Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-12-22
02:33:07 openstackgerrit Yikun Jiang (Kero) proposed openstack/nova master: Fix error when deleting any root resource provider https://review.openstack.org/529519
02:33:12 mriedem sqlite 3.7+ should work with foreign keys
02:33:13 mriedem see TestDatabaseArchive
02:33:36 mriedem engine.connect().execute("PRAGMA foreign_keys = ON")
02:33:48 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Test allocation candidates: multiple aggregates https://review.openstack.org/518633
02:33:49 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Fix accumulated nits in refactor series https://review.openstack.org/521189
02:33:49 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Test helper: validate provider summaries https://review.openstack.org/518982
02:33:50 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Fix missing rps in allocation candidates https://review.openstack.org/522409
02:33:50 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Add aggregates check in allocation candidates https://review.openstack.org/522407
02:33:51 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: placement: func tests for multiple shared RPs https://review.openstack.org/498737
02:34:03 yikun o? I will take a look, and try it in my local env, thanks. : )
02:34:08 yikun mriedem,
02:34:10 mriedem yw
02:38:33 mriedem yikun: fyi https://review.openstack.org/#/c/529672/
02:38:37 mriedem turns out we needed that after all
02:42:33 yikun mriedem, emm, yes, +1, the deadlock on action event start is possible,
02:42:40 yikun because action and event doesn't have some relation ship, so sqlalchemy session mechanism can't ensure the update order.
02:43:57 yikun some ref: http://www.aosabook.org/en/sqlalchemy.html # 20.9. Unit of Work
02:45:16 Kevin_Zheng consider add relationship?
02:45:34 openstackgerrit Merged openstack/nova master: Fix unit test failures when direct IO not supported https://review.openstack.org/529568
02:46:00 Kevin_Zheng mriedem we were thinking about a full scan about these kind of things in Nova
02:46:37 Kevin_Zheng https://review.openstack.org/#/c/527362/
02:47:33 Kevin_Zheng ^ the deadlock retry method is not perfect, so we might be able to improve performance here
02:58:45 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Test allocation candidates: multiple aggregates https://review.openstack.org/518633
02:58:46 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Fix accumulated nits in refactor series https://review.openstack.org/521189
02:58:46 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Test helper: validate provider summaries https://review.openstack.org/518982
02:58:47 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Fix missing rps in allocation candidates https://review.openstack.org/522409
02:58:47 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: Add aggregates check in allocation candidates https://review.openstack.org/522407
02:58:48 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: placement: func tests for multiple shared RPs https://review.openstack.org/498737
03:04:04 openstackgerrit OpenStack Proposal Bot proposed openstack/nova master: Updated from global requirements https://review.openstack.org/529783
03:37:22 openstackgerrit Tetsuro Nakamura proposed openstack/nova master: trivial: use cn instead of rp https://review.openstack.org/529786
05:03:18 openstackgerrit Jianghua Wang proposed openstack/nova master: XenAPI: provide VGPU_DISPLAY_HEAD inventory in compute node https://review.openstack.org/523342
09:33:36 openstackgerrit Maciej Jozefczyk proposed openstack/nova master: Update resources once in update_available_resource https://review.openstack.org/520024
09:37:08 openstackgerrit Maciej Jozefczyk proposed openstack/nova master: Update resources once in update_available_resource https://review.openstack.org/520024
10:14:29 ralonsoh stephenfin: hi, about https://review.openstack.org/#/c/449257/52/nova/pci/request.py@93
10:18:34 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova master: Read Neutron port 'binding_profile' during boot https://review.openstack.org/507481
10:18:34 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/nova master: Change 'InstancePCIRequest' spec field https://review.openstack.org/449257
10:49:22 openstackgerrit Sylvain Bauza proposed openstack/nova master: libvirt: create vGPU for instance https://review.openstack.org/528832
10:55:51 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Add abstract OVSDB API https://review.openstack.org/476612
11:01:30 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Add abstract OVSDB API https://review.openstack.org/476612
11:05:50 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Add native implementation OVSDB API https://review.openstack.org/482226
11:16:01 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Migrate from 'ip' commands to 'pyroute2' https://review.openstack.org/484386
11:22:23 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Add memoize function using oslo.cache https://review.openstack.org/472773
11:33:21 openstackgerrit rahul bardia proposed openstack/python-novaclient master: nova limits ERROR (Exception): Field names must be unique https://review.openstack.org/527922
12:55:25 jaypipes good morning nova comrades.
12:57:59 cdent morning jaypipes
12:58:09 cdent get your dishwasher and dinner sorted?
12:58:33 jaypipes no and yes.
12:59:28 jaypipes cdent: friggin electricians hooked up the dishwasher's electric but said they could touch the plumbing. so I hooked up the plumbing and there's a leak. so my wife's uncle is out grabbing plumber's tape and we're going to pull it out and have a look at where the leak is. the saga continues.
