Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-09-29
15:31:05 sahid sean-k-mooney: yes really, since the beginning the point was that, vgpus can be exposed with sriov or mdev
15:31:40 cdent sean-k-mooney: while you’re around, can you have a look at https://review.openstack.org/#/c/504540/ you were one of the people who had some ideas on how to do allocation candidate limiting, and there now lots of ideas on that spec, but some confusion on what things we are trying to optimize
15:31:55 sean-k-mooney cdent: sure opening it now
15:32:00 cdent thanks
15:32:07 bauzas leakypipes: sean-k-mooney: that's one of the reasons why the vGPU spec is saying we will only pass GPU resource classes
15:32:17 bauzas if libvirt uses mdev, then meh
15:32:27 bauzas if xen is using anything else, then meh
15:32:39 sean-k-mooney bauzas: its nto a libvirt vs xen thing
15:32:41 bauzas we shouldn't just leak out the technical details
15:32:42 sahid it's not libvirt who is using mdev, it's the hardware driver...
15:33:04 sean-k-mooney at the hardware level amd uses hardware based partitioning of the gpu via sriov
15:33:12 bauzas whatever the solution is, the interface with the compute manager is just a resource class
15:33:17 sean-k-mooney intel and nvidia do it via the diriver with mdevs
15:33:48 bauzas for the moment, libvirt supports virtual GPUs by mdevs but that's just a temporary thing AFAICU
15:33:51 andreykurilin mriedem: I agree that it is difficult to compare the results from different hardware(which we have in infra), but it is not a good time for openstack and I do not know the company who are ready to give reserved hardware for a performance ci
15:34:15 bauzas sean-k-mooney: yeah that's the driver which provides mdev types, libvirt just gets them
15:34:17 sean-k-mooney bauzas: the hypervisror did nto need modifcation to work with vgpus with sriov
15:34:51 bauzas sean-k-mooney: not sure I get your point :)
15:35:11 sean-k-mooney my point is that libvirt works with amd vgpus also
15:35:27 sean-k-mooney you just to a pci passthough of the sriov vf
15:35:47 bauzas ah okay, and amd driver uses SR-IOV for that?
15:35:57 bauzas the AMD driver isn't using mdevs ?
15:36:05 sean-k-mooney yes they do the virtualisation in hardware
15:36:09 bauzas I see
15:36:11 sean-k-mooney they do not use mdevs
15:36:11 bauzas so, the thing is
15:36:20 bauzas I see baby steps here
15:36:38 bauzas step 1/ make sure we have something workable without adding more tech deby
15:37:07 bauzas step 2/ work on the generic thingy from efried and "do the work" (c)
15:37:37 bauzas step 3/ migrate how we track PCI devices from the old world to the new world
15:38:02 fried_rice ++
15:38:12 bauzas hah, Friday!
15:38:28 bauzas holy shit, I said many times I need a ZNC bot for me
15:38:40 mriedem andreykurilin: http://forumtopics.openstack.org/cfp/details/55
15:38:43 mriedem superdan: melwitt: ^
15:39:23 melwitt cool
15:40:27 sean-k-mooney bauwser: makes sense, my concern is coming for interop side in that i would hope at the api level we will not need to expose if it is an mdev or sriov based vgpu
15:40:51 bauwser sean-k-mooney: that's where I'm pretty firm in my mind
15:41:02 bauwser sean-k-mooney: outside the virt driver, we shouldn't leak out the details
15:41:23 sean-k-mooney bauwser: we may need to expose a vender in some form for guest driver reasons but other then that how we virtualise should be internal to nova
15:41:26 bauwser the virt driver will just report inventories
15:41:46 bauwser how those inventories will be populated will be different based on the driver
15:41:59 bauwser but that will be consistent
15:42:17 openstackgerrit Eric Fried proposed openstack/nova master: WIP: Use ksa adapter for cinder client https://review.openstack.org/508345
15:42:41 sean-k-mooney sure and i think we can use traits on the provider to hanel the "i only have a nvida driver in my vm image" problem
15:42:44 fried_rice mordred ^ would you mind checking that I did my version discovery sanely here?
