| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-12-18 | |||
| 14:34:55 | bauzas | leakypipes: in theory, a magical libvirt API would just tell me if it's assigned or not, but that's not existing | |
| 14:35:27 | bauzas | leakypipes: so the only way I found would be to loop over all the instances and see their XML | |
| 14:36:03 | bauzas | leakypipes: worst case scenario, one created mdev and 10K instances on a host, you can imagine the performance to find which guest is assigned to that mdev | |
| 14:36:30 | bauzas | leakypipes: so, tbh, the only reasonable way I see is to persist the guest assignment information we do in nova | |
| 14:37:09 | bauzas | and consider that if for some specific mdev, we don't have the assignment in the Nova DB, then it means it's free to be allocated | |
| 14:37:09 | ildikov | mriedem: hi | |
| 14:37:26 | bauzas | leakypipes: but you see my point, it would require some persistence model... | |
| 14:37:38 | ildikov | mriedem: IIRC you were pointing me at some point to a review that added some locking similar to what I will need to add for multi-attach | |
| 14:37:40 | bauzas | thoughts on that ? | |
| 14:38:18 | ildikov | mriedem: in case it really is a memory and not a dream and you know what I'm talking about it would be great if you could point me to that one again :) | |
| 14:39:48 | ebbex | mdbooth: nova/virt/libvirt/imagecache.py function "_scan_base_images", should that have changes/deprecations in pike? | |
| 14:40:02 | mriedem | ildikov: i think that's pseudocode in the multiattach spec | |
| 14:40:07 | bauzas | leakypipes: wait, I think about a possible way to hack that without needing to persist that | |
| 14:40:17 | mdbooth | ebbex: Yep, but mriedem wanted changes and I never followed it up. | |
| 14:40:30 | mdbooth | ebbex: Lemme dig out the review, maybe you have time | |
| 14:40:44 | ildikov | mriedem: oh, ok, I will re-read that one again then :) | |
| 14:40:47 | bauzas | leakypipes: if we have a config opt that says "here are my already created mdevs", nova can use it and assume all of them as free | |
| 14:42:04 | mdbooth | ebbex: https://review.openstack.org/#/c/423392/ | |
| 14:42:28 | ebbex | mdbooth: Cool, I'll have a look, thanks | |
| 14:43:03 | mdbooth | ebbex: If you want to fix it up just go ahead, btw. No need to ask again. | |
| 14:46:12 | mriedem | artom: are you adding the upgrade reno to https://review.openstack.org/#/c/528330/ or would you like me to wordsmith that? | |
| 14:47:18 | artom | mriedem, yeah - question though about the migration being optional in the pike backport | |
| 14:47:50 | mriedem | artom: i think it's fine | |
| 14:47:53 | mriedem | we've done this before | |
| 14:47:53 | artom | mriedem, so I've pushed https://review.openstack.org/#/c/528712/ to test it | |
| 14:48:03 | artom | mriedem, ok, I just wanted to be sure | |
| 14:48:13 | artom | Dropping kids off at daycare, back in a bit | |
| 14:51:07 | stephenfin | bauzas: That sounds like something that could be libvirt really should have. If it's not present, could we do something with libvirt and carry the change locally til nova libvirt version > libvirt version with your fix? | |
| 14:51:35 | bauzas | stephenfin: WDYM ? | |
| 14:51:48 | bauzas | hacking it with a TODO ? | |
| 14:52:24 | stephenfin | bauzas: Aye. Figure out if it's something we could do with libvirt internals and propose the change upstream if possible | |
| 14:52:38 | bauzas | I dunno | |
| 14:52:47 | bauzas | lemme write something and we'll discuss | |
| 14:52:53 | stephenfin | Neither do I. Just as a suggestion | |
| 14:53:05 | stephenfin | Given that I imagine other people would benefit if libvirt had it too | |
| 15:07:31 | openstackgerrit | Matt Riedemann proposed openstack/nova stable/pike: Make request_spec.spec MediumText https://review.openstack.org/528330 | |
| 15:07:34 | mriedem | artom: added the release note to the pike backport | |
| 15:07:36 | mriedem | lyarwood: ^ | |
| 15:11:14 | leakypipes | bauzas: sorry, went for breakfast... reading back now | |
| 15:11:40 | bauzas | leakypipes: no worries, just trying a way to not have a huuuuuuge tech debt because $libvirt | |
| 15:11:57 | bauzas | leakypipes: maybe a hangout would be cool for discussing | |
| 15:11:59 | leakypipes | bauzas: ack. | |
| 15:12:09 | leakypipes | bauzas: gimme about 10 minutes and I can do that, sure | |
| 15:12:30 | bauzas | leakypipes: I'll need to get my girls in 15 but we can discuss that later | |
| 15:12:35 | leakypipes | np | |
| 15:13:11 | bauzas | or prozac | |
| 15:13:22 | leakypipes | lol :) | |
| 15:13:36 | openstackgerrit | Matt Riedemann proposed openstack/nova stable/ocata: Make request_spec.spec MediumText https://review.openstack.org/528332 | |
| 15:18:22 | openstackgerrit | Matthew Booth proposed openstack/nova master: Pass DriverBlockDevice to driver.attach_volume https://review.openstack.org/528363 | |
| 15:18:22 | openstackgerrit | Matthew Booth proposed openstack/nova master: Expose driver_block_device fields as attributes https://review.openstack.org/528362 | |
