| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-02-22 | |||
| 10:48:11 | efried | Now I have kids and a side business. | |
| 10:48:26 | bauzas | I have kids too | |
| 10:48:37 | bauzas | but taking them at school at 4:30pm kills me | |
| 10:48:51 | bauzas | so I tend to try working later in the evening | |
| 10:51:34 | efried | My wife *usually* handles the school runs on both ends. She has multiple flexible part-time jobs, so it's fair :) | |
| 10:52:28 | efried | hm | |
| 10:52:41 | efried | bauzas: How do you feel about this: | |
| 10:52:45 | bauzas | I recently invested in a Bose QC35 II, that helps a lot | |
| 10:53:05 | efried | I think when you do GET /allocation_candidates?required=BOGUS_TRAIT you get a 404 (because the trait isn't found). | |
| 10:53:21 | efried | Should that be a 400 instead? (bad query params) | |
| 10:53:43 | bauzas | I'm not an API specialist | |
| 10:53:52 | efried | Yeah, but you're *here* | |
| 10:53:53 | bauzas | but it's not a malformed query ? | |
| 10:53:56 | bauzas | right? | |
| 10:54:04 | efried | Well, "malformed"... | |
| 10:54:06 | bauzas | it's just that the resource isn't found | |
| 10:54:11 | bauzas | hence the 404 | |
| 10:54:17 | efried | But it's not the *resource* that's not found. It's the *trait*. | |
| 10:55:07 | efried | I think all of our placement APIs return 200 with an empty list when you use filter criteria that yield zero results (but are otherwise valid) | |
| 10:55:51 | efried | Anyway, this seems wrong to me. I'm going to raise a bug and see how others feel. | |
| 10:58:21 | efried | never mind, I think I'm wrong. | |
| 10:59:05 | openstackgerrit | Surya Seetharaman proposed openstack/nova-specs master: Support disabling a cell https://review.openstack.org/546684 | |
| 11:01:30 | openstackgerrit | Surya Seetharaman proposed openstack/nova-specs master: Support disabling a cell https://review.openstack.org/546684 | |
| 11:01:45 | openstackgerrit | Eric Fried proposed openstack/nova master: rp: GET /resource_providers?required= |
|
| 11:01:49 | efried | ka-blam ^ | |
| 11:02:22 | bauzas | efried: sorry, my computer decided on its own without asking me first to consider taking a rest | |
| 11:02:31 | bauzas | so it froze | |
| 11:02:44 | efried | heh, lovely. | |
| 11:03:23 | efried | So yeah, it was my mistake; GET /allocation_candidates probably returns 400 in that case. I was missing an exception handler. | |
| 11:03:35 | bauzas | efried: sweet | |
| 11:03:43 | efried | Feel free to review the above placement microversion bump with new query param :) | |
| 11:03:56 | bauzas | efried: sure, /me needs to look a few things | |
| 11:04:08 | efried | Who knows, if we get the right people on IRC today, we could have that blueprint completed before the PTG... | |
| 11:08:18 | efried | Ah, dangit, forgot the reference docs. | |
| 11:08:23 | bauzas | efried: well, sure, but I'm triaging the bugs now | |
| 11:08:44 | bauzas | 18 new bugs yesterday, now 22 :' | |
| 11:08:45 | bauzas | :( | |
| 11:08:46 | efried | It's totally not urgent. Just be a nice tickymark to have | |
| 11:14:16 | openstackgerrit | Eric Fried proposed openstack/nova master: rp: GET /resource_providers?required= |
|
| 11:14:20 | efried | API ref added | |
| 11:21:23 | hrw | monrign | |
| 11:24:22 | bauzas | efried: interesting bug https://bugs.launchpad.net/nova/+bug/1750790 | |
| 11:24:22 | openstack | Launchpad bug 1750790 in OpenStack Compute (nova) "resources on target host aren't released if resize fails" [Undecided,New] | |
| 11:24:31 | bauzas | efried: I wonder if we should support that | |
| 11:24:39 | bauzas | some possible exception we forgot | |
| 11:25:42 | efried | bauzas: Just based on the title, it sounds like a valid bug. | |
| 11:25:58 | bauzas | efried: the problem was that the folk forgot to provide SSH keys | |
| 11:26:00 | bauzas | hence the error | |
| 11:26:59 | efried | Does the instance exist on both hosts while it's in the error state? | |
| 11:27:26 | bauzas | no, because https://docs.openstack.org/nova/latest/admin/configuration/resize.html | |
| 11:28:00 | efried | Oh, I see, so it didn't even get started. | |
| 11:28:56 | efried | bauzas: 2.16 - that's Pike, right? | |
| 11:29:24 | efried | You may want to check with dansmith on this one, but I'm pretty sure in Pike we were still doing the "duplicate allocations" thing the wrong way. | |
