Posted	Nick	Remark
#openstack-nova - 2017-10-05
14:39:44	andreas_s_	mriedem, Thanks!
14:41:20	cdent	efried: that’s kind of a special case and was sort of the genesis of custom resource classes
14:41:58	cdent	because the idea was that an ironic node is a single consumable thing
14:42:32	efried	edleafe I updated https://wiki.openstack.org/wiki/Meetings/Scheduler to point to https://wiki.openstack.org/wiki/Meetings/NovaScheduler
14:42:57	efried	cdent jaypipes So does that mean we actually don't have a way to specify generic resources in flavors yet?
14:43:27	jaypipes	efried: in the flavor's extra_specs. there's a special key pattern called "resources:$RESOURCE_CLASS" and the value is the amount of that resource class the flavor wants.
14:43:52	efried	jaypipes Ah, perfect. And that's been implemented already?
14:43:57	jaypipes	efried: those special extra_specs resources: things are called resource overrides.
14:44:03	jaypipes	efried: yes.
14:44:16	jaypipes	efried: edleafe and alex_xu implemented that IIRC
14:44:44	efried	jaypipes Cool; and (as the name "overrides" implies) I can specify "standard" resources there too, like VCPU, DISK_GB, etc., and their corresponding hardcoded fields will be ignored?
14:47:09	openstackgerrit	Gábor Antal proposed openstack/nova master: Transform instance.rebuild_scheduled notification https://review.openstack.org/473929
14:47:33	efried	jaypipes So in a nutshell, this new thing will take resources$X:$RESOURCE_CLASS=$N and traits$X:$TRAIT=required and translate them into GET /allocation_candidates?resources$X=$N...&required$X=$TRAIT
14:47:45	efried	jaypipes Does that align with your thinking?
14:48:10	jaypipes	efried: precisely.
14:48:14	efried	beaut
14:48:40	mriedem	gibi: thanks for https://review.openstack.org/#/c/509759/ i was just looking at fixing that
14:48:44	mriedem	gibi: comment within
14:48:51	openstackgerrit	Dan Smith proposed openstack/nova master: Fix minor input items from previous patches https://review.openstack.org/506416
14:48:51	openstackgerrit	Dan Smith proposed openstack/nova master: Merge build requests into the sortmaster 9000 https://review.openstack.org/508595
14:48:52	openstackgerrit	Dan Smith proposed openstack/nova master: Add CellMapping.get_by_project_id() query method https://review.openstack.org/509002
14:48:52	openstackgerrit	Dan Smith proposed openstack/nova master: Make get_instance_objects_sorted() be smart about cells https://review.openstack.org/509003
14:49:04	dansmith	mriedem: rebased now that your metadata cleanup has merged ^
14:49:17	mriedem	ok
14:49:56	mriedem	says the bottom change is in merge conflict... ?
14:50:19	dansmith	um
14:50:20	mriedem	but it's queued up so idk
14:50:23	dansmith	Iliterally just pushed
14:50:27	mriedem	maybe a zuulv3 thing
14:52:20	gibi	mriedem: I will push a new ps soon with your suggestion implemented
14:52:46	mriedem	dansmith: i see we're running singleconductor mode in grenade between pike->queens, couldn't we be running superconductor?
14:53:08	mriedem	maybe that would require making grenade aware of that kind of mode?
14:53:36	dansmith	yeah I think we just punted on any sort of telling grenade about that since we couldn't upgrade from one layout to the other
14:53:47	dansmith	I dunno how much it really needs to know about, but I haven't looked
14:55:05	openstackgerrit	Zhenyu Zheng proposed openstack/nova-specs master: Improve the performance of filtering instances by IP. https://review.openstack.org/509326
14:55:06	mriedem	https://github.com/openstack-dev/grenade/blob/03de9e0fc7f4fc50a00db5d547413e26cf0780dd/grenade.sh#L222
14:55:29	mriedem	might just be worth removing that from master grenade and see how it goes
14:55:59	openstackgerrit	Gábor Antal proposed openstack/nova master: Transform aggregate.update_prop notification https://review.openstack.org/462576
14:56:32	openstackgerrit	Dan Smith proposed openstack/nova master: Pre-create migration object https://review.openstack.org/498950
14:56:33	openstackgerrit	Dan Smith proposed openstack/nova master: Revert allocations by migration uuid https://review.openstack.org/498949
14:56:33	openstackgerrit	Dan Smith proposed openstack/nova master: Refactor resource tracker to account for migration allocations https://review.openstack.org/506419
14:56:34	openstackgerrit	Dan Smith proposed openstack/nova master: Make migration uuid hold allocations for migrating instances https://review.openstack.org/506420
14:56:57	dansmith	jaypipes: cdent: I didn't see that failure in ^ locally, so this is a rebase and we'll see how it goes this time
14:57:04	jaypipes	kk
15:00:43	jaypipes	dansmith, mriedem: I'm ready on this one after a number of back and forths with johnthetubaguy. https://review.openstack.org/#/c/507052/
15:01:01	openstackgerrit	Gábor Antal proposed openstack/nova master: Transform instance.live_migration_rollback_dest notification https://review.openstack.org/480214
15:01:06	johnthetubaguy	jaypipes: thanks for your time on that, glad to move that forward
15:02:51	dansmith	edleafe: is this fail related to your change? http://logs.openstack.org/15/486215/14/check/gate-nova-tox-functional-py35-ubuntu-xenial/f7d2653/testr_results.html.gz
15:04:16	mriedem	jaypipes: i'll put it in the tab qeuue
15:04:19	mriedem	*queue
15:04:51	mriedem	dansmith: i'll go through https://review.openstack.org/498950 again
15:05:11	dansmith	mriedem: thanks, I don't think you had any feedback on that did you?
