0ne has joined #openstack-placement | 14:38 | |
| *** irclogbot_2 has joined #openstack-placement | 14:41 | |
| *** irclogbot_2 has quit IRC | 14:52 | |
| gibi | cdent: there are some failed unit tests after switching to GreenThreadPoolExecutor so needs more time before I can push an update | 14:55 |
| cdent | oh, great :( | 14:55 |
| *** irclogbot_2 has joined #openstack-placement | 15:03 | |
| gibi | cdent: needs to change some unit test fixture and the requirements.txt as well so I think I will make this change a separate patch and put your's top of it | 15:24 |
| cdent | sounds good | 15:25 |
| *** dtantsur|brb is now known as dtantsur | 15:25 | |
| *** tssurya has quit IRC | 15:57 | |
| cdent | edleafe: do you have ideas (beyond what I already responded to you on the review) on how to make that test more tasty? | 16:09 |
| edleafe | cdent: in meeting hell. Will reply soon | 16:10 |
| cdent | aye aye | 16:11 |
| cdent | edleafe: If I'm gone, feel free to change it yourself | 16:11 |
| edleafe | ah, this meeting ended early due to holiday no-shows :) | 16:18 |
| edleafe | cdent: Length won't cut it either | 16:18 |
| edleafe | since the order could have been changed. | 16:18 |
| edleafe | IMO, this is where humans doing review come into play | 16:19 |
| cdent | sure, but it is merely a sanity check, not a perfection check | 16:19 |
| cdent | the idea of having the sanity check is making sure that the originator of the change is conscious of the issue | 16:19 |
| cdent | so the test fails, they go read it (and the improved comment that it doesn't yet have) | 16:19 |
| cdent | and they are enlightened | 16:19 |
| edleafe | OK, it will help in the lone case of someone inserting a new entry somewhere other than at the end | 16:20 |
| edleafe | But since all such changes will go through review, the red/green lines in gerrit would tell us exactly what the reviewer needs to know | 16:21 |
| edleafe | If anything, the test gives a false confidence | 16:22 |
| * cdent has limited faith in reviews based on loads of evidence | 16:22 | |
| edleafe | It won't hurt anything, since reviews are still needed, hence my +1 | 16:22 |
| cdent | i'll do the length thing | 16:23 |
| edleafe | what about copying the list wholesale? | 16:23 |
| edleafe | And comparing the copied list to the value in the lib? | 16:23 |
| edleafe | That will take care of insertions, position changes, etc. | 16:23 |
| cdent | that's even more annoying | 16:24 |
| edleafe | But it still won't catch the case of the invalid change being made in both places | 16:24 |
| cdent | the idea is something limited, but somewhat useful | 16:24 |
| cdent | all I'm really after here is someone making a change and not considering the tests | 16:24 |
| mriedem | cdent: some small stuff in the os-resource-classes docs patch https://review.openstack.org/#/c/624384/ | 16:25 |
| cdent | thanks mriedem | 16:25 |
| *** ttsiouts has quit IRC | 16:29 | |
| *** ttsiouts has joined #openstack-placement | 16:30 | |
| *** ttsiouts has quit IRC | 16:30 | |
| *** ttsiouts has joined #openstack-placement | 16:31 | |
| *** tssurya has joined #openstack-placement | 16:51 | |
| openstackgerrit | Chris Dent proposed openstack/os-resource-classes master: Sync STANDARDS with placement/nova rc_fields https://review.openstack.org/624370 | 16:53 |
| openstackgerrit | Chris Dent proposed openstack/os-resource-classes master: Tune up documentation to be more useful https://review.openstack.org/624384 | 16:53 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: Use os-resource-classes in placement https://review.openstack.org/623556 | 17:03 |
| *** helenafm has quit IRC | 17:08 | |
| *** e0ne has quit IRC | 17:21 | |
| *** dtantsur is now known as dtantsur|afk | 17:26 | |
| *** ttsiouts has quit IRC | 17:33 | |
| *** ttsiouts has joined #openstack-placement | 17:34 | |
| *** ttsiouts has quit IRC | 17:38 | |
| *** mriedem is now known as mriedem_lunch | 18:21 | |
| * cdent waves | 18:31 | |
| *** cdent has quit IRC | 18:31 | |
| *** tssurya has quit IRC | 18:51 | |
| *** mriedem_lunch is now known as mriedem | 19:11 | |
| *** e0ne has joined #openstack-placement | 19:20 | |
| *** e0ne_ has joined #openstack-placement | 19:27 | |
| *** e0ne has quit IRC | 19:27 | |
| *** e0ne_ has quit IRC | 19:30 | |
| *** e0ne has joined #openstack-placement | 19:34 | |
| *** e0ne_ has joined #openstack-placement | 19:39 | |
| *** e0ne has quit IRC | 19:41 | |
| *** e0ne has joined #openstack-placement | 19:44 | |
| *** e0ne_ has quit IRC | 19:45 | |
| *** e0ne has quit IRC | 19:45 | |
| *** e0ne has joined #openstack-placement | 19:46 | |
| *** e0ne has quit IRC | 19:54 | |
| *** e0ne has joined #openstack-placement | 20:15 | |
| *** e0ne has quit IRC | 20:16 | |
| *** takashin has joined #openstack-placement | 21:54 | |
| *** smcginnis has quit IRC | 22:29 | |
| openstackgerrit | Matt Riedemann proposed openstack/os-resource-classes master: Sync STANDARDS with placement/nova rc_fields https://review.openstack.org/624370 | 23:15 |
| openstackgerrit | Matt Riedemann proposed openstack/os-resource-classes master: Tune up documentation to be more useful https://review.openstack.org/624384 | 23:15 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!