| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-01-19 | |||
| 16:28:44 | openstackgerrit | Pawel Koniszewski proposed openstack/nova: Change exponential function to linear https://review.openstack.org/407518 | |
| 16:28:50 | _gryf | it will still fail if there is no keystone around | |
| 16:28:54 | openstackgerrit | Pawel Koniszewski proposed openstack/nova: Mark live_migration_downtime_steps as deprecated for removal https://review.openstack.org/408002 | |
| 16:30:33 | _gryf | failure was caused by the fact, that compute was started earlier than keystone. So probably the right choice would be to make it wait a bit (using warn_limit) | |
| 16:30:50 | ildikov | scottda: done, thanks | |
| 16:39:02 | markus_z | esberglu: hey, you're the guy managing the nova-powervm releases, right? It looks like your team doesn't use the "openstack/releases" repo for that. Why is that? | |
| 16:39:42 | openstackgerrit | Davanum Srinivas (dims) proposed openstack/nova: [WIP] ignore me - trying experimental jobs https://review.openstack.org/421209 | |
| 16:41:13 | BobBall | Stupid question... Is there a way to run the nova unit tests with fewer than the number of CPUs on the host? I've got a shared development host with 40 CPUs and while the unit tests finish very quickly with that number of CPUs using them all is very antisocial and wedges the rest of the system | |
| 16:42:52 | openstackgerrit | Sujitha proposed openstack/nova: Allow finding instance with invalid regex name https://review.openstack.org/392305 | |
| 16:44:39 | mriedem | BobBall: --concurrency=1 i think | |
| 16:45:45 | esberglu | markus_z: Didn't know that was the way to do it | |
| 16:45:56 | BobBall | Awesome - thanks. Needed an extra -- to pass the args through to pretty_tox, but yes: tox -epy27 -- --concurrency=4 | |
| 16:47:00 | markus_z | esberglu: I'm still investigating, not yet 100% sure that's the way to go for "out-of-tree" drivers. | |
| 16:47:09 | jaypipes | bhagyashris: ok, sorry, meetings. alright, so to answer your question, patch 417961 doesn't update any allocations information in the placement API at all. | |
| 16:47:50 | jaypipes | bhagyashris: the only thing that updates allocation information currently is the nova-compute node (the resource tracker contains a scheduler reporting *client* which is badly named but doesn't actually talk to the scheduler. | |
| 16:47:50 | adreznec | markus_z: Part of that is an artifact of the releases repo not existing when we started, and part if it is that I haven't really seen non-official drivers using that process | |
| 16:48:39 | adreznec | markus_z: Not to say they can't, just haven't really seen a pattern for it | |
| 16:49:31 | jaypipes | bhagyashris: so, perhaps you are just confused by the fact that the "scheduler reporting client" doesn't actually talk to the scheduler :) | |
| 16:49:36 | markus_z | adreznec: You simply push a tagged commit to Gerrit in merge it, I assume? | |
| 16:49:44 | markus_z | s/in/and | |
| 16:50:31 | adreznec | Yep, we have our own releases team defined for the projects and handle branch creation and tag commit pushing through that ACL | |
| 16:51:06 | markus_z | adreznec: I see. Any major drawback with that approach? | |
| 16:52:11 | adreznec | markus_z: Not really that we've noticed, other than that the release cycle is completely self-enforced. | |
| 16:54:08 | markus_z | adreznec: yeah, sounds good enough IMO. I'll double-check with the release mgmt team though. Thanks! | |
| 16:54:27 | adreznec | markus_z: Ok, I'd be curious to see what feedback you get from them | |
| 16:54:44 | markus_z | adreznec: yep, will keep you in the loop | |
| 17:04:30 | openstackgerrit | Wang Qing wu proposed openstack/nova: WIP(5): Introduce base PowerVM driver (ovs vif) https://review.openstack.org/422512 | |