| 15:18:39 | mdbooth | lyarwood: That second one ^^^^ | |
| 15:21:10 | mdbooth | lyarwood: It's a bit disruptive in tests, but amongst its advantages is that it will expose both attachment_id and connection_info without further changes. | |
| 15:21:25 | mdbooth | (Also bdm.uuid, which is what I want it for) | |
| 15:23:52 | openstackgerrit | Matt Riedemann proposed openstack/nova stable/newton: Make request_spec.spec MediumText https://review.openstack.org/528338 | |
| 15:24:48 | artom | mriedem, ah, thanks | |
| 15:25:34 | artom | mriedem, and https://review.openstack.org/#/c/528712/ passed Zuul, so I think we're good, yeah | |
| 15:26:57 | mriedem | ok i've got the ocata and newton ones updated too | |
| 15:27:49 | openstackgerrit | Mr Rambo proposed openstack/nova master: Fix the problems that volume-backed server rebuild https://review.openstack.org/528740 | |
| 15:29:41 | artom | mriedem, I'll push a more complete reno if you don't mind, mentionning that the migration is optional unless you've been affected by the bug, and how to tell if you've been affected | |
| 15:30:03 | mriedem | go nuts | |
| 15:32:04 | artom | http://www.risd.edu/news/stories/go-nads/ | |
| 15:37:32 | openstackgerrit | Artom Lifshitz proposed openstack/nova stable/pike: Make request_spec.spec MediumText https://review.openstack.org/528330 | |
| 15:40:24 | openstackgerrit | Hironori Shiina proposed openstack/nova master: virt/ironic: Implement rescue and unrescue https://review.openstack.org/416487 | |
| 15:40:30 | openstackgerrit | Artom Lifshitz proposed openstack/nova stable/ocata: Make request_spec.spec MediumText https://review.openstack.org/528332 | |
| 15:41:58 | openstackgerrit | Artom Lifshitz proposed openstack/nova stable/newton: Make request_spec.spec MediumText https://review.openstack.org/528338 | |
| 15:42:46 | openstackgerrit | Claudiu Belu proposed openstack/nova master: tests: autospecs all the mock.patch usages https://review.openstack.org/470775 | |
| 15:55:10 | openstackgerrit | Jay Pipes proposed openstack/nova master: allow compute nodes to be associated with host agg https://review.openstack.org/526753 | |
| 16:09:44 | openstackgerrit | Jianghua Wang proposed openstack/nova master: XenAPI: create vGPU for instance https://review.openstack.org/516899 | |
| 16:11:25 | jianghuaw | leakypipes, bauzas: ^ Please help to check if it has addressed your comments? Thanks. | |
| 16:12:56 | mdbooth | mriedem: I vaguely recall there's some reason we don't use docstrings in tests. Am I imagining things? | |
| 16:13:42 | mriedem | mdbooth: that's no longer an issue | |
| 16:13:46 | mriedem | it was because of limitations with nose | |
| 16:13:49 | mriedem | but we don't use nose | |
| 16:14:23 | mdbooth | mriedem: Ah, ok. Glad I didn't stick a -1 on lyarwood 's patch :) Figured that stuff should be in HACKING anyway if it's to be enforceable. | |
| 16:14:38 | mdbooth | lyarwood: Ignore that one ^^^ | |
| 16:14:55 | leakypipes | jianghuaw: +2 from me. :) | |
| 16:15:02 | bauzas | jianghuaw: sorry, today I'm working on implementing the libvirt change, so tomorrow will look | |
| 16:15:21 | jianghuaw | leakypipes, thanks:-) | |
| 16:15:40 | jianghuaw | bauzas, no worries. tomorrow will be very good also. Thanks. | |
| 16:15:42 | mdbooth | leakypipes: Is this now super Monday? | |
| 16:15:45 | leakypipes | jianghuaw: no problem! hope I picked the correct Chinese characters from Google translate in my review comment ;) | |
| 16:15:52 | leakypipes | oh, crap... | |
| 16:16:01 | mdbooth | Hehe | |
| 16:16:02 | jaypipes | thanks mdbooth :) | |
| 16:18:09 | jianghuaw | jaypipes, haha. I was thinking why your Chinese was so good:-) | |
| 16:18:39 | kashyap | Folks using Ubuntu, there's some repo that provides newer packages than the system packages | |
| 16:18:42 | kashyap | What's it called again? | |
| 16:18:59 | mdbooth | jaypipes: https://boingboing.net/2008/07/15/chinese-restaurant-c.html | |
| 16:19:24 | jaypipes | mdbooth: ha! that's awesome | |
| 16:20:03 | mriedem | kashyap: UCA | |
| 16:20:05 | mriedem | ubuntu cloud archive | |
| 16:20:34 | jaypipes | lol | |
| 16:37:26 | kashyap | mriedem: Ah, thank you. | |
| 17:03:14 | melwitt | stephenfin: hey, are you around? | |
| 17:03:19 | stephenfin | I am | |
| 17:04:16 | melwitt | I was just thinking about the xen ci failure on the patch and I realize, TLS is supposed to be disabled by default and you have to opt-in, and I don't think xen ci is doing anything to opt in. so I'm not sure yet why it's going through that path | |
| 17:05:30 | melwitt | so something is probably wrong there if it's going through the TLS path without being enabled. I'm still looking at it but just wanted to bring it up | |
| 17:05:40 | stephenfin | Yeah, I saw that comment on Thursday. I don't know what it is either | |
| 17:05:54 | melwitt | like, if they didn't enable it, their CI should be passing :\ | |
| 17:06:06 | stephenfin | I know sross has some tests merged to DevStack way back related to this, but I don't know how that could possibly trigger it | |