Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2018-05-15
15:29:35 dansmith they also seem to focus on "graphics acceleration", which would be in line with being OpenGL-only
15:29:46 openstackgerrit Vladyslav Drok proposed openstack/nova-specs master: Allow having placement inventories with reserved value equal to total https://review.openstack.org/568613
15:30:18 dansmith jaypipes: ah, they say use P3 for GPGPU
15:30:39 dansmith jaypipes: https://aws.amazon.com/ec2/instance-types/p3/
15:30:42 dansmith jaypipes: tesla v100 only
15:30:46 dansmith jaypipes: so, yeah, easy :)
15:32:12 dansmith and note they're not saying "you'll get some nvidia somethingorother" they say .. "it'll be a damned v100, alwas"
15:32:13 jaypipes dansmith: I see them say tesla v100. those are passthrough direct GPU devices, though, not vGPU "types", right?
15:32:35 jaypipes "Alternatively, you can also use the NVIDIA AMI with GPU driver and CUDA toolkit pre-installed." ... interesting.
15:32:52 dansmith jaypipes: yeah I dunno if it's a vGPU or them just giving you a whole card
15:33:04 jaypipes dansmith: it's the whole card. or 8 of them :)
15:33:32 dansmith doesn't have anything to do with our discussion though right? they're very specific about what you get, and thus what driver you need
15:35:08 jaypipes dansmith: yes and no? :) the "elastic GPUs" are very much non-vendor-specific. The P3 instances are clearly very much "here is this specific machine configuration with 8 NVIDIA GPUs", etc.
15:35:24 jaypipes dansmith: but I see your point.
15:35:38 jaypipes dansmith: we need to support both cloudy and non-cloudy things.
15:35:52 dansmith jaypipes: but they're not for GPGPU, they're for graphics acceleration
15:36:01 jaypipes dansmith: I've just been trying my best to keep the vendor-specific stuff out of the way.
15:36:05 dansmith and they virtualize the OpenGL interface, so you're far removed
15:36:09 jaypipes dansmith: yes, I understand.
15:38:02 openstackgerrit Vladyslav Drok proposed openstack/nova-specs master: Allow having placement inventories with reserved value equal to total https://review.openstack.org/568613
15:38:24 dansmith and, they achieve the vendor neutrality of the elastic gpu thing by creating their own vendor so to speak, and provide signed drivers for supported operating systems
15:38:28 dansmith if they didn't have that,
15:38:45 dansmith they'd have to have specific details about which drivers you needed to have to support all their stuff, or provide flavors to let you choose which ones,
15:38:56 dansmith because drivers are required one way or the other
15:39:20 dansmith so unless you're going to write a driver and backend system to do that, I don't see how even the elastic gpu case alters the discussion
15:40:22 jaypipes dansmith: sounds like we need traits that represent the vendor software drivers, then...
15:41:53 dansmith or we expect that people have to do that on their own by using custom traits
15:42:11 dansmith which is what I said in that review, but then that generates a much less interoperable situation for everyone
15:42:33 jaypipes agreed.'
15:42:48 jaypipes dansmith: ok, thx for talking through this with me. was very useful.
15:43:01 jaypipes and for the record, I did not try to shame you. sorry if anything came across that way.
15:43:16 dansmith we also can punt the problem for a while, saying people need to use custom traits for now and see how it goes
15:45:45 jaypipes ack
15:46:21 openstackgerrit Merged openstack/nova master: placement: Fix HTTP error generation https://review.openstack.org/568567
16:01:39 openstackgerrit Chris Dent proposed openstack/nova master: [placement] default to accept of application/json when */* https://review.openstack.org/568630
16:08:27 mriedem stephenfin: have to talked with claudiu about whether or not the hyperv driver can support https://review.openstack.org/#/c/541290/ also?
16:09:07 stephenfin mriedem: I have not, but HyperV's support for all things NUMA is minimal at best
16:09:23 stephenfin Far as I can tell, they let you define guest topologies but not anything to do with the host topology
16:10:23 stephenfin So they don't do things like "restrict my instance's vCPU processes to CPUs from the same NUMA nodes as the attached PCI device". Wouldn't make sense (yet!) to do that for networks either
16:20:02 mriedem stephenfin: ok, as dansmith mentioned in the spec, it would be good to call out what is lacking in hyper-v to restrict them from being able to do the new magic
16:20:20 mriedem stephenfin: if you haven't seen yet, lots of comments on that spec, mostly concerned about the data modeling changes being proposed
16:20:26 dansmith yeah, sounds like it's cool, just a sentence saying "they don't do host topo" would be good I think
16:21:06 mriedem i'd also like to know what's going on with bauzas' numa topology modeling in placement spec
16:21:19 mriedem since this could technically depend on that and then we wouldn't need to build in more claims logic in the RT for this stuff
16:21:34 bauzas mriedem: I need to update my spec
16:21:44 mriedem like, if numa support is not even close for rocky, then i just need to know that
16:22:12 mriedem bauzas: you should probably read https://review.openstack.org/#/c/541290/ if you haven't already
16:22:54 bauzas mriedem: I'm a bit on and off given the previous week and the next one, but okay
16:23:16 bauzas maybe we could discuss that in Vancouver ?
16:24:55 cfriesen anyone know offhand where the code is that determines how much disk to consume if the flavor specifies a size of zero?