| 11:29:36 | efried | So this is at least a duplicate bug, and possibly a wontfix | |
| 11:29:39 | bauzas | efried: 16.0.3 is the Pike 3rd stable release | |
| 11:30:34 | efried | Yeah, IIRC, we punted on actually trying to fix the duplicate allocations thing in pike. I actually don't even know if we fully fixed it in queens - it depended on cdent's POST allocations thing - but I want to say we did. | |
| 11:30:59 | efried | But don't quote me on this. dansmith will know for sure, mriedem probably will too. | |
| 11:31:47 | efried | um, I also may be thinking of live migration - which isn't the same as resize?? I can never keep it straight. | |
| 11:32:03 | efried | Gotta run for a couple hours. | |
| 12:14:57 | openstackgerrit | Carlos Goncalves proposed openstack/os-vif master: Introduce network namespace aware plugging https://review.openstack.org/546961 | |
| 12:29:09 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Transform servergroup.addmember notification https://review.openstack.org/541101 | |
| 13:42:43 | mriedem | melwitt: are you running the meeting today? | |
| 13:46:44 | mriedem | any other takers? | |
| 13:46:49 | mriedem | i've got a kid i have to get on the bus | |
| 13:47:50 | jaypipes | mriedem: when is it? | |
| 13:48:13 | mriedem | 12 minutes | |
| 13:48:41 | melwitt | mriedem: yeah, I didn't realize gibi was on vacation this week | |
| 13:48:53 | mriedem | ok | |
| 13:48:54 | kashyap | melwitt: Wauw, isn't it like 05:45 AM for you there? | |
| 13:48:59 | melwitt | tis | |
| 13:49:12 | kashyap | Jee | |
| 13:50:40 | melwitt | oh, fyi everyone I'm traveling today and tomorrow to the PTG. so I won't be around later today and tomorrow | |
| 13:51:00 | openstackgerrit | Merged openstack/nova stable/queens: Only pull associated *sharing* providers https://review.openstack.org/546740 | |
| 13:51:14 | jaypipes | efried: you around? | |
| 13:51:20 | efried | jaypipes: Yup. | |
| 13:51:30 | sean-k-mooney | melwitt: nice :) going to do some siteseeing over the weekend? | |
| 13:51:47 | jaypipes | efried: you pinged me last night about something but it's gone from my scrollback... what was it? | |
| 13:52:09 | efried | jaypipes: I desperately needed some SQL help. | |
| 13:52:22 | melwitt | sean-k-mooney: on saturday yes. on sunday we have an internal meeting day | |
| 13:52:27 | jaypipes | efried: I don't know anything about SQL | |
| 13:52:29 | efried | jaypipes: But I slogged through it, with much cursing, and put up a patch for GET /resource_providers?required=<traits> | |
| 13:52:45 | jaypipes | efried: ah, cool. | |
| 13:53:11 | efried | jaypipes: We could conceivably get this blueprint closed before the PTG :) | |
| 13:53:24 | jaypipes | efried: nothing wrong with that, IMHO. | |
| 13:53:25 | efried | jaypipes: https://review.openstack.org/#/c/546837/ | |
| 13:53:33 | efried | Yeah, I thought it would be a nice tickymark. | |
| 13:53:54 | jaypipes | efried: that multiattach failure related? | |
| 13:54:04 | jaypipes | don't see how it could be, but worth asking... | |
| 13:54:06 | efried | jaypipes: No, random timeout. | |
| 13:54:08 | efried | I rechecked. | |
| 13:54:10 | jaypipes | k | |
| 13:54:44 | jaypipes | melwitt, dansmith, mriedem, bauzas: one of the easier specs to +2/W here: https://review.openstack.org/#/c/546009/2 | |
| 13:55:36 | bauzas | mriedem: melwitt: gibi_off told me about that, I was planning to chair it | |
| 13:55:43 | bauzas | but fine by me | |
| 13:56:10 | sean-k-mooney | jaypipes: efried: if your makeing a list of blueprints that could be closed before the ptg the nic feature based scheduling blueprint has been code complete since pike and is fully rebased for rocky. spec needs to be reapproved and patches re reviwed though | |
| 13:56:36 | melwitt | bauzas: okay, cool. yeah, I realized I'd have to get up early to make sure you were able to run it anyway so I might as well do it. it was my bad. thanks though | |
| 13:56:38 | jaypipes | sean-k-mooney: link pls | |
| 13:57:28 | sean-k-mooney | jaypipes: code https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/enable-sriov-nic-features spec https://review.openstack.org/#/c/545951/ | |
| 13:57:46 | sean-k-mooney | on tox faild on the spec ill go fix that now | |
| 13:58:10 | jaypipes | sean-k-mooney: danke | |