Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-08-14
16:03:08 mriedem but i'm not too worried about non-ironic instances because this is admin-only api stuff and if someone is abusing it for non-ironic instances and it doesn't work as expected....meh?
16:03:18 dansmith okay
16:03:45 tssurya so what't it going to be: manager or driver ?
16:03:55 tssurya at this point I don't mind either
16:04:12 dansmith "at this point" .. translation: "at this point I'm about to shoot both of you guys"
16:04:13 mriedem i forgot the question
16:04:26 tssurya dansmith: ;)
16:04:35 dansmith tssurya: I think it doesn't matter that much, figure it out when you move the code and it's probably okay
16:04:44 tssurya mriedem: meaning we move the task state and https://review.opendev.org/#/c/645611/12/nova/compute/api.py@250
16:04:46 dansmith tssurya: also when are you actually leaving?
16:05:13 mriedem right now :)
16:05:32 dansmith lol
16:05:36 dansmith #ragequit
16:05:37 mriedem (11:04:47 AM) dansmith: tssurya: also when are you actually leaving?
16:05:38 tssurya sorry bad connection
16:05:49 tssurya I leave on Friday
16:05:53 dansmith okay
16:05:56 tssurya so I can work on it this evening and tomorrow
16:06:02 dansmith okay cool
16:06:24 dansmith tssurya: so I said I think manager vs. driver probably doesn't matter that much, you choose while moving the code and it'll probably be okay
16:06:37 dansmith I don't have a strong opinion without seeing it, so maybe just have to pick one and see
16:06:41 tssurya ok awesome thanks
16:06:48 dansmith mriedem: sound okay?
16:06:49 tssurya so first I'll try with the manager and then we can see
16:06:50 mriedem i'll leave a couple of comments
16:07:22 tssurya thanks a lot dansmith and mriedem :D
16:10:48 mriedem done https://review.opendev.org/#/c/645611/12
16:13:10 tssurya ty
16:13:49 dansmith mriedem: yeah, so assuming your "don't do this for libvirt" assertion I think what you said in there is fine
16:13:49 openstackgerrit Merged openstack/nova stable/stein: Fix misuse of nova.objects.base.obj_equal_prims https://review.opendev.org/676289
16:42:22 alex_xu dansmith: mriedem sean-k-mooney anyway thanks for your time today, it is indeed I use a lot today :)
16:43:17 dansmith alex_xu: it's okay, and I think sean-k-mooney has more than 24 hours in his days, so he has plenty to spare :)
16:43:59 mriedem no problem
16:44:39 dansmith mars has 25 hour long days, so maybe sean-k-mooney is from mars
16:45:05 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Hook resource_tracker to remove stale node information https://review.opendev.org/676461
16:52:11 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: rt: only map compute node if we created it https://review.opendev.org/676463
16:54:51 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Hook resource_tracker to remove stale node information https://review.opendev.org/676461
16:55:04 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: rt: only map compute node if we created it https://review.opendev.org/676463
17:00:52 efried dustinc: Ima rebase the sdk series and bump ksa/sdk requirements and get it tracking against this devstack patch https://review.opendev.org/#/c/676268/
17:01:02 efried as soon as sdk 0.34.0 lands in u-c
17:01:36 efried when I say rebase - it's currently based on a really old nova commit, so I'm going to bring it up to current master.
17:05:03 openstackgerrit Matt Riedemann proposed openstack/nova stable/ocata: Hook resource_tracker to remove stale node information https://review.opendev.org/676467
17:12:37 rouk is there a particular reason why OS-FLV-DISABLED:disabled is not exposed via api client? is it intended to be manual api queries only?
17:15:28 rouk i see an old proposed patch from... a long time ago, but it was abandoned.
17:18:43 dansmith efried: on this:
17:18:44 dansmith https://review.opendev.org/#/c/675705/2/nova/tests/functional/regressions/test_bug_1839560.py
17:19:20 efried yah
17:19:22 dansmith efried: you're just noting that in real life we have a service with hostname "host1" but would normally get uuids for the node names, is that right?
17:19:59 efried dansmith: It appears to me as though, in real life, we actually go and create a node for the host, and then delete it when it doesn't come back in the list of nodes.