16:30:01 openstackgerrit Eric Fried proposed openstack/nova master: Use placement.inventory.inuse in report client https://review.openstack.org/568639
16:30:43 mriedem bauzas: i don't think we have any related forum sessions for something like this
16:30:52 mriedem unless you just mean between sessions
16:30:57 bauzas between, yeah
16:30:59 bauzas anyway
16:31:09 bauzas I'll look at the spec tomorrow if I can
16:31:16 bauzas I still have 3 free days
16:55:22 openstackgerrit Matt Riedemann proposed openstack/nova stable/queens: Don't reschedule on RequestedVRamTooHigh errors https://review.openstack.org/568642
17:10:53 openstackgerrit Merged openstack/nova stable/queens: Added ability to configure default architecture for ImagePropertiesFilter https://review.openstack.org/568575
17:10:59 openstackgerrit Merged openstack/nova master: Update auth_url in install docs https://review.openstack.org/568002
17:13:06 mriedem dansmith: that queens backport is ready to launch now https://review.openstack.org/#/c/568363/
17:13:27 dansmith right on
17:16:53 openstackgerrit Merged openstack/nova master: Add INVENTORY_INUSE to DELETE /rp/{u}/inventories https://review.openstack.org/568578
17:32:26 melwitt gibi, efried: there's not much of a plan. I'm going to put together a short slide deck similar to this cinder one from the boston summit (in case we need it) https://www.slideshare.net/JayBryant2/openstack-cinder-onboarding-education-boston-summit-2017
17:33:31 melwitt I was thinking I'll also prepare a set of logs from a gate run showing how to trace a request through the various nova services, in case anyone is interested in that
17:34:21 melwitt if you have any other topic ideas, it would help to have more choices available because what we usually do is poll the room and ask what topic they most like to hear about
17:35:58 efried melwitt: When sdague did that (walk through logs) in Boston I found it super useful. So ++ to that idea.
17:36:48 efried um, Boston or Austin, can't remember.
17:37:21 melwitt I think you're right that it was boston
17:41:44 openstackgerrit Eric Fried proposed openstack/nova master: Use placement.inventory.inuse in report client https://review.openstack.org/568639
18:20:24 openstackgerrit Eric Fried proposed openstack/nova master: Use placement.inventory.inuse in report client https://review.openstack.org/568639
18:20:46 openstackgerrit Eric Fried proposed openstack/nova master: Use placement.inventory.inuse in report client https://review.openstack.org/568639
18:28:28 openstackgerrit melanie witt proposed openstack/nova-specs master: Propose counting quota usage from placement and API database https://review.openstack.org/509042
18:37:15 openstackgerrit Matt Riedemann proposed openstack/nova master: Add granular policy rules for resource providers inventories https://review.openstack.org/568666
18:49:20 openstackgerrit melanie witt proposed openstack/nova-specs master: Propose counting quota usage from placement and API database https://review.openstack.org/509042
18:59:34 efried melwitt: Is it cheaper to get a list of instance UUIDs than whatever "instance_mappings" is?
19:04:30 melwitt efried: what do you mean? instance_mappings (API database table) is the cheapest way to get a list of instance UUIDs. otherwise you have to query multiple cell databases to get lists of UUIDs to merge
19:04:49 efried okay, cool, that's what I was asking.
19:27:54 openstackgerrit sunku ranganath proposed openstack/nova-specs master: Submitting blueprint describing usage of resource management daemon to control and use cache as a resource https://review.openstack.org/568678
19:32:03 openstackgerrit Matt Riedemann proposed openstack/nova master: Implement granular policy rules for placement https://review.openstack.org/524425
19:32:04 openstackgerrit Matt Riedemann proposed openstack/nova master: Add granular policy rules for /resource_classes* https://review.openstack.org/565578
19:32:05 openstackgerrit Matt Riedemann proposed openstack/nova master: Add granular policy rules for resource providers inventories https://review.openstack.org/568666
19:33:43 artom Do we do nested objects printing in debug mode?
19:34:09 artom So, InstancePCIRequests has a list of InstancePCIRequest
19:34:31 artom Logging the former at debug level... it is supposed to print out in full the latter as well?
19:35:24 artom Looking at https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/base.py#L309, I don't think so
19:35:58 artom Looks like it only prints the object names of the "nested" objects, right?
19:38:02 openstackgerrit Zack Cornelius proposed openstack/nova-specs master: Libvirt file backed memory https://review.openstack.org/563704
19:38:32 dansmith artom: that calls this for an object field: https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/fields.py#L819
19:38:44 dansmith artom: which tries to get some reasonable identifying information to print,
19:38:55 dansmith but doesn't recurse and repr() the object inside
19:39:21 dansmith just because it can get really hard to eye-parse that in a useful way, but if you see Instance(...., service=Service(123),...) then you can make sense out of that
19:39:30 artom Ah, thanks for choosing the right words - "recurse and repr()" indeed
19:39:49 artom Yeah, the rationale makes sense
19:40:07 artom Except I need resursive repr in this particular case :/
19:51:19 mriedem esberglu: just to check sanity, the in-tree powervm ci isn't running some snapshot tests b/c they are on power states that the driver doesn't yet support, like test_create_image_from_paused_server and test_create_image_from_suspended_server right?
19:54:37 esberglu mriedem: Correct
19:57:56 efried mriedem: Because https://github.com/powervm/powervm-ci/blob/master/devstack/master/intree/local.conf#L100
20:22:00 mriedem esberglu: efried: what passes the disk_path into the StreamToGlance task? https://review.openstack.org/#/c/543023/20/nova/virt/powervm/tasks/image.py@73
20:22:08 mriedem i thought it was the output of the last task in the flow,

Earlier   Later