Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-01-04
13:24:57 openstackgerrit Zhenyu Zheng proposed openstack/nova: Add query parameters white list for server list/detail https://review.openstack.org/408571
13:54:37 openstackgerrit Ed Leafe proposed openstack/nova: Do not post allocations that are zero https://review.openstack.org/407180
13:58:16 cdent johnthetubaguy: are rpc calls to the compute node's actually expensive (in the sense of we've got hard numbers) enough to be damaging, or something that we talk about as expensive because there are several things that combine to make slowness an issue?
13:58:45 johnthetubaguy cdent: its more just the act of waiting for the RPC call reply
13:59:01 cdent which presumably is unpredictable?
13:59:16 johnthetubaguy cdent: I guess, yeah
14:00:40 johnthetubaguy cdent: scaling wise, I don't really like having to scale up rabbit when you get more API calls, only want to scale that up when you get more instances/hypervisors, at least ideally
14:00:48 johnthetubaguy but thats largely make believe I guess
14:12:33 sdague alex_xu: do we care in https://review.openstack.org/#/c/408571/26/nova/api/openstack/compute/schemas/servers.py that ip6 doesn't show up until a later microversion?
14:13:06 openstackgerrit Sivasathurappan Radhakrishnan proposed openstack/nova: [WIP] Add version field to LibvirtLiveMigrateData https://review.openstack.org/404607
14:13:06 openstackgerrit Sivasathurappan Radhakrishnan proposed openstack/nova: [WIP] Refactor _post_live_migration https://review.openstack.org/413021
14:15:02 macsz johnthetubaguy: there is no need, the change that introduced this was merged after newton release
14:15:27 johnthetubaguy macsz: awesome, thanks for checking
14:15:49 macsz johnthetubaguy: np :)
14:44:17 openstackgerrit Sean Dague proposed openstack/nova: Add aggregate notification related enum values https://review.openstack.org/396209
15:03:06 gabor_antal Hey, would someone please review this patch? https://review.openstack.org/#/c/411816/
15:09:05 openstackgerrit Matt Riedemann proposed openstack/nova: Enable Neutron by default https://review.openstack.org/416083
15:13:42 johnthetubaguy mriedem: is that a disney themed snack?
15:13:57 mriedem sdague: sfinucan: thanks for going through that use_neutron series
15:14:02 mriedem johnthetubaguy: touche
15:14:05 mriedem but not this time
15:14:12 sdague mriedem: no prob, I'm still chugging up the stack right now
15:14:18 mriedem johnthetubaguy: it's -25 F with wind chill here today
15:14:29 mriedem keeps out the riff-raff
15:14:30 johnthetubaguy mriedem: thats proper nasty
15:14:40 johnthetubaguy the temperature that is :)
15:15:37 kashyap mriedem: Holy moly, -25...
15:15:48 kashyap Here I am trying to keep myself warm without heating from -1
15:16:16 mriedem coffee helps
15:16:28 dansmith mriedem: wow.. it's 24F here now, forecast for 17 tonight and we're all raising hell about it
15:16:29 edleafe mriedem: Yeah, it's like the arctic here in San Antonio - just under 50F!
15:16:58 mriedem dansmith: it's a good thing all those hipsters have beards and stocking caps year round
15:16:59 sfinucan mriedem: np. I'll grab the last one once Jenkins is happy
15:16:59 dansmith 7F with the wind chill I guess
15:17:03 dansmith mriedem: heh
15:17:33 mriedem edleafe: my wife was in phoenix last week complaining about it being about that temp
15:18:09 edleafe mriedem: yeah, having to wear a light hoodie is too much to bear
15:23:24 mriedem bauzas: are you around?
15:23:48 bauzas mriedem: indeed
15:23:56 bauzas just doing lots of internal stuff
15:24:25 mriedem bauzas: was wondering where you're at with https://blueprints.launchpad.net/nova/+spec/resource-providers-scheduler-db-filters
15:24:48 bauzas mriedem: the API change is waiting for reviews and I'm working on implementing the scheduler change
15:25:06 bauzas which is the only change needed for ^
15:25:19 openstackgerrit sahid proposed openstack/nova: numa: add numa constraints for emulator threads policy https://review.openstack.org/385023
15:25:21 openstackgerrit sahid proposed openstack/nova: objects: add new field cpus_overhead in NUMACell https://review.openstack.org/400874
15:25:21 openstackgerrit sahid proposed openstack/nova: libvirt: return a CPU overhead if isolate emulator threads requested https://review.openstack.org/385364
15:25:22 openstackgerrit sahid proposed openstack/nova: numa: fit instance toplogy with cpu overhead onto cores https://review.openstack.org/400875
15:25:23 openstackgerrit sahid proposed openstack/nova: virt: update overhead to take into account vCPUs https://review.openstack.org/385024
15:25:25 openstackgerrit sahid proposed openstack/nova: libvirt: make emulator threads to run on the reserved pCPU https://review.openstack.org/400878
15:25:26 openstackgerrit sahid proposed openstack/nova: numa: take into account cpus overhead https://review.openstack.org/400876
15:25:27 openstackgerrit sahid proposed openstack/nova: numa: update numa usage incording the cpu overhead https://review.openstack.org/400877
15:25:40 mriedem bauzas: ok i knew about https://review.openstack.org/#/c/392569/
15:25:44 mriedem just haven't gone through it yet
15:25:58 mriedem but wanted to know if you were at least doing the scheduler side stuff locally already, sounds like you are
15:26:46 bauzas indeed
15:26:55 mriedem btw, i was wondering, even though we're going to use the placement API for simple resource classes/filters like cpu/memory/disk, don't we still end up calling ComputeNodeList.get_all() for the other more advanced filters?