| 17:21:26 | mnestratov | hi all | |
| 17:23:34 | mnestratov | any chances to have some cores' eyes on https://review.openstack.org/#/c/312488/ and maybe get it merged before freeze? | |
| 17:29:57 | bauzas | mriedem: jaypipes: FWIW, I'll need to disappear a bit tonight because of a meetup so I could miss the nova meeting | |
| 17:30:49 | bauzas | mriedem: my two concerns are #1 the critical bug, but the success path is found | |
| 17:31:10 | bauzas | mriedem: #2 the placement upcall, but the ML thread is open with my reply, so I'd love thoughts on that | |
| 17:31:20 | jaypipes | bauzas: just finishing up response to the ML thread. | |
| 17:32:06 | bauzas | jaypipes: I really need to disappear in like 2 mins, but I'll check my emails | |
| 17:34:29 | jaypipes | bauzas: response sent (to mwhahaha not your reply) | |
| 17:34:47 | mwhahaha | jaypipes: i shall read and flame appropriately :) | |
| 17:34:59 | jaypipes | mwhahaha: I would expect nothing less :) | |
| 17:39:16 | bauzas | jaypipes: roger | |
| 17:39:30 | bauzas | jaypipes: I made two proposals, I'd love to see your thoughts | |
| 17:40:24 | bauzas | jaypipes: nevermind, saw your reply | |
| 17:40:26 | jaypipes | bauzas: k. if you could look at https://review.openstack.org/#/c/398473/ later, would be great. | |
| 17:40:55 | bauzas | jaypipes: oh that one is absolutely gorgeous | |
| 17:41:06 | jaypipes | lol | |
| 17:41:06 | bauzas | I'll try to look at it when I have time | |
| 17:41:39 | bauzas | jaypipes: I'm a serious lover of your RT cleanup series | |
| 17:41:55 | openstackgerrit | Jay Pipes proposed openstack/nova: placement: create aggregate map in report client https://review.openstack.org/407309 | |
| 17:42:02 | openstackgerrit | Roman Dobosz proposed openstack/nova: Prevent compute crash on discovery failure https://review.openstack.org/422780 | |
| 17:42:02 | jaypipes | bauzas: cool, thx :) | |
| 17:42:25 | bauzas | I just need to teleport now, see you folks | |
| 17:57:13 | dims | sean-k-mooney : jaypipes : gah! nova/nova/network/linux_net.py has diverged from os-vif/vif_plug_linux_bridge/iptables.py the os-vif one does not have any tests and is outdated :( | |
| 17:58:18 | jaypipes | dims: yeah, an unfortunate side effect of not having an oslo.netutils library :( | |
| 17:58:35 | sean-k-mooney | dims: is there a bug for that against os-vif | |
| 17:58:36 | dims | :) | |
| 17:59:00 | sean-k-mooney | also alot of the nova/virt/libvirt/vif.py is not tested ... | |
| 17:59:01 | dims | i just spotted it when i tried to use linuxbridge in a py35 job sean-k-mooney (http://logs.openstack.org/09/421209/16/experimental/gate-tempest-dsvm-nova-py35-ubuntu-xenial/15a4fc7/logs/screen-n-cpu.txt.gz?level=ERROR) | |
| 18:00:09 | dims | so back to devstack defaults now | |
| 18:00:53 | sean-k-mooney | dims: so is linux bridge not deployable currently? | |
| 18:01:04 | dims | not on py35 | |
| 18:01:06 | sean-k-mooney | dims: with python 3.5 | |
| 18:01:11 | sean-k-mooney | ok | |
| 18:01:13 | dims | correct | |
| 18:01:25 | sean-k-mooney | well not ok but we should adress it | |
| 18:02:52 | dims | sean-k-mooney : got this far with a number of depends-on. so yea, there's a bunch of work. i am trying to find a golden path that works (https://review.openstack.org/#/c/421209/) | |
| 18:02:56 | jaypipes | sean-k-mooney: you want to push a fix or should I? | |
| 18:03:01 | sean-k-mooney | jaypipes: the experimenatl job for os-vif seams to work fine so i would like to move it to the check pipepline. perhaps we should also add a linuxbridge varient once we fix it | |
| 18:03:11 | jaypipes | sean-k-mooney: ++ | |
| 18:04:18 | sean-k-mooney | jaypipes: do you have a linux bridge enviroment that you can test with? or a local.conf you can share. | |