15:43:03 bauwser so that a flavor asking for an amount of 2 of the VGPU resource class and a trait of 'K800' will get 2 K800 vGPUs
15:43:21 bauwser that's where the API is
15:43:46 bauwser actually a good call for trying to fit the fried_rice's model with vGPUs
15:44:19 sean-k-mooney bauwser: perhaps yes or as image metadata where i say i can support x,y,z and the and flavor say i want y and we compare to make sure its compatible
15:45:10 bauwser sean-k-mooney: mmm
15:45:26 bauwser sean-k-mooney: I wonder if that wouldn't be just a filter for the image metadata
15:45:47 bauwser like, I can find all the hosts that can support y (if y is asked by the flavor)
15:45:48 sean-k-mooney im thinking we can add traits to the image metadata
15:45:52 bauwser that is a placement thing
15:46:14 bauwser well
15:46:25 bauwser honestly that's a thought
15:47:22 mordred fried_rice: looks good on quick glance - but I haven't ooked deeply yet
15:47:37 fried_rice mordred Cool, thanks.
15:48:22 bauwser well, I need to call it a day, actually
15:48:48 mriedem fried_rice: let's let mordred deal with the infra fire that's raging
15:48:52 mriedem :)
15:49:19 fried_rice mriedem mordred Sorry, didn't smell the smoke.
15:49:27 mriedem #openstack-infra
15:49:42 sean-k-mooney leakypipes: am qq. is there a way we can retrive the compute node resocue provide by the hostid? and if not would you be against adding one. it could just be a trait
15:50:33 leakypipes sean-k-mooney: each resource provider has a name attribute.
15:50:41 leakypipes sean-k-mooney: that is set to hypervisor_hostname value
15:50:49 leakypipes sean-k-mooney: so yeah, you can already do that.
15:50:55 leakypipes sean-k-mooney: and that name is unique BTW
15:51:24 sean-k-mooney leakypipes: yes i know we require that all hostnames be unique in the openstack cloud
15:51:47 sean-k-mooney leakypipes: the issue is the get resouce provider api to day i think need you to pass the uuid
15:51:50 mriedem sean-k-mooney: you can find the compute node by hostname, and from the compute node uuid look up the RP
15:51:53 mriedem that's what our functional tests do
15:51:58 sean-k-mooney i dont know if we have a rest api call to get it by name
15:52:06 leakypipes sean-k-mooney: no, I mean when resource providers are added by the compute node, the resource provider's name attribute is set to the hypervisor_hostname
15:52:32 sean-k-mooney fried_rice: cool if GET /resource_providers?name={hostname} works that perferct
15:52:34 leakypipes sean-k-mooney: GET /resource_providers?name=$hypervisor_hostname
15:52:43 leakypipes sean-k-mooney: sorry, yeah, fried_rice beat me to it.
15:52:59 fried_rice Yeah, to be technically correct, in case $hypervisor_hostname isn't the same as `hostname`
15:52:59 leakypipes he's a wily coyote, that fried_rice
15:53:16 leakypipes fried_rice: which is the case for Ironic nodes :)
15:53:27 fried_rice right
15:53:39 sean-k-mooney the reason i was asking is for the neutron placement work. when nested provider are added we will need to find the computenode provider for the current host hence question
15:54:17 fried_rice So as long as you know how virt is reporting the hypervisor hostname, you're golden.
15:54:54 sean-k-mooney in neutron i belive we have the host id in a config that must be set to match the one used by nova
15:55:23 leakypipes right
15:55:32 fried_rice # Hostname to be used by the Neutron server, agents and services running on
15:55:32 fried_rice # this machine. All the agents and services running on this machine must use
15:55:32 fried_rice # the same host value. (unknown value)
15:55:32 fried_rice #host = example.domain
15:55:35 fried_rice in neutron.conf ?
15:55:39 leakypipes device_id == hostname, right?
15:55:42 leakypipes sean-k-mooney: ^\
15:55:59 sean-k-mooney device_id?
15:56:41 leakypipes sean-k-mooney: in neutron port response...
15:57:35 leakypipes sean-k-mooney: never mind.
15:57:36 sean-k-mooney oh binding:host_id is
15:57:40 mriedem superdan: http://forumtopics.openstack.org/cfp/details/60
15:57:45 mriedem melwitt: tssurya: ^
15:59:06 superdan mriedem: hmm, did the gate just get dumped?

Earlier   Later