| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-09-06 | |||
| 10:46:31 | artom | stephenfin, sean-k-mooney, actually, while we're on the topic, I thought you could get CPU pinning with https://libvirt.org/formatdomain.html#elementsCPUTuning without NUMA pinning with https://libvirt.org/formatdomain.html#elementsNUMATuning | |
| 10:46:47 | artom | For example by using only the cpu_policy=dedicated extra spec | |
| 10:46:49 | sean-k-mooney | artom: yes you can | |
| 10:46:57 | sean-k-mooney | no | |
| 10:47:05 | sean-k-mooney | if you use cpu_policy=dedicated | |
| 10:47:11 | sean-k-mooney | you will have a numa toplogy | |
| 10:47:20 | sean-k-mooney | if you set vcpu_pin_set on the host | |
| 10:47:22 | artom | But looking at the code of _get_guets_numa_config, IIUC if it sets vcpupin, it'll also set the memnode stuff | |
| 10:47:26 | sean-k-mooney | and set nothing on the flavor | |
| 10:47:42 | sean-k-mooney | you get <vcpu placement='static' cpuset="1-4,^3,6" current="1">2</vcpu> | |
| 10:47:45 | artom | sean-k-mooney, right, that's the case with <vcpu cpuset=blah> that I'm not handling | |
| 10:47:55 | sean-k-mooney | or whatever the vcpu_pin_set is but no numa affinity | |
| 10:48:05 | artom | I was talking about <cputune> without <numatune> | |
| 10:48:17 | artom | Is that possible? Don't think so | |
| 10:48:19 | sean-k-mooney | oh libvirt allows that but we never generate it | |
| 10:48:23 | artom | Aha, ok | |
| 10:48:53 | artom | Confirms what I was seeing in the code, thanks | |
| 10:49:03 | sean-k-mooney | we only generate the cputune element if you have pinning nebaled which give you a numa toplogy | |
| 10:49:16 | sean-k-mooney | we do generate numatune without cpu tune i think | |
| 10:49:26 | sean-k-mooney | e.g. if you just enable hugepages but no pinning | |
| 10:49:32 | sean-k-mooney | or hw:numa_nodes=X | |
| 10:50:23 | artom | sean-k-mooney, you sure? It would happen only if https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L4609 doesn't iterate at all | |
| 10:50:52 | sean-k-mooney | we definitly numa affine hugepages | |
| 10:50:52 | openstackgerrit | Luyao Zhong proposed openstack/nova master: libvirt: report VPMEM resources by provider tree https://review.opendev.org/678454 | |
| 10:50:53 | openstackgerrit | Luyao Zhong proposed openstack/nova master: libvirt: Support VM creation with vpmems and vpmems cleanup https://review.opendev.org/678455 | |
| 10:50:53 | openstackgerrit | Luyao Zhong proposed openstack/nova master: Parse vpmem related flavor extra spec https://review.opendev.org/678456 | |
| 10:50:54 | openstackgerrit | Luyao Zhong proposed openstack/nova master: libvirt: Enable driver configuring PMEM namespaces https://review.opendev.org/679640 | |
| 10:50:54 | openstackgerrit | Luyao Zhong proposed openstack/nova master: Add functional tests for virtual persistent memory https://review.opendev.org/678470 | |
| 10:50:55 | openstackgerrit | Luyao Zhong proposed openstack/nova master: doc: attaching virtual persistent memory to guests https://review.opendev.org/680300 | |
| 10:51:10 | sean-k-mooney | im just not sure if we use vcpupin to pin the cores to the numa node or cpuset | |
| 10:51:12 | artom | sean-k-mooney, right, I'm saying in those cases we also affine guest CPUs with cputune | |
| 10:51:19 | sean-k-mooney | ah ok | |
| 10:51:34 | sean-k-mooney | well as i said cpuset is just a shortcurt for vcpupin | |
| 10:51:43 | sean-k-mooney | let me give an example | |
| 10:52:28 | sean-k-mooney | those are the same | |
| 10:52:31 | sean-k-mooney | <vcpu placement='static' cpuset="1-4" current="1">2</vcpu> | |
| 10:52:33 | sean-k-mooney | <cputune> | |
| 10:52:35 | sean-k-mooney | <vcpupin vcpu="0" cpuset="1-4"/> | |
| 10:52:37 | sean-k-mooney | <vcpupin vcpu="1" cpuset="1,2,3,4"/> | |
| 10:52:39 | sean-k-mooney | <cputune> | |
| 10:52:54 | sean-k-mooney | libvirt internally generate the vcpupin element form the cpu set | |
| 10:53:04 | sean-k-mooney | and generge the same qemu commandline | |
| 10:53:12 | artom | Wait a second. | |
| 10:53:24 | artom | My *entire* code triggers if the instance has a NUMA topology | |
| 10:53:32 | sean-k-mooney | yes | |
| 10:53:53 | artom | So the insance with no NUMA topology on a host with vcpu_pin_set case will have to...? | |
| 10:54:12 | sean-k-mooney | you will need to add support for it | |
| 10:54:22 | sean-k-mooney | well you or stepen | |
| 10:54:30 | sean-k-mooney | technically this is a long standing issue | |
| 10:54:43 | sean-k-mooney | we always should have been updating the xml | |
