Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-08-25
07:15:37 bauzas gibi: good idea, I'll jump onto reviewing those patches
07:45:27 brinzhang_ gibi, bauzas: optimiza create and bind arqs logical patch, need to review https://review.opendev.org/#/c/726564/6
07:46:16 gibi brinzhang_: hi! on my list for today
07:46:19 brinzhang_ gibi, bauzas: the cyborg evacuate support patch rebased on this patch, and it merge conflict, we will resolve the merge conflict
07:46:34 brinzhang_ gibi: cool, thanks
08:06:45 openstackgerrit Brin Zhang proposed openstack/nova master: Optimize _create_and_bind_arqs logic in conductor https://review.opendev.org/726564
08:06:46 openstackgerrit Brin Zhang proposed openstack/nova master: Cyborg evacuate support https://review.opendev.org/715326
09:03:49 swp20 gibi: Hi, gibi. https://review.opendev.org/#/c/719163/16/nova/scheduler/client/report.py@L2168. can u give some idea to process these exceptions?
09:04:42 xinranwang__ gibi: sean-k-mooney: Hi all, as we discussed during virtual PTG, I have proposed an overall spec https://review.opendev.org/#/c/742785/ to discuss sriov nic support in nova, neutron, and cyborg. Hope you guys can review it and give comments on it, there still some open questions need to be nailed down which really need you guys' comments and opinion. Let's take this spec as a place to discuss it. Please
09:04:43 xinranwang__ review it when you guys got time. Thanks a lot!
09:07:32 gibi swp20: I will check that comment from efried and try to recover the context
09:07:53 swp20 Thanks gibi.
09:10:03 gibi xinranwang__: you need to re-propose that spec to wallaby as we are past victory spec freeze. I cannot promise I will have time in the coming weeks to dig deep into this as the victoria feature freeze is coming fast. But please add this as a topic for the PTG so we won't forget about it
09:10:18 gibi xinranwang__: https://etherpad.opendev.org/p/nova-wallaby-ptg
09:17:30 xinranwang__ gibi: Understood, sure. I will re-propose it.
09:20:28 xinranwang__ gibi: I'd like to point it out here because it is better that we can do some reviews before next ptg. No rush, maybe after the feature freeze in V.
09:21:05 gibi xinranwang__: sure, I will try to prepare myself for the PTG by reading up on the possible PTG topics
09:23:58 xinranwang__ gibi: great, thanks
09:41:52 Yumeng hi gibi and sean-k-mooney: just one more question for the smartnic spec, would you like xinranwang__ and I to split the discussions before next PTG and start discuss(we introduce one or two of the open issues first, and then once you guys have any questions, we can answer and discuss.) the open issues directly one by one at your convenience? so
09:41:52 Yumeng that we don't need to think and decide all of them at a very short time in PTG.
09:43:24 openstackgerrit Stephen Finucane proposed openstack/nova master: functional: Enable real policy fixture by default https://review.opendev.org/745906
09:43:25 openstackgerrit Stephen Finucane proposed openstack/nova master: functional: Add and use 'GlanceFixture' https://review.opendev.org/745907
09:43:25 openstackgerrit Stephen Finucane proposed openstack/nova master: tests: Remove '_FakeImageService' https://review.opendev.org/745908
09:45:51 stephenfin gmann: Addressed your comments on that "use real policy" patch ^ Let me know if I missed anything :)
09:58:19 itandops Hi folks, I have some issue to restart nova API after rebooting the controller http://paste.openstack.org/show/797116/
09:58:51 itandops There is a database with empty values nova_cell0
09:59:37 itandops It seems to be the root cause of this issue. Any suggestion to solve this ?
10:03:35 lyarwood itandops: `Lost connection to MySQL server during query` it isn't that the db is empty
10:06:03 itandops lyarwood: I used 2 controllers running nova API.
10:07:16 lyarwood itandops: okay well there's a clear db connectivity issue there for you to debug now :)
10:15:44 sean-k-mooney gibi are you going to try and resolve the sriov attach/detach feature before m3
10:16:50 sean-k-mooney gibi: im still trying to find time to work on numa in placement but im fully expecting that to slip to next cycle. but im going to try and at least have a poc of it before repoposing the spec next cycle
10:19:01 gibi sean-k-mooney: I will try to push that feature later this week but I think at some point I have to accept that it will slip
10:19:54 gibi Yumeng: if there is a clear list of open issues in the spec then that is enough
10:20:53 sean-k-mooney ok, do you know what the status of provider.yaml is? https://review.opendev.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/provider-config-file
10:22:00 gibi sean-k-mooney: I will go back to provider.yaml today, just finished reviewing the cyborg evac support. (going through the runway items)
10:22:35 gibi so I can tell more in couple of hours
10:23:47 sean-k-mooney ok it looks like the first patch merged and the second has a +2 from alex the 3rd or 4th patch makes the feature functional
10:24:27 sean-k-mooney so that look like its making progress ill try to review the runway items but maybe not today.
10:25:17 jsuchome ok gibi, lyarwood, https://review.opendev.org/#/c/574301/ is green again, chance for new reviews?
10:25:58 gibi jsuchome: ohh, I was +2 on that before, so I hope I can make a quick rereview of that now
10:27:00 lyarwood jsuchome: ack looking
10:27:27 lyarwood gibi: I went ahead and moved a few things around with that, there's another change underneath from me now moving rbd_utils out of nova.virt.libvirt for example
10:27:45 lyarwood gibi: and this change also introduces some generic configurables now
10:27:58 gibi lyarwood: yeah, I just opened your change below
10:28:48 lyarwood jsuchome: https://review.opendev.org/#/c/728095/ can you update the releasenote to include the new configurables please
10:31:09 lyarwood bauzas: https://review.opendev.org/#/q/I32908b77c18f8ec08211dd67be49bbf903611c34 - would you mind hitting these today?