15:05:24	mriedem	yes...
15:05:54	dansmith	oh the fail cleanup, right
15:09:29	tikitavi	look to the nova-conductor logs, please (http://logs.openstack.org/91/500091/8/check/gate-functional-neutron-dsvm-ec2api-ubuntu-xenial/206b664/). We have problem with starting instances in stable/ocata and stable/pike.
15:10:35	dansmith	tikitavi: you mean the traceback?
15:10:44	tikitavi	yes!
15:10:44	dansmith	2017-10-05 10:48:16.014 30844 ERROR oslo_messaging.rpc.server IOError: [Errno 2] No such file or directory: '/usr/local/lib/python2.7/dist-packages/six-1.10.0.dist-info/METADATA'
15:10:47	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: de-ORM ResourceProvider.get_by_uuid() https://review.openstack.org/509025
15:10:48	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: Move RP._get|set_aggregates() to module scope https://review.openstack.org/509026
15:10:48	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: Remove RP.get_traits() method https://review.openstack.org/509027
15:10:49	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: move RP._set_traits() to module scope https://review.openstack.org/509028
15:10:49	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: remove CRUD operations on Inventory class https://review.openstack.org/509029
15:10:50	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: streamline InventoryList.get_all_by_rp_uuid() https://review.openstack.org/509030
15:10:50	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: remove dead code in Allocation._create_in_db() https://review.openstack.org/509031
15:10:51	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: remove ability to delete 1 allocation record https://review.openstack.org/509032
15:10:51	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: fix up AllocList.get_by_resource_provider_uuid https://review.openstack.org/509033
15:10:52	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: rework AllocList.get_all_by_consumer_id() https://review.openstack.org/509035
15:10:52	dansmith	tikitavi: seems clearly like some installation problem
15:10:52	openstackgerrit	Jay Pipes proposed openstack/nova master: rp: remove _HasAResourceProvider mixin https://review.openstack.org/509036
15:11:40	tikitavi	dansmith: it is repeatable. we can't pass this review for more than a month
15:12:03	dansmith	tikitavi: I understand, but still, it's a package problem, nothing to do with any nova ode
15:12:04	dansmith	*code
15:13:00	edleafe	dansmith: I ran those tests locally w/o a fail, so I don't think so.
15:13:18	dansmith	edleafe: yea, I rechecked and noted that the next patch didn't hit the same failure
15:16:08	mriedem	sdague: did your "wait for host mapping" change in devstack get backported to stable/pike? i'm seeing http://logs.openstack.org/56/509656/1/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/6851396/logs/old/screen-n-cond.txt.gz#_2017-10-04_23_25_13_456 in stable/pike grenade runs only
15:17:09	mriedem	oh hm https://review.openstack.org/#/c/490635
15:17:38	sdague	right, that will need to land there. I can push it in
15:18:16	mriedem	+1 fwiw
15:18:17	mriedem	:)
15:21:25	dansmith	johnthetubaguy: a spec from you with no discernible typos.. jaypipes must have been up all night. :P
15:21:49	jaypipes	lol
15:21:58	johnthetubaguy	dansmith: him and efried got most of them the previous time :)
15:22:09	dansmith	johnthetubaguy: heroes, those guys.
15:22:24	mriedem	ooo -1
15:22:30	mriedem	i'll fix after it's merged
15:25:12	openstackgerrit	Matt Riedemann proposed openstack/nova-specs master: Rename ironic-driver-traits spec https://review.openstack.org/509812
15:25:53	openstackgerrit	sean mooney proposed openstack/nova-specs master: Use neutron's new port binding API https://review.openstack.org/375580
15:26:28	openstackgerrit	Balazs Gibizer proposed openstack/nova master: fix unstable shelve offload functional tests https://review.openstack.org/509759
15:26:50	gibi	mriedem: fixed your comment in the shelve offload test bugfix ^^
15:27:01	shvepsy	Hi all! Can you help me with my issue.
15:27:02	shvepsy	How I can call quiesce instance (freeze fs) from python API?
15:27:05	mdbooth	dansmith: I vaguely recall some error from the dim and distants depths of time which meant that an error in a periodic task would prevent that task from continuing to run. Does that ring any bells with you?
15:27:37	mriedem	johnthetubaguy: the thing on L37 of your spec confuses me
15:27:47	mriedem	aren't you just describing the flavor-access API?
15:27:50	dansmith	mdbooth: no, not that I know of
15:28:00	johnthetubaguy	mriedem: I guess
15:28:21	johnthetubaguy	mriedem: its just a public flavor and a private flavor, mapping to different groups of hosts
15:28:36	mriedem	johnthetubaguy: so flavor-access with host aggregates?
15:30:01	dansmith	mriedem: I'm not sure what your point is? that you can accomplish the same thing with aggregates and a private flavor?

Earlier   Later