15:27:18 mriedem or is the idea to widdle the original list of compute nodes based on what the placement API gives us
15:27:34 mriedem regardless of the order of the scheduler filters in the config
15:27:37 johnthetubaguy (I thought it was the later)
15:27:48 bauzas mriedem: the latter yes
15:27:51 mriedem ok
15:28:28 bauzas (sorry, I'm awfully busy today, hopefully being less impacted tomorrow :/ )
15:28:45 mriedem np. on that note, i'm running to a meeting.
15:35:18 sdague cdent: are you fine with this - https://review.openstack.org/#/c/407180/6 ?
15:38:33 cdent sdague: I think the code has some stylistic issues, but functionally it's okay. I can go ahead and leave those comments now, but I wouldn't want it to block.
15:39:16 openstackgerrit Christopher Martin proposed openstack/nova: Bugfix: Changes to Instance Soft Delete https://review.openstack.org/386756
15:39:35 openstackgerrit Wang Qing wu proposed openstack/nova: WIP(2): Introduce base PowerVM driver (disk drv) https://review.openstack.org/409401
15:43:48 cdent sdague, edleafe: I've left those comments
15:56:16 openstackgerrit Gábor Antal proposed openstack/nova: Added missing decorator for instance.create.error https://review.openstack.org/416597
16:02:05 jaypipes edleafe: hey! you know I love my C code :)
16:09:54 cfriesen any chance of getting some reviewers for https://review.openstack.org/#/c/260339/ ? (It's not my patch, but it's a useful fix.) There's a note about a possible tempest race in the bug, so would we need to fix the tempest test before fixing nova?
16:10:44 openstackgerrit Gábor Antal proposed openstack/nova: Removed unnecessary parantheses and fixed formation https://review.openstack.org/370549
16:19:33 openstackgerrit melanie witt proposed openstack/nova: WIP Count resources to check quota for cells https://review.openstack.org/416521
16:20:42 edleafe cdent: Not sure what I was going for there, but yeah, it looks wrong. Will fix.
16:21:21 cdent ✔
16:30:14 openstackgerrit Gleb Stepanov proposed openstack/python-novaclient: Add evacuate host method https://review.openstack.org/349179
16:30:52 sfinucan mriedem: Boom https://review.openstack.org/#/c/416083
16:32:09 openstackgerrit Gábor Antal proposed openstack/nova: Removed unnecessary parantheses and fixed formation https://review.openstack.org/370549
16:32:33 sfinucan :P
16:32:42 openstackgerrit Wang Qing wu proposed openstack/nova: WIP(3): Introduce base PowerVM driver (console) https://review.openstack.org/409402
16:36:29 openstackgerrit Ed Leafe proposed openstack/nova: Do not post allocations that are zero https://review.openstack.org/407180
16:36:43 edleafe cdent: should be better ^^
16:37:29 cdent edleafe: decided you didn't like my concern about self.used?
16:38:57 edleafe cdent: you mean protecting against negative used values?
16:39:24 cdent yes, and also being explicit about the conditional being numeric
16:39:48 edleafe cdent: not sure I follow
16:40:39 cdent I just think that self.used == 0 or self.used <= 0 is more explicit/readable than not self.used
16:41:02 mriedem sdague: mtreinish: easy one https://review.openstack.org/#/c/416401/
16:41:06 mriedem sfinucan: thanks
16:41:09 edleafe cdent: self.used == 0 is major ugly in Python
16:41:27 edleafe <= 0 is another thing
16:41:31 mtreinish mriedem: +2
16:41:45 cdent but it is _exactly_ what you mean to be testing for there, especially based on the message raised if it is true
16:42:00 edleafe cdent: if you think we need to guard against negative values, I'll add <= 0
16:42:40 cdent a) just out of curiosity why is self.used == 0 ugly? b) I'm not sure if we do or not, I'm asking, not telling
16:43:07 mriedem dansmith: there will be a thrilling cells v2 meeting in 17 minutes yes?
16:43:34 mriedem dare i say titilating
16:44:09 edleafe cdent: 0, False, empty string/dict/list/tuple/set/etc are all boolean False, so you don't test for == 0, or == {}, or the like
16:44:43 cdent edleafe: I understand that, and that's my point: you _are_ testing for 0 here. Not False or empty
16:45:03 openstackgerrit Sylvain Bauza proposed openstack/nova: Expose a REST API for a specific list of RPs https://review.openstack.org/392569

Earlier   Later