Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2018-05-23
17:51:56 efried Correct level for audience, good questions.
17:52:20 efried One attendee wants instance count quotas :P
17:52:35 jaypipes efried: good to hear!
17:56:59 gibi efried: I'm glad you did not used the fridge and sandwich example to talk about removing allocations
17:57:24 efried gibi: ikr, the analogy really breaks down there o.O
17:57:49 efried I did actually mean to mention that
18:01:40 gibi :)
18:38:38 cdent validate all the knobs!
18:40:06 efried cdent: It's not that we couldn't validate the release of allocations in the fridge model. It's that... what does that mean? I don't want to think about how to un-eat my sandwich and return the pickles to the jar.
18:42:00 cdent efried: that's a paraphrase of mriedem in an NVF session. However, I'm certain that a virtual sandwich can be un-eaten, and in any case, when you describe and claim the sandwich, whether you eat it or not is out of scope.
18:42:09 cdent nfv
18:42:18 efried :)
18:43:43 artom Observing your sessions from afar with no context is entertaining :)
18:44:38 artom ... and appetite-buliding
19:14:12 bauzas jaypipes: efried: I wrapped my head around finding a good way for describing traits with GPUs and I finally ended up having no real possibility but having custom traits defined by the operator
19:14:32 bauzas jaypipes: efried: the problem is that standard traits about display heads et al. aren't sufficient
19:15:01 efried bauzas: Does this have to do with the issue that we talked about yesterday: needing specific drivers for specific card models etc?
19:15:33 bauzas like, I have a CUDA application that uses my super expensive nvidia P100 but I wouldn't like ending up on a host which is either having Intel GPU or an older nvidia GPI
19:15:38 bauzas efried: indeed
19:16:06 bauzas efried: vendor drivers is one part of the problem
19:16:43 bauzas efried: so I'm about to update https://review.openstack.org/#/c/557065/4/specs/rocky/approved/vgpu-rocky.rst and add a possibility to set traits on PCI IDs
19:17:18 efried bauzas: Which you can already do.
19:17:38 bauzas efried: by osc-placement ?
19:17:41 efried yes
19:17:45 bauzas efried: with nested resource providers ?
19:17:53 bauzas then, we don't need a conf opt
19:18:06 bauzas it's just documentation as we now support placement RBAC
19:18:17 efried bauzas: The tricky part, which I yelled about for a bit in Dublin, is representing the line between traits the virt driver "owns" and traits the user "owns".
19:19:00 bauzas efried: and I think I considered the operator traits as only being additive
19:19:30 bauzas if you set a trait on a node, it's your responsibility to make it consistent from what you got from the driver
19:22:16 efried1 Because the virt driver (or really, the compute service) "owns" certain traits even when they're not set for a given RP.
19:22:25 efried1 and is in fact responsible for unsetting them if they're not applicable.
19:25:09 bauzas efried: sure, and I don't see a problem with that
19:25:22 bauzas efried: that's just two sets of things
19:25:36 bauzas efried: the virt driver owns its set of traits
19:25:50 bauzas efried: the operator owns another set
19:26:24 bauzas efried: if for some reason the operator provides a traits that the virt driver could own, having the virt driver removing that trait authoritatively isn't a problem for me
20:51:56 openstackgerrit Matt Riedemann proposed openstack/nova master: Add encrypted volume support for feature matrix docs https://review.openstack.org/570255
21:00:15 openstackgerrit Matt Riedemann proposed openstack/nova master: Add encrypted volume support to feature matrix docs https://review.openstack.org/570255
21:00:16 mriedem lyarwood: ^ keep me honest
21:32:11 openstackgerrit Sylvain Bauza proposed openstack/nova-specs master: Proposes Multiple GPU types https://review.openstack.org/557065
21:37:19 mgagne hi, could I get a review on that one when someone has time? thanks! https://review.openstack.org/#/c/523559/
21:54:08 openstackgerrit Chris Dent proposed openstack/nova master: Optional separate database for placement API https://review.openstack.org/362766
21:54:09 openstackgerrit Chris Dent proposed openstack/nova master: Add PLACEMENT_DB_ENABLED=True to the nova-next job https://review.openstack.org/564067
21:56:22 openstackgerrit Chris Dent proposed openstack/nova master: Ensure that os-traits sync is attempted only at start of process https://review.openstack.org/553857
22:00:51 openstackgerrit Chris Dent proposed openstack/nova master: Isolate placement database config https://review.openstack.org/541435
22:01:00 openstackgerrit Chris Dent proposed openstack/nova master: Ensure that os-traits sync is attempted only at start of process https://review.openstack.org/553857
22:23:12 openstackgerrit Dan Smith proposed openstack/nova stable/queens: Add ssdb and virt-ssdb flags to cpu_model_extra_flags whitelist https://review.openstack.org/570271
22:39:21 bauzas jaypipes: you could be interested in https://review.openstack.org/#/c/569132/ and above
22:39:52 bauzas I'm blocked for my vGPU series up until https://review.openstack.org/#/c/560444/ merges
22:47:00 openstackgerrit Dan Smith proposed openstack/nova stable/queens: Add ssdb and virt-ssdb flags to cpu_model_extra_flags whitelist https://review.openstack.org/570271
22:47:45 openstackgerrit Dan Smith proposed openstack/nova stable/queens: Add ssdb and virt-ssdb flags to cpu_model_extra_flags whitelist https://review.openstack.org/570271
22:53:36 mriedem cfriesen_: https://www.redhat.com/archives/libvir-list/2018-May/msg01562.html
23:12:47 openstackgerrit Dan Smith proposed openstack/nova stable/queens: Add ssdb and virt-ssdb flags to cpu_model_extra_flags whitelist https://review.openstack.org/570271
