| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-11-21 | |||
| 14:26:49 | mriedem | stephenfin: the core/ram/disk filters are no longer part of the default list of enabled filters b/c of placement | |
| 14:26:50 | bauzas | stephenfin: ralonsoh: so the problem is that we don't have a os-traits tag, right? | |
| 14:26:55 | stephenfin | (not saying it's something we want, but placement is not the reason not to want it) | |
| 14:27:03 | efried | stephenfin Agree. | |
| 14:27:08 | stephenfin | bauzas: Correct. We need to wait for the next release | |
| 14:27:11 | bauzas | stephenfin: ralonsoh: honestly, we could just create a new release | |
| 14:27:12 | bauzas | mriedem: ^ | |
| 14:27:18 | stephenfin | Also, I wouldn't want to block that patch while we wait for it | |
| 14:27:24 | jaypipes | stephenfin: eventually the *inputs* to the weighers will be created from the returned results of the GET /allocation_candidates call's provider_summaries response, instead of a constructing from the ComputeNode object/cell DB table. | |
| 14:27:26 | stephenfin | That's also an option :D | |
| 14:27:32 | bauzas | mriedem: context is https://review.openstack.org/#/c/466051/12..15/nova/virt/libvirt/config.py | |
| 14:27:43 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Enable cold migration with target host(1/2) https://review.openstack.org/408955 | |
| 14:28:01 | stephenfin | jaypipes: But filters would still be a nova (vs placement thing), right? | |
| 14:28:05 | bauzas | mriedem: I could just provide a new release for os-traits, fine with you ? | |
| 14:28:12 | stephenfin | Sorry - weighters | |
| 14:28:16 | mriedem | bauzas: sure | |
| 14:28:23 | ralonsoh | bauzas: you are right, os-traits is not updated. But the patch can work as is | |
| 14:28:24 | jaypipes | stephenfin: weighers will continue to be in nova-scheduler, yes | |
| 14:28:30 | mriedem | bauzas: you'll be working this week if the gate explodes and the rest of us are on vacation? | |
| 14:28:37 | bauzas | stephenfin: ralonsoh: I'll just provide a new release change for os-traits | |
| 14:28:54 | bauzas | mriedem: sure, I can | |
| 14:29:06 | jianghuaw | mriedem, both should work. But I think bauzas suggested MultiStrOpt when reviewing the spec. | |
| 14:29:07 | stephenfin | jaypipes, efried: If either of you want to review that once I do, I'd appreciate it https://review.openstack.org/#/c/379525/ | |
| 14:29:43 | jaypipes | stephenfin: k | |
| 14:30:14 | stephenfin | Cheers :) Seems I'm on the hook for nested RP reviews this week too, you'll be happy to hear | |
| 14:30:19 | mriedem | stephenfin: https://review.openstack.org/#/c/379525/ isn't happening for queens | |
| 14:30:20 | efried | stephenfin This is for a 3-year-old blueprint that's not approved for Q ? | |
| 14:30:23 | stephenfin | /unhappy | |
| 14:30:28 | bauzas | jianghuaw: I'd love to see you reviewing my inventory patch for the VGPU in libvirt | |
| 14:30:51 | stephenfin | Ahh, hold on so. I didn't realize there was a BP attached | |
| 14:30:55 | jianghuaw | bauzas, sure. I will. Actually I've opened that patch. | |
| 14:31:11 | stephenfin | In that case, I'll reopen and target for Rocky | |
| 14:31:22 | efried | SGTM | |
| 14:31:26 | bauzas | jianghuaw: I have multiple problems with libvirt mdevs, but that change would be okay | |
| 14:31:43 | mriedem | bauzas: any specific reason to use MultiStrOpt over ListOpt here? https://review.openstack.org/#/c/512580/ | |
| 14:31:53 | stephenfin | jianghuaw, bauzas, mriedem: MultiStrOpt is deprecated, afaik | |
| 14:31:58 | stephenfin | Or will be soon. Don't use it | |
| 14:32:03 | bauzas | mriedem: I don't remember me commenting that | |
| 14:32:19 | bauzas | lemme verify | |
| 14:32:26 | mriedem | stephenfin: it doesn't say that it is https://github.com/openstack/oslo.config/blob/master/oslo_config/cfg.py#L1571 | |
| 14:32:50 | bauzas | hem, no, I haven't said that https://review.openstack.org/#/c/512580/1/nova/conf/vgpu.py | |
| 14:33:15 | stephenfin | mriedem: Hmm, odd. I was sure gcb talked about that at the PTG | |
| 14:33:22 | mriedem | i'm asking in -oslo | |
| 14:33:42 | bauzas | mriedem: stephenfin: yeah, don't see it in https://docs.openstack.org/oslo.config/latest/reference/opts.html#oslo_config.cfg.MultiStrOpt | |
| 14:33:56 | bauzas | if that's deprecated, then please change that | |
| 14:34:09 | stephenfin | I'm pretty sure gcb did talk about it, so maybe it's on his list. In any case, I wouldn't use it. It's very confusing | |
| 14:34:21 | bauzas | honestly, I don't see the difference | |
| 14:34:24 | bauzas | but meh | |
| 14:34:28 | bauzas | I trust you | |