| 10:54:54 | sean-k-mooney | it will be fixed via a hard reboot however | |
| 10:55:15 | sean-k-mooney | libvirt will reject the migration if the cores do not exist on the dest | |
| 10:55:25 | artom | Right, true | |
| 10:55:33 | artom | I'm surprised more people haven't been hitting that | |
| 10:55:54 | openstackgerrit | Brin Zhang proposed openstack/nova master: Follow-up: Add delete_on_termination to volume-attach API https://review.opendev.org/680655 | |
| 10:56:05 | sean-k-mooney | well the havent be casue we tell peole ot use the same cpu_set on all host in teh same aggreagte/az | |
| 10:56:11 | stephenfin | artom: You wouldn't see an issue if you had the same hardware and the same vcpu_pin_set on all hosts | |
| 10:56:15 | stephenfin | yeah | |
| 10:56:42 | aspiers | sean-k-mooney: I've implemented your suggestions from yesterday in https://review.opendev.org/#/c/680065/ and https://review.opendev.org/#/c/644565/ | |
| 10:56:48 | sean-k-mooney | and as i said since its fixed on a hard reboot sicne we dont claim anything and just regenreate it form the config | |
| 10:57:02 | sean-k-mooney | aspiers: did it fix the import loop? | |
| 10:57:06 | aspiers | sean-k-mooney: yes | |
| 10:57:10 | sean-k-mooney | :) | |
| 10:57:16 | stephenfin | sean-k-mooney: Has https://blueprints.launchpad.net/nova/+spec/image-metadata-prefiltering been punted to U? | |
| 10:57:17 | aspiers | sean-k-mooney: it worked very nicely as you should see from the comments I just posted | |
| 10:57:25 | stephenfin | If so, can I kick it out of the runway? | |
| 10:57:50 | sean-k-mooney | stephenfin: so i was suggesting since your code and aspiers keeps breaking it | |
| 10:57:56 | sean-k-mooney | and no one is revieing it | |
| 10:58:10 | stephenfin | I was, in fairness | |
| 10:58:21 | sean-k-mooney | that we could kick it out to get your stuff landed instead | |
| 10:58:22 | aspiers | sean-k-mooney: I'm hoping SEV is now done | |
| 10:58:50 | aspiers | all feedback should be now addressed, even stephenfin's request for extra API checks | |
| 10:58:51 | sean-k-mooney | stephenfin: well yes but you wer also reviing the sev stuff which keeps breaking it | |
| 10:59:00 | stephenfin | True | |
| 10:59:13 | sean-k-mooney | i had hoped that it would merge before either the cpu or sev stuff starte merging to avoid that | |
| 10:59:26 | sean-k-mooney | so im ok with punting ot U | |
| 10:59:32 | sean-k-mooney | we did it internally | |
| 10:59:42 | sean-k-mooney | or i can rebase and we can merge it quickly | |
| 11:00:07 | stephenfin | I was going to suggest rebasing on top of cpu-resources and I'd review | |
| 11:00:09 | stephenfin | since it's small | |
| 11:00:14 | stephenfin | but just on master could be fine too | |
| 11:00:28 | sean-k-mooney | erric alredy -2'd it since i said he could | |
| 11:00:32 | stephenfin | the merge conflicts should be slight | |
| 11:00:34 | sean-k-mooney | so ill kick it out of the run way | |
| 11:00:40 | sean-k-mooney | stephenfin: they are | |
| 11:01:01 | aspiers | stephenfin: the list of SEV patches got slightly bigger but don't be fooled, it's just 2 trivial extra refactorings plus the API checks you requested | |
| 11:02:01 | sean-k-mooney | stephenfin: focus on the sev series and your own | |
| 11:04:22 | aspiers | sean-k-mooney: hopefully there will be no more major changes to SEV to cause further merge conflicts | |
| 11:05:10 | sean-k-mooney | aspiers: sev conflict with the cpu series i think | |
| 11:05:29 | sean-k-mooney | but we should finish the sev seires as soon as we can i think | |
| 11:05:38 | sean-k-mooney | then push to land the cpu series | |
| 11:05:38 | aspiers | +1 for that ;-) | |
| 11:06:02 | aspiers | It's unfortunate that there was a Gerrit issue causing the current Zuul backlog | |
| 11:06:17 | aspiers | But hopefully it should catch up over the weekend | |
| 11:14:23 | bauzas | aspiers: gtk, then we should avoid asking for respins in case of nits | |
| 11:14:58 | aspiers | bauzas: yes, currently some jobs are in the check queue for >20 hours | |
| 11:16:51 | gibi | aspiers: I think those jobs stuck, as other, newly proposed patches went through | |
| 11:17:27 | aspiers | gibi: hrmm :-( I'll ask on #openstack-infra | |
| 11:19:00 | aspiers | gibi: Which ones do you see going through which are newer? I don't see them | |
| 11:19:07 | openstackgerrit | Brin Zhang proposed openstack/nova master: Add user_id and project_id colume to Migration https://review.opendev.org/673990 | |