23:16:46 openstackgerrit Dan Smith proposed openstack/nova stable/queens: Add ssdb and virt-ssdb flags to cpu_model_extra_flags whitelist https://review.openstack.org/570271
23:22:59 mriedem lbragstad[m]: attempted to reply to your scope types question in https://review.openstack.org/#/c/524425/16/nova/api/openstack/placement/policies/resource_provider.py
23:23:39 mriedem tl;dr if i understand scope properly, i think resource providers would be system scope (they aren't scoped to a project at all) and things like resource allocations and usages would be scoped to a project (they are specific to something like a server instance which is per project)
23:40:54 openstackgerrit Merged openstack/nova master: Convert libvirt's RBD storage to using processutils.execute() https://review.openstack.org/569862
#openstack-nova - 2018-05-24
01:13:25 openstackgerrit Merged openstack/nova master: Remove mox in libvirt/test_driver.py (2) https://review.openstack.org/568757
01:42:03 openstackgerrit Zhenyu Zheng proposed openstack/nova master: Only run placement request filters when Placement will be called https://review.openstack.org/569969
04:31:16 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: Initial change set of z/VM driver https://review.openstack.org/523387
04:31:17 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: Spawn and destroy function of z/VM driver https://review.openstack.org/527658
04:31:18 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add snapshot function https://review.openstack.org/534240
04:31:19 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add power actions https://review.openstack.org/543340
04:31:20 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add get console output https://review.openstack.org/543344
05:41:20 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add snapshot function https://review.openstack.org/534240
05:41:21 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add power actions https://review.openstack.org/543340
05:41:22 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add get console output https://review.openstack.org/543344
06:10:45 openstackgerrit jichenjc proposed openstack/nova master: WIP: Remove support for /os-fixed-ips REST API https://review.openstack.org/568516
06:12:37 openstackgerrit jichenjc proposed openstack/nova master: raise exception early https://review.openstack.org/569024
06:54:57 openstackgerrit Merged openstack/nova master: Honor availability_zone hint via placement https://review.openstack.org/546282
06:56:58 openstackgerrit Zhenyu Zheng proposed openstack/nova master: Only run placement request filters when Placement will be called https://review.openstack.org/569969
07:42:00 openstackgerrit sahid proposed openstack/nova master: metadata: add vf_trusted field to device metadata https://review.openstack.org/566344
07:49:38 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add snapshot function https://review.openstack.org/534240
07:49:39 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add power actions https://review.openstack.org/543340
07:49:40 openstackgerrit jichenjc proposed openstack/nova master: z/VM Driver: add get console output https://review.openstack.org/543344
08:51:46 openstackgerrit Hamdy Khader proposed openstack/nova master: Adding NVMEoF for libvirt driver https://review.openstack.org/482640
08:53:07 openstackgerrit Hamdy Khader proposed openstack/nova master: Adding NVMEoF for libvirt driver https://review.openstack.org/482640
09:27:03 openstackgerrit Kashyap Chamarthy proposed openstack/nova master: libvirt: Deprecate support for monitoring Intel CMT `perf` events https://review.openstack.org/565242
09:44:28 guest_ PCI alias gtx630 is not defined (HTTP 400) (Request-ID: req-a506add4-6167-47c3-ac87-dc31bf9a7be7)
09:44:57 guest_ PCI alias gtx630 is not defined (HTTP 400) (Request-ID: req-a506add4-6167-47c3-ac87-dc31bf9a7be7)
09:45:00 guest_ kolla-queens
09:46:26 openstackgerrit Zhenyu Zheng proposed openstack/nova master: nova-manage db archive_deleted_rows is not multi-cell aware https://review.openstack.org/507486
11:23:03 openstackgerrit MultipleCrashes proposed openstack/nova master: rerunning for jenkins ci zull check https://review.openstack.org/570370
12:45:12 openstackgerrit Peter Hamilton proposed openstack/nova-specs master: Add support for certificate validation https://review.openstack.org/570381
13:49:41 openstackgerrit Peter Hamilton proposed openstack/nova-specs master: Add support for certificate validation https://review.openstack.org/570381
13:54:25 openstackgerrit Dan Smith proposed openstack/nova stable/queens: Add ssbd and virt-ssbd flags to cpu_model_extra_flags whitelist https://review.openstack.org/570271
14:11:45 openstackgerrit Eric Berglund proposed openstack/nova stable/pike: Stringify instance UUID https://review.openstack.org/569421
15:10:50 openstackgerrit Chen proposed openstack/nova master: Fix some inconsistencies in doc https://review.openstack.org/570407
15:50:05 mriedem o/
15:55:11 mriedem dansmith: with https://review.openstack.org/#/c/546282/ merged, i think that bp is done now yeah?
15:55:17 mriedem https://review.openstack.org/#/q/topic:bp/placement-req-filter+(status:open+OR+status:merged)
16:03:42 dansmith mriedem: yeah, those are all the req filters I had ready to go, so .. yeah
16:04:37 mriedem done
16:22:47 dansmith cool
16:34:25 openstackgerrit Dan Smith proposed openstack/nova stable/queens: Add ssbd and virt-ssbd flags to cpu_model_extra_flags whitelist https://review.openstack.org/570271
16:34:38 mriedem1 note to self, my ip continually changes during a session so authorship colors on the etherpads are nonsense now
16:35:35 mtreinish mriedem: heh, it is annoying
16:35:57 mriedem "so i know you guys said no to LTS, but seriously, at which branch do we start LTS?"
16:36:46 mtreinish lol, yeah they never give up

Earlier   Later