17:20:31 efried and because that messes with the test, it was easier to name the host 'node1', so that we don't do that deletion.
17:20:51 dansmith we *don't* do that in real life though
17:21:02 efried well
17:21:20 efried what I did was pull down the patch, change node1 to host1 where it was talking about the host, and run the test.
17:21:31 efried and saw the message indicating that we deleted the host1 "node"
17:21:40 dansmith right, it's just an artifact of the test that we create a host and node from the same name
17:21:58 dansmith but in real life that's not how that works,
17:22:05 efried oh, it's start_service that's doing that?
17:22:06 dansmith it's just because of the test plumbing
17:22:27 dansmith no, I think it's because of the fake driver and the service which has already been started by the time we get to the test
17:22:41 dansmith well,
17:22:45 dansmith maybe not been started,
17:22:50 dansmith but because we're using the fake driver,
17:22:58 dansmith we've already run get_available_nodes() by the time start_service has returned
17:23:08 dansmith and that returns the same name as host.hostname or something
17:23:25 dansmith so when he then overrides the node list to be node1, node2, you'd get a deletion of the host1 node that was already created
17:24:18 dansmith anyway, I think your "I pulled down the patch" comment explains what I was asking, just that you were noticing test behavior and not concerned about relevance to real life or something else
17:25:28 efried I was concerned about relevance to real life. But if you're telling me it's just test harness noise, I'm okay with that.
17:25:35 dansmith ack
17:25:43 efried however, if it is test harness noise, I feel like we ought to be able to fix it.
17:25:50 efried more easily than if it's in the real code.
17:26:03 dansmith we could for sure (just made a comment to that effect)
17:26:07 dansmith I just don't think it's that important
17:26:26 dansmith normally having them match is handy
17:26:45 efried "normally" not-ironic
17:26:50 efried no doubt
17:27:24 efried but for ironic cases in general I would think it's never handy, n/a most of the time, but in certain cases like this one really confusing.
17:27:38 dansmith yes, normally not ironic
17:27:44 dansmith yep
17:27:58 dansmith well, yep to the first part, not the second
17:28:10 dansmith I don't really think it's confusing, but probably just because I know the details
17:28:28 dansmith the test isn't testing the ironic case, it's testing the bring-back-the-dead case,
17:28:46 dansmith which is related to how it works with ironic, but... not entirely ironic specific
17:28:49 dansmith ANYway
17:30:17 efried we could add to the fragility of FakeDriver's init_host by special-casing another CONF.host, say ironic-compute, and making self._nodes [] in that case.
18:19:09 mriedem efried: fwiw,
18:19:22 mriedem when we had a fake.set_nodes global i could have used that to avoid this problem,
18:19:28 mriedem but since that was removed, you get the thing described above
18:19:35 mriedem "so when he then overrides the node list to be node1, node2, you'd get a deletion of the host1 node that was already created"
18:20:23 mriedem so i could have started with host1:host1, and then returned available nodes host1:host1 and host1:node2 but that's also weird - which was my reply
18:25:44 efried I let it go, but if y'all keep talking about it, I'm going to have to fix it.
18:25:58 openstackgerrit Merged openstack/nova master: lxc: make use of filter python3 compatible https://review.opendev.org/676263
18:26:16 mriedem do you want to talk about AZ design in a new private cloud instead?
18:26:36 mriedem sean-k-mooney: do you want to backport https://review.opendev.org/#/c/676263/ ?
18:45:00 openstackgerrit Matt Riedemann proposed openstack/nova stable/ocata: Log compute node uuid when the record is created https://review.opendev.org/676487
19:15:07 sean-k-mooney mriedem: i can backport it. how far?
19:15:20 sean-k-mooney mriedem: also i looked at what devstack is doing
19:15:32 sean-k-mooney its useing a cirros image for lxc by defult
19:15:58 sean-k-mooney i have not unpacked it but they use a different init system or rather non standard
19:16:13 sean-k-mooney so that coudl be the issue with the other error
19:16:40 sean-k-mooney i might try to set up an lxc deployment locally and see if i can figure it out later in the week

Earlier   Later