| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-03-27 | |||
| 14:00:59 | bauzas | except the fact we don't use traits atm | |
| 14:01:29 | bauzas | anyway, will just propose something soon so we could see if that requires some spec | |
| 14:01:43 | bauzas | or at least some spec amendment | |
| 14:01:50 | efried | bauzas: Right, so you're setting up the types & inventories the first time you load up. The first time update_provider_tree runs, it reads that config file and sets up the provider tree with types & inventories according to what the admin put in there. Right? | |
| 14:01:59 | bauzas | ++ | |
| 14:02:12 | efried | Cool cool | |
| 14:02:16 | alex_xu_ | the race condidates only happened one time, if people accept that, it also ok | |
| 14:02:20 | bauzas | efried: https://specs.openstack.org/openstack/nova-specs/specs/queens/implemented/add-support-for-vgpu.html describes that partially | |
| 14:02:34 | jaypipes | guh, it would have been nice to have more than a few hours of sleep before spec sprint day... :( | |
| 14:02:41 | efried | bauzas: I can also see a solution that goes halfway in the future... | |
| 14:02:55 | bauzas | jaypipes: take caffeine or drugs, either. | |
| 14:03:32 | bauzas | jaypipes: FWIW, https://review.openstack.org/#/c/541290/7/specs/rocky/approved/numa-aware-vswitches.rst will require me some paracetamol | |
| 14:03:36 | bauzas | stephenfin: ^ ;) | |
| 14:03:50 | cdent | jaypipes: start on an easy one: https://review.openstack.org/#/c/418393/ | |
| 14:03:52 | edleafe | jaypipes: have you tried the blue meth? | |
| 14:04:02 | bauzas | always take the blue pill | |
| 14:04:04 | stephenfin | I hear the blue meth is real good | |
| 14:04:13 | bauzas | never choose the red one | |
| 14:04:35 | efried | bauzas: In a fashion similar to neutron port creation, the user could create the VGPU resource before doing the boot request. The VGPU creation would eventually hit vendor-specific code (maybe virt, maybe some agent - cyborg?) which actually *would* tweak the inventories and available types. Then you would attach that VGPU to your boot request, just like you would a port. | |
| 14:05:03 | alex_xu_ | efried: any legal drugs in Florida? | |
| 14:05:26 | efried | alex_xu_: Legal schmegal. But I'm in Texas. | |
| 14:05:30 | bauzas | efried: creating the mediated devices in advance is already a possibility | |
| 14:05:32 | efried | In Florida, they wash up on the beach. | |
| 14:08:46 | kashyap | alex_xu_: Hey there | |
| 14:09:12 | kashyap | alex_xu_: About your comment here: https://review.openstack.org/#/c/534384/16/nova/tests/unit/virt/libvirt/test_config.py | |
| 14:09:21 | efried | bauzas: But having that creation impact placement inventories would be new. | |
| 14:09:34 | bauzas | efried: no, it's already the case | |
| 14:09:42 | kashyap | alex_xu_: The test you pointed out is slightly different in that, it is testing: 'obj.extra_flags' | |
| 14:09:43 | efried | oh? | |
| 14:09:57 | bauzas | efried: if you create a mediated device, libvirt will just add that to the total | |
| 14:10:11 | bauzas | I should write a blogpost on this... | |
| 14:10:34 | bauzas | because creating a mediated device doesn't mean you *use* it for a guesrt | |
| 14:10:53 | bauzas | so, placement is getting an inventory of 'available+created' as a total already | |
| 14:11:20 | efried | cool | |
| 14:12:11 | jaypipes | stephenfin: and ack on the OVS documentation being awful. | |
| 14:12:19 | jaypipes | OVS-DPDK that is. | |
| 14:14:57 | dansmith | efried: mriedem: I can never remember the moving target of translations.. we're translating exceptions now, but not translating logs _at all_ (even warning) ? | |
| 14:15:26 | bauzas | dansmith: right IIRC | |
| 14:15:28 | efried | dansmith: correct | |
| 14:15:49 | bauzas | because translating logs is too much for the i18n team | |
| 14:15:57 | bauzas | they aren't able to scale | |
| 14:16:24 | efried | Nice patch for a new contributor: remove _Lx from nova.i18n, see what breaks, fix it. | |
| 14:17:36 | dansmith | wtfever | |
| 14:18:07 | dansmith | not marking them means they _can't_ be translated which seems like a stupid step back, but whatever | |
| 14:21:09 | openstackgerrit | Patricia Domingues proposed openstack/nova master: load up the volume drivers by checking architecture https://review.openstack.org/541393 | |
| 14:23:50 | jaypipes | efried: "The first time update_provider_tree runs, it reads that config file and sets up the provider tree with types & inventories according to what the admin put in there. Right?" <-- you mean, exactly like my provider-config-file spec enabled? | |
| 14:24:56 | efried | jaypipes: could be, could be. I'm not sure if I read that before it was declared moot, but if I did, I've purged it :( | |
| 14:25:47 | efried | jaypipes: Anyway, it sounds like a lovely idea you had :) | |
| 14:26:06 | jaypipes | efried, bauzas: I'm the person that would shit a brick if you start adding dynamic inventory creation based on whatever a user requested. | |
| 14:26:34 | efried | Yes, but not the only person. | |
| 14:26:52 | bauzas | jaypipes: I'm entering the danger timezone of me needing to go get my children at school | |
| 14:27:15 | bauzas | jaypipes: tl;dr I just want the inventory for that specific physical GPU to be config-driven | |
| 14:27:31 | bauzas | jaypipes: if that sounds crazy, tell me more in 20 mins :) | |
