| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-04-19 | |||
| 14:01:31 | mriedem | my offer to mail it still stands | |
| 14:02:16 | fried_rice | Then it'll be, "the nickel's in the mail." I've heard that one before too. | |
| 14:04:19 | fried_rice | I get that same NoSuchGroupError in stein and rocky. So I'm guessing there's a bug in the test setup where the test env doesn't run through the code path that registers the options if I'm just running those two tests. | |
| 14:05:55 | mriedem | https://photos.app.goo.gl/giiB1e8RD8WztkaQ8 | |
| 14:05:57 | mriedem | i've set a reminder | |
| 14:06:23 | mriedem | fried_rice: yeah i get the same: $ tox -r -e py36 -- nova.tests.unit.test_rpc --until-failure | |
| 14:07:38 | fried_rice | so that's a red herring for our gate failure :( | |
| 14:12:12 | fried_rice | If you actually pay me off, I won't be able to keep giving you crap about it. We need a different plan. | |
| 14:14:46 | mriedem | what i'd like to know is what is calling oslo.config when the death spiral starts | |
| 14:23:22 | mriedem | well i fixed that latent traceback at least :) | |
| 14:23:25 | fried_rice | Running all of py36 locally I got these: | |
| 14:23:26 | fried_rice | {2} nova.tests.unit.pci.test_utils.PciDeviceMatchTestCase.test_spec_extra_key [] ... inprogress | |
| 14:23:26 | fried_rice | {7} nova.tests.unit.objects.test_instance_group.TestInstanceGroupObject.test_get_by_hint [] ... inprogress | |
| 14:24:53 | fried_rice | oh, by the way, that ^ was on rocky | |
| 14:25:54 | fried_rice | what have we backported lately? | |
| 14:26:57 | mriedem | shrug, lots of stuff | |
| 14:27:07 | mriedem | what is getting an option with a namespace? https://github.com/openstack/oslo.config/blob/6.8.1/oslo_config/cfg.py#L2614 | |
| 14:27:12 | mriedem | wtf is a config option namespace? | |
| 14:27:15 | mriedem | stephenfin: ^? | |
| 14:27:31 | mriedem | http://logs.openstack.org/45/649345/7/check/openstack-tox-py36/ba15c17/job-output.txt.gz#_2019-04-18_18_10_53_423952 | |
| 14:27:37 | mriedem | that's where the stack overflow starts | |
| 14:28:24 | fried_rice | and why am I suddenly getting | |
| 14:28:24 | fried_rice | Error: pg_config executable not found. | |
| 14:28:24 | fried_rice | when rebuilding my venv?? | |
| 14:28:46 | mriedem | https://github.com/openstack/oslo.config/blob/6.8.1/oslo_config/cfg.py#L2183 | |
| 14:29:33 | mriedem | might be due to newer psycopg2? | |
| 14:29:42 | mriedem | are you on 18.04? | |
| 14:29:46 | mriedem | bionic i mean | |
| 14:30:38 | fried_rice | yes. Have been for three months. Apparently I suddenly needed libpq-dev | |
| 14:31:01 | fried_rice | four months | |
| 14:31:05 | fried_rice | time flies | |
| 14:31:46 | fried_rice | On master I just got | |
| 14:31:46 | fried_rice | {6} nova.tests.unit.objects.test_instance.TestRemoteInstanceObject.test_create_with_extras [] ... inprogress | |
| 14:31:46 | fried_rice | '>' not supported between instances of 'NoneType' and 'datetime.datetime' | |
| 14:31:58 | fried_rice | so it's not RPC I guess | |
| 14:32:40 | mriedem | heh i can't recreate oslo_config.cfg.NoSuchGroupError: no such group [oslo_messaging_notifications] now | |
| 14:32:51 | yan0s | So, novnc produces the error I mentioned earlier when both a key pair and a cloud-config (setting user password) are set | |
| 14:33:12 | mriedem | fried_rice: that's what i saw yesterday when looking at this | |
| 14:33:33 | mriedem | it's not in the logs everytime the stack overflow happens, but i know the cells 1 removal stuff touched some of that code | |
| 14:55:08 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Only set oslo_messaging_notifications.driver if using RPCFixture https://review.openstack.org/653954 | |
| 15:43:28 | kashyap | Hmm, https://review.openstack.org/ is offline | |
| 15:44:20 | dansmith | kashyap: see topic | |
| 15:45:45 | kashyap | dansmith: Yeah, noticed it. _Just_ when I was hitting send on a comment I saw it go down | |
| 15:46:10 | kashyap | dansmith: But, I trained my muscle memory to copy it into an editor; before pasting it :-) | |
| 15:46:15 | kashyap | dansmith: Shouldn't you be offline, too? | |
| 15:46:46 | dansmith | no :) | |
| 15:46:58 | dansmith | this day has no special meaning for me | |
| 15:47:17 | kashyap | Hehe, nod. | |
| 15:48:33 | edleafe | It *is* the 244th anniversary of the start of the American Revolution. Surely *that* has meaning for you? | |
| 16:00:57 | mriedem | weeeee https://bugs.launchpad.net/nova/+bug/1825537 | |
| 16:00:59 | openstack | Launchpad bug 1825537 in OpenStack Compute (nova) "finish_resize failures incorrectly revert allocations" [Medium,Triaged] | |