12:59:41 cdent argh
12:59:50 jaypipes cdent: cost me almost as much for the 1.25 hours of electrician as for the new dishwasher :(
12:59:53 jaypipes $400...
12:59:57 cdent jeebus
13:00:00 jaypipes ya.
13:00:37 jaypipes so after dinner, the eleven of us just made a Hobbit-dwarf-esque factory line and polished off the dish cleanup team-style.
13:00:58 cdent well if it makes you feel any better, one christmas at my parents, where they had no dishwasher, I was responsible for peeling the potatoes and somehow managed to completely screwn up the garbage disposal such that the sink wouldn't drain
13:01:03 cdent we did the dishes in the bathtub
13:01:31 cdent did you sing?
13:06:33 jaypipes cdent: heh, that's... awesome. :) and no, I did not sign.
13:06:34 jaypipes sing
13:12:37 maciejjozefczyk jaypipes: good morning
13:12:43 jaypipes maciejjozefczyk: mornin!
13:12:52 jaypipes maciejjozefczyk: or afternoon for you I suppose :)
13:13:20 leakypipes cdent: my nick actually reflects current status today.
13:13:37 cdent woot
13:13:44 maciejjozefczyk jaypipes: here in PL - yes, afternoon, but also good ;)
13:14:13 leakypipes maciejjozefczyk: are you in Poznan or Gdansk?
13:14:25 maciejjozefczyk leakypipes: sorry for saying that but this change https://review.openstack.org/#/c/520024 breaks resource tracking in placement :) I need to dig more
13:14:28 maciejjozefczyk leakypipes: Wroclaw
13:14:31 leakypipes ah
13:15:37 leakypipes maciejjozefczyk: k, I'll take a look at that patch again (on the resource provider issue) shortly.
13:17:00 leakypipes maciejjozefczyk: what you can try in the meantime is just replacing each of those calls to self._update() one by one and see which specific one causes the invalid inventory exception to go away...
13:17:38 openstackgerrit rahul bardia proposed openstack/python-novaclient master: nova limits ERROR (Exception): Field names must be unique https://review.openstack.org/527922
13:17:47 maciejjozefczyk leakypipes: I found that its in _update_inventory_for_provider /opt/stack/nova/nova/objects/resource_provider.py(232
13:18:51 leakypipes maciejjozefczyk: yes, I see that. but I'm wondering which of the locations in https://review.openstack.org/#/c/520024/8/nova/compute/resource_tracker.py where you removed the call to self._update() causes that exception to be hit.
13:19:27 leakypipes maciejjozefczyk: so you might try replacing (one by one) those call locations of self._update() in the resource tracker and determining which of those calls fixes the reserved inventory invalid exception.
13:19:32 leakypipes that's all I was saying :)
13:19:55 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Migrate from 'ip' commands to 'pyroute2' https://review.openstack.org/484386
13:19:56 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Add support for Windows network commands https://review.openstack.org/487405
13:23:31 maciejjozefczyk leakypipes: I found that because I've removed duplications the self._resource_change(compute_node) always says false (yeah, thats right) so the ComputeNode object is not updated directly (I mean with cn.save()), and without this call the exceptions is raised by placement
13:23:33 maciejjozefczyk https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L854
13:23:50 maciejjozefczyk leakypipes: anyway, need to switch context because of other tasks... We'll dig it later ;)
13:24:12 maciejjozefczyk s/we/I/ :)
13:24:13 leakypipes k, no prob
13:24:27 bauzas hola, folks
13:24:58 bauzas (yeah, it seems weird that I'm looking to be in an US TZ, but no worries, I was not looking at the IRC this morning :p)
13:26:32 leakypipes bauzas: where are ya?
13:26:52 bauzas just still in Le Versoud, FR :p
13:27:13 bauzas wed, I was off because $skiing
13:27:39 bauzas so, I needed to work off the IRC by this morning for other things :(
13:28:05 bauzas oh, and Friday, right
13:33:12 openstackgerrit Merged openstack/nova master: Don't try to delete build request during a reschedule https://review.openstack.org/528835
13:42:40 mriedem mdbooth: ping me once https://review.openstack.org/#/c/242602/ is fixed up and i'll +2
13:43:03 mdbooth mriedem: Thanks, will do.
13:43:43 bauwser mriedem: today is my last day for 2017, how can I help for today ?
13:44:14 bauwser mriedem: I worked most of the time off IRC because *VGPU, but in case you need stable reviews or anything else
13:45:46 mriedem bauwser: we have these ocata and newton backports yet https://review.openstack.org/#/q/I214a44f0eee7d90be5cd89f32f6e0017b19a3fd6
13:46:05 bauwser ah right, I saw the email
13:46:22 bauwser mriedem: thanks for helping on that bug, btw.
13:46:38 mriedem bauwser: and these https://review.openstack.org/#/q/I0b3ec6bb098ca32ffd32a61d4f9dcf426c3faf46

Earlier   Later