| 14:34:40 | bauzas | if you say ListOpt should be used, then -1 the change | |
| 14:34:42 | stephenfin | Every other option is specified once except that special one :) | |
| 14:34:56 | bauzas | AFAIK, you're an oslo core :p | |
| 14:35:13 | stephenfin | Heh, I hate -1ing for that kind of stuff, but ok | |
| 14:35:30 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: cleanup mapping/reqspec after archive instance https://review.openstack.org/515034 | |
| 14:36:04 | jianghuaw | bauzas, stephenfin, mriedem: Got it. So will proceed to change it from MultiStrOpt to ListOpts. Thanks. | |
| 14:36:10 | bauzas | ++ | |
| 14:36:21 | bauzas | I'll need to rebase my change, but meh | |
| 14:36:32 | stephenfin | ++ as commented, those things are hard to change after the fact | |
| 14:36:37 | bauzas | jianghuaw: just ping me when it's uploaded so I can rebase my series | |
| 14:36:49 | stephenfin | bauzas: I'm a pbr core - not oslo (yet ;)) | |
| 14:36:50 | jianghuaw | bauzas, Sure. | |
| 14:37:03 | bauzas | :p | |
| 14:37:03 | bauzas | stephenfin: WAT? | |
| 14:37:10 | stephenfin | Probably for the better 🙈 | |
| 14:37:48 | bauzas | ralonsoh: FWIW, working on a new release for os-traits | |
| 14:38:30 | ralonsoh | bauzas: thank you very much | |
| 14:46:42 | openstackgerrit | Jianghua Wang proposed openstack/nova master: vgpu: add enabled white list https://review.openstack.org/512580 | |
| 14:46:43 | openstackgerrit | Jianghua Wang proposed openstack/nova master: XenAPI: provide vGPU inventory in compute node https://review.openstack.org/516217 | |
| 14:46:43 | openstackgerrit | Jianghua Wang proposed openstack/nova master: XenAPI: get vGPU stats from hypervisor https://review.openstack.org/512965 | |
| 14:46:44 | openstackgerrit | Jianghua Wang proposed openstack/nova master: XenAPI: restrict to only enable one vGPU type https://review.openstack.org/521769 | |
| 14:46:44 | openstackgerrit | Jianghua Wang proposed openstack/nova master: XenAPI: create vGPU for instance https://review.openstack.org/516899 | |
| 14:46:58 | jianghuaw | bauzas, ^ | |
| 14:48:01 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add CPUWeigher https://review.openstack.org/379525 | |
| 14:48:37 | bauzas | jianghuaw: roger, will rebase | |
| 14:49:06 | jianghuaw | bauzas, cool. | |
| 14:49:25 | moshele | stephenfin: hi can you review https://review.openstack.org/#/c/519066/ ? | |
| 14:49:40 | stephenfin | moshele: Big backlog today, but it's on the list :) | |
| 14:50:17 | moshele | stephenfin: so tomorrow :) | |
| 14:50:38 | stephenfin | Depends on how quickly I work through the backlog, heh | |
| 14:50:56 | moshele | stephenfin: :) | |
| 14:51:03 | openstackgerrit | Balazs Gibizer proposed openstack/nova stable/pike: Document the real behavior of notify_on_state_change https://review.openstack.org/521870 | |
| 14:51:52 | openstackgerrit | Dan Smith proposed openstack/nova master: Alternate fix for validating image on rebuild https://review.openstack.org/521186 | |
| 14:51:53 | openstackgerrit | Dan Smith proposed openstack/nova master: Fix doubling allocations on rebuild https://review.openstack.org/521662 | |
| 14:51:53 | openstackgerrit | Dan Smith proposed openstack/nova master: Add regression test for rebuild with new image doubling allocations https://review.openstack.org/521153 | |
| 14:52:15 | dansmith | mriedem: you wanna write the reno for ^? I'm struggling to formulate english sentences at the moment | |
| 14:52:22 | bauzas | mriedem: ralonsoh: stephenfin: https://review.openstack.org/#/c/521869/ | |
| 14:52:34 | bauzas | mriedem: need your lovely +1 | |
| 14:53:01 | mriedem | dansmith: sure | |
| 14:53:26 | stephenfin | bauzas: Do we want to merge the requirements updates first | |
| 14:53:30 | stephenfin | https://review.openstack.org/#/q/project:openstack/os-traits+status:open | |
| 14:53:30 | stephenfin | ? | |
| 14:53:44 | stephenfin | The rest are WIP | |
| 14:53:46 | bauzas | stephenfin: I looked at those | |
| 14:53:53 | openstackgerrit | Balazs Gibizer proposed openstack/nova stable/ocata: Document the real behavior of notify_on_state_change https://review.openstack.org/521871 | |
| 14:54:03 | bauzas | stephenfin: and honestly, I don't see why we need to block a new release because of those | |
| 14:54:14 | stephenfin | Sweet. Then you have your +1 | |
| 14:54:25 | bauzas | stephenfin: most of cores will be on vacations by the next days | |
| 14:54:51 | bauzas | so if we wanna merge a new release and modify g-r, then it needs to be soon | |
| 14:56:04 | bauzas | mriedem: FWIW, I'm not sure we need to bump g-r given nova reqs.txt is os-traits>=0.3.2 # Apache-2.0 | |
| 14:56:08 | bauzas | you okay with that ? | |