10:31:24 bauzas lyarwood: I can take a look
10:32:30 lyarwood thanks
10:33:03 lyarwood https://review.opendev.org/#/q/I7eb86edc130d186a66c04b229d46347ec5c0b625 is the other change that's part of this bugfix
10:49:10 openstackgerrit Luyao Zhong proposed openstack/nova master: Track error migrations in resource tracker https://review.opendev.org/745281
10:49:11 openstackgerrit Luyao Zhong proposed openstack/nova master: Improve orphans tracking in resource tracker https://review.opendev.org/745282
10:49:11 openstackgerrit Luyao Zhong proposed openstack/nova master: Rollback when reset running live migration https://review.opendev.org/747899
11:00:52 openstackgerrit Luyao Zhong proposed openstack/nova master: Rollback when reset running live migration https://review.opendev.org/747899
11:03:22 luyao stephenfin, alex_xu: Thank you for your comments. if you get time, vpmem enhancement is ready for new round review. https://review.opendev.org/#/c/714653/
11:09:41 openstackgerrit Jiri Suchomel proposed openstack/nova master: Bring back allowed_direct_url_schemes in support of RBD image download https://review.opendev.org/728095
11:21:05 sean-k-mooney it would be really nice if you could resize a shelved instance
11:23:17 gibi sean-k-mooney: you can automat that on the client side :) unshelve; resize; shelve
11:23:23 sean-k-mooney im going to file a bug for numa blancing later today and work on a patch to resolve it later in the week
11:23:35 sean-k-mooney gibi: thats what im doing with horizon manually
11:23:46 sean-k-mooney while i finished reading my email
11:23:54 sean-k-mooney which is now alos done
11:24:29 sean-k-mooney gibi: it just feels like something that would be easy to do in the shevled offloaded state as its just a db update
11:24:51 sean-k-mooney resize while shelve offloaded that is
11:25:04 gibi hm, that makes more sense yes
11:25:12 sean-k-mooney but thats an api change and needs a spec so not this cycle
11:41:47 gibi swp20: left some comments in https://review.opendev.org/#/c/719163/18/nova/scheduler/client/report.py@2170
11:42:51 openstackgerrit Balazs Gibizer proposed openstack/nova master: Undeprecate the vmwareapi driver https://review.opendev.org/742407
11:45:52 sean-k-mooney gibi: i know vmware jobs are now working again but do they trigger automatically or is it only when the recheck comment is left
11:46:15 gibi sean-k-mooney: it is triggered automatically
11:46:29 gibi I just pushed a new ps and it will vote on it
11:46:41 sean-k-mooney ok cool
11:46:59 sean-k-mooney initally it looked like it was manual but i guess they made it automatic over the cycle
11:47:42 sean-k-mooney if its still it seams to be relitivly stable so if they continue to maintain it then un deprecating it makes sense.
11:48:17 gibi looking at the recent history it of the job I think it is fairly stable. I also see some patches targeting the vmware driver so there is real maintenance on the code now
11:48:45 gibi e.g. https://review.opendev.org/#/c/742398/
11:49:30 sean-k-mooney ya
11:49:54 sean-k-mooney reading the commit message for your change it makes sense to me
11:59:48 openstackgerrit sean mooney proposed openstack/nova master: libvirt: delegate ovs plug to os-vif https://review.opendev.org/602432
12:19:33 lyarwood stephenfin: https://review.opendev.org/#/q/topic:bp/nova-image-download-via-rbd any chance you could add this to your queue, I'll get to the vTPM stuff tomorrow if that's okay
12:28:55 gibi stephenfin, sean-k-mooney: could you confirm that the "any number" here is a doc bug? https://github.com/openstack/nova/blame/master/doc/source/user/flavors.rst#L549-L551 I belive if hw:numa_nodes is not specified then it is defaulted to 1
12:36:25 jsuchome gibi, lyarwood so who could we ask for another +2?
12:36:50 gibi jsuchome: I think lyarwood just asked stephenfin above
12:38:48 lyarwood yup I did
12:39:46 sean-k-mooney gibi: it depends on the context but ill check now
12:40:44 sean-k-mooney gibi: the docs are correct
12:41:17 gibi sean-k-mooney: they why I remember that nova does not automatically split resources between host NUMA nodes?
12:41:27 gibi for a single instance
12:41:46 sean-k-mooney gibi: it does not but it does not crete a numa toplogy by default either
12:42:15 sean-k-mooney hw:numa_cpus and hw:numa_mem are only used if hw:numa_nodes>=2
12:42:40 gibi OK, so if nothing numa related are requested then we split resource between host numa nodes
12:42:43 sean-k-mooney and by default vms without a numa toplogy float over all host numa nodes
12:42:57 sean-k-mooney gibi: no we dont split them
12:42:58 gibi but if we request huge pages then a numa topology is created
12:43:11 sean-k-mooney we simply do not limit them to a specific numa node
12:43:15 gibi and then numa_nodes defaults to 1 if not specified
12:43:18 sean-k-mooney the kernel will split and move tehm as it sees fit
12:43:25 sean-k-mooney yes
12:43:44 sean-k-mooney if you request hugepage or cpu pinning or pmem we default to 1 numa node
12:43:50 gibi OK, cool

Earlier   Later