| 18:04:57 | sean-k-mooney | if you can fix quickly go ahead other wise i can try and set up a dev environment but like wont get to it till next week. | |
| 18:05:29 | jaypipes | sean-k-mooney: I can fix the filter() thing quickly. don't have a LB env to use at the moment no. | |
| 18:07:32 | sean-k-mooney | dims: that is a faily deep dependcy tree for your experimental job | |
| 18:08:36 | sean-k-mooney | dims: are they all required? | |
| 18:09:16 | openstackgerrit | Jay Pipes proposed openstack/nova: placement: adds REST API for nested providers https://review.openstack.org/384807 | |
| 18:09:16 | sean-k-mooney | jaypipes: will the filter change need to be made to nova or os-vif | |
| 18:09:17 | openstackgerrit | Jay Pipes proposed openstack/nova: placement: allow filter providers in tree https://review.openstack.org/377215 | |
| 18:09:18 | openstackgerrit | Jay Pipes proposed openstack/nova: placement: integrate ProviderTree to report client https://review.openstack.org/415921 | |
| 18:09:18 | openstackgerrit | Jay Pipes proposed openstack/nova: pci: add uuid field to PciDevice object https://review.openstack.org/414768 | |
| 18:09:19 | openstackgerrit | Jay Pipes proposed openstack/nova: placement: adds ProviderTree for nested resources https://review.openstack.org/415920 | |
| 18:09:19 | openstackgerrit | Jay Pipes proposed openstack/nova: placement: SRIOV PF devices as child providers https://review.openstack.org/414769 | |
| 18:09:20 | openstackgerrit | Jay Pipes proposed openstack/nova: placement: add nested resource providers https://review.openstack.org/377138 | |
| 18:09:20 | openstackgerrit | Jay Pipes proposed openstack/nova: placement: update client to set parent provider https://review.openstack.org/385693 | |
| 18:09:40 | jaypipes | sean-k-mooney: os-vif. which means becaus eof the non-client freeze we need to do it today. | |
| 18:09:58 | jaypipes | sean-k-mooney: shouldn't be a hard fix. will work on it shortly. | |
| 18:10:02 | jaypipes | it's just a py35 thing. | |
| 18:10:21 | dims | sean-k-mooney : y, nothing too big in there, all small stuff. very close 1 failure left (http://logs.openstack.org/09/421209/15/experimental/gate-tempest-dsvm-nova-py35-ubuntu-xenial/b47de57/logs/testr_results.html.gz) | |
| 18:10:23 | sean-k-mooney | yes but it means we need to do a 1.4.1 release also correct | |
| 18:11:57 | sean-k-mooney | dims: no reason other then if i wanted to replicate locally its a little trickyer to recreate | |
| 18:13:48 | openstackgerrit | Vladik Romanovsky proposed openstack/nova: Adding vlans filed to Device tagging metadata https://review.openstack.org/405951 | |
| 18:13:49 | openstackgerrit | Vladik Romanovsky proposed openstack/nova: libvirt: expose virtual interfaces with vlans to metadata https://review.openstack.org/405954 | |
| 18:19:19 | jaypipes | dims: so you're already working on a fix for that os-vif thing? or should I do that? | |
| 18:20:04 | dims | jaypipes : please go ahead. | |
| 18:20:50 | jaypipes | dims: you got it :) | |
| 18:21:10 | dims | thanks much jaypipes ! | |
| 18:24:09 | openstackgerrit | Eric Fried proposed openstack/nova: Introduce base PowerVM driver (console) https://review.openstack.org/409402 | |
| 18:26:40 | openstackgerrit | Eric Fried proposed openstack/nova: WIP(4): Introduce base PowerVM driver (cfg drv) https://review.openstack.org/409404 | |
| 18:50:03 | jaypipes | dims: did you have a bug report about this os-vif thing? | |
| 18:50:14 | jaypipes | dims: if not, I'll add one. | |
| 18:50:38 | dims | jaypipes : no i have not added one | |
| 18:51:19 | jaypipes | dims: k, no problemo. | |
| 18:51:35 | dims | starting hard at logs :) | |
| 19:21:14 | mriedem | what os-vif issue do we have? | |