| 14:27:35 | jaypipes | bauzas: yet another whitelist conf option, yes, I know. | |
| 14:28:15 | openstackgerrit | Dan Smith proposed openstack/nova master: Make get_allocation_candidates() honor aggregate restrictions https://review.openstack.org/547990 | |
| 14:28:16 | openstackgerrit | Dan Smith proposed openstack/nova master: Add an index on aggregate_metadata.value https://review.openstack.org/555851 | |
| 14:28:16 | openstackgerrit | Dan Smith proposed openstack/nova master: Add AggregateList.get_by_metadata() query method https://review.openstack.org/544728 | |
| 14:28:17 | openstackgerrit | Dan Smith proposed openstack/nova master: Add require_tenant_aggregate request filter https://review.openstack.org/545002 | |
| 14:28:17 | openstackgerrit | Dan Smith proposed openstack/nova master: WIP: Honor availability_zone hint via placement https://review.openstack.org/546282 | |
| 14:29:12 | sahid | jaypipes: if you can enqueue this https://review.openstack.org/#/c/511188/ :) | |
| 14:29:44 | mriedem | efried: can we abandon https://review.openstack.org/#/c/497978/ or do you plan on updating it? | |
| 14:30:17 | efried | mriedem: Eventually. But I can abandon for now and resurrect at that time. | |
| 14:30:35 | mriedem | ok. i'm just starting backward for specs, oldest to newest and cleaning out the cruft. | |
| 14:30:56 | openstackgerrit | Balazs Gibizer proposed openstack/nova-specs master: Network bandwidth resource provider https://review.openstack.org/502306 | |
| 14:31:29 | mriedem | johnthetubaguy: should we abandon https://review.openstack.org/#/c/438134/ for service-protected-servers? does that align with anything keystone is working on with RBAC? | |
| 14:31:32 | mriedem | edmondsw: ^ | |
| 14:32:35 | edmondsw | in a mtg, will look in a few | |
| 14:32:44 | mriedem | dansmith: this was the thing that prompted my ML thread about volume type proxy https://review.openstack.org/#/c/466595/ | |
| 14:32:53 | mriedem | i think... | |
| 14:33:22 | dansmith | ack | |
| 14:35:47 | mriedem | bauzas: i'm going to start an ops list thread about https://review.openstack.org/#/c/446446/ since if it's just a bug fix for broken behavior, we don't need a spec or a microversion | |
| 14:37:29 | bhagyashris | mriedem, alex_xu_: Hi, I have proposed revised (as per the discussion in Dublin PTG) spec: https://review.openstack.org/#/c/511825/2 ( | |
| 14:37:29 | bhagyashris | Disallow rotation parameter 0 for 'createBackup' API) Request you to review the same. | |
| 14:37:54 | mriedem | bhagyashris: cool i'll review today | |
| 14:38:25 | alex_xu_ | kashyap: strange...I didn't see there is extra_flags in LibvirtConfigGuestCPU obj | |
| 14:38:39 | alex_xu_ | bhagyashris: cool, will try to reach that | |
| 14:38:45 | kashyap | alex_xu_: But, thanks to your comment, I could actually remove another line in the test | |
| 14:39:01 | alex_xu_ | kashyap: np | |
| 14:39:14 | kashyap | alex_xu_: I'll upload a new version once I re-run my local tests. | |
| 14:39:21 | kashyap | alex_xu_: As I caught another bug from a functional test. | |
| 14:39:48 | kashyap | The casing of the config options was lost, after I moved to Oslo StrOpt(). Glad I tested locally w/ both casings | |
| 14:39:54 | kashyap | (Now fixed it) | |
| 14:41:14 | jaypipes | gibi: "The backend information is needed for the NeutronBackendWeigher that tries to simulate the Neutron backend selection mechanism by implementing a preference order between backends." <-- this is what concerns me about the introduction of those NET_BACKEND_XXX traits. I don't see those traits as being germane to scheduling. Rather, I just see them as being a part of the port configuration information that we send down to os-vif. And we have | |
| 14:41:14 | jaypipes | the port binding information in Neutron for that. I still don't see why those should be traits. | |
| 14:42:40 | alex_xu_ | kashyap: actually I mean I didn't find a extra_flags field for the LibvirtConfigGuestCPU even for now... | |
| 14:42:55 | gibi | jaypipes: the backend traits are not needed for the placement query or for the scheduler filters | |
| 14:43:04 | efried | jaypipes: Do you have a spec queued up for NRP-in-a_c yet? | |
| 14:43:31 | gibi | jaypipes: but as the filter scheduler makes the allocation, it implicitly decides about the backend as well | |
| 14:43:49 | gibi | jaypipes: and before the bandwidth feature this decision was made by neutron | |
| 14:44:07 | gibi | jaypipes: but after it, the decision is made by the filter scheduler | |
| 14:44:11 | bhagyashris | mriedem, alex_xu_: thank you :) | |
| 14:44:23 | gibi | jaypipes: so we want to save some of the freedom of neutron here | |
| 14:44:25 | kashyap | alex_xu_: It's not in that class. But take a look at LibvirtConfigCPUFeature() | |
| 14:44:38 | gibi | jaypipes: by adding a weigher that can express backend preference order | |
| 14:45:28 | kashyap | alex_xu_: Typo, actually this one: LibvirtConfigGuestCPUFeature() | |
| 14:45:39 | jaypipes | efried: crap. haven't finished it. | |
| 14:45:44 | jaypipes | efried: I can push what I have. | |
| 14:46:08 | efried | jaypipes: as you see fit. Just thought I'd ask, it being spec review day and all. | |
| 14:47:06 | openstackgerrit | Matthew Booth proposed openstack/nova-specs master: Add serial numbers for local disks https://review.openstack.org/556565 | |
| 14:47:10 | jaypipes | efried: ack, thx for the reminder. | |