| 16:01:12 | mriedem | mnaser: eandersson: ^ will add to the placement allocations are wrong woes | |
| 16:01:34 | mriedem | i've got a recreate test locally which i'll push up once gerrit is back | |
| 16:11:11 | mriedem | dansmith: interested in your opinion on the options i've laid out in https://bugs.launchpad.net/nova/+bug/1825537 for fixing it - comment 2 | |
| 16:11:12 | openstack | Launchpad bug 1825537 in OpenStack Compute (nova) "finish_resize failures incorrectly revert allocations" [Medium,Triaged] | |
| 16:11:57 | dansmith | mriedem: this is getting really old | |
| 16:12:04 | dansmith | can't you do something positive to help the project? | |
| 16:12:22 | mriedem | would you like auto-formatted code style? | |
| 16:12:33 | mriedem | it will bring the new contribs | |
| 16:12:57 | dansmith | yes please! | |
| 16:13:48 | dansmith | mriedem: I'm highly allergic to changing the host/node assignment timing | |
| 16:14:08 | dansmith | not because what we do today is good or right, but just because I'm afraid of breaking tons of other things that encode those assumptuons | |
| 16:14:11 | dansmith | *assumptions | |
| 16:14:51 | mriedem | dansmith: i.e. "I think this is probably not really an option because finish_resize has never done this on failure and we don't really know what state the instance is in" | |
| 16:15:09 | dansmith | sure | |
| 16:15:41 | mriedem | i don't think that's really a good option either, i just listed it for completeness | |
| 16:16:46 | mriedem | an auto-revert is appealing but i very much doubt the revert code is graceful enough to handle it, b/c of baked in assumptions about the state of the world when a revert is started | |
| 16:17:14 | mriedem | e.g. "cleaning up networking...wtf there is no networking on this host KABOOM!!!" | |
| 16:22:47 | fried_rice | mriedem: Did you already do the e-r thing for bug 1825435? | |
| 16:22:48 | openstack | bug 1825435 in OpenStack Compute (nova) "TestRPC unit tests intermittently fail with "'>' not supported between instances of 'NoneType' and 'datetime.datetime'" - maybe due to "Fatal Python error: Cannot recover from stack overflow."" [High,Confirmed] https://launchpad.net/bugs/1825435 | |
| 16:22:54 | mriedem | yeah | |
| 16:22:58 | fried_rice | k | |
| 16:23:33 | mriedem | http://status.openstack.org/elastic-recheck/#1825435 | |
| 16:28:04 | mriedem | dansmith: ok option 1 fix is quick and easy and already done locally | |
| 16:28:08 | mriedem | it's essentially just doing what confirm_resize does | |
| 16:28:12 | dansmith | cool | |
| 18:00:39 | eandersson | Sweet mriedem_lunch | |
| 18:01:02 | eandersson | fried_rice, We have this in Rocky as well. I don't think it ever worked. | |
| 18:01:58 | eandersson | https://github.com/openstack-dev/devstack/blob/master/lib/nova_plugins/hypervisor-ironic#L44 | |
| 18:02:13 | eandersson | We don't add a region_name, so that path was never really tested. | |
| 18:43:07 | fried_rice | eandersson: Okay. It's nearly impossible to test with all the permutations of ksa opts, times all the permutations of service catalog and endpoint setups. Inevitable that we missed things. | |
| 18:43:20 | eandersson | Yep | |
| 18:43:36 | eandersson | For sure. | |
| 18:43:47 | eandersson | I opened a PR to have it added. | |
| 18:44:12 | fried_rice | eandersson: IIUC this is only reproducible when there's a real ironicclient listening on the other end? | |
| 18:44:38 | eandersson | The problem is really that the real ironicclient hides this. | |
| 18:44:41 | fried_rice | right | |
| 18:44:50 | fried_rice | oh | |
| 18:44:57 | fried_rice | wait, you mean the *real* real ironicclient? | |
| 18:45:03 | eandersson | Since when you pass on a None value, the ironicclient tries to figure it out. | |
| 18:45:19 | eandersson | Yep | |
| 18:45:35 | fried_rice | so... where is it a problem? | |
| 18:46:09 | eandersson | Well the problem is that ironicclient is smart, but not smart enough | |
| 18:46:22 | eandersson | when you pass on a None endpoint it just gets the first endpoint in the catalog | |
| 18:47:15 | eandersson | and does not take region into account | |
| 18:48:47 | eandersson | There has been multiple bugs with this code pass, so been difficult to know exactly where things are failing | |
| 18:49:12 | eandersson | e.g. https://github.com/openstack/python-ironicclient/commit/466be3b6568b643605d826e5aa26d9a344cc74ae | |
| 18:49:40 | eandersson | *code path | |
| 18:50:48 | fried_rice | I don't actually see where we're taking region into account at all. | |
| 18:51:04 | eandersson | Yea - that was my first assumption. | |