| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-03-27 | |||
| 13:18:29 | efried | We added some wrappers around our TaskFlow usage that log time spent in a particular Task (https://review.openstack.org/#/c/438729/11/nova/virt/powervm/tasks/base.py) | |
| 13:18:38 | efried | Those metrics would be neat to have natively. | |
| 13:27:37 | thorst | harlowja: also, huge fan of taskflow. Love it. | |
| 13:45:29 | edleafe | Nova scheduler meeting in 15 minutes in #openstack-meeting-alt | |
| 13:58:40 | sfinucan | johnthetubaguy: Correct | |
| 13:58:51 | sfinucan | self.stub_out is mock-based | |
| 13:59:25 | openstackgerrit | Sean Dague proposed openstack/nova master: WIP: proposed different wording on quota doc https://review.openstack.org/450244 | |
| 14:00:13 | johnthetubaguy | sfinucan: I was talking to the osic folks about focusing on the mox stuff instead, as we can leave all the old self.stub_out, for saftey | |
| 14:02:52 | openstackgerrit | Maxim Nestratov proposed openstack/nova master: libvirt: Virtuozzo containers config drive support https://review.openstack.org/449818 | |
| 14:03:23 | johnthetubaguy | sfinucan: I added a note on the spec | |
| 14:03:29 | sfinucan | johnthetubaguy: Yeah, that would make sense. iirc the 'stub_out' thing was something sdague wrote to ease migration from Mox. No reason to stop using it now | |
| 14:03:31 | sfinucan | spec? | |
| 14:03:48 | johnthetubaguy | sfinucan: I meant blueprint | |
| 14:04:06 | johnthetubaguy | sfinucan: yeah, self.stubs.Set just gets a simple re-write thats easy to review | |
| 14:04:29 | sfinucan | exactly | |
| 14:04:40 | sfinucan | johnthetubaguy: If you had an OSIC'er that wanted something interesting to do for a week or so: write a cliff plugin for Sphinx | |
| 14:04:52 | sfinucan | a la https://sphinx-argparse.readthedocs.io | |
| 14:05:25 | johnthetubaguy | sfinucan: they are all overbooked right now | |
| 14:05:39 | sfinucan | johnthetubaguy: Good to hear :) | |
| 14:21:20 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: do not log error for missing _save_tags https://review.openstack.org/450265 | |
| 14:27:47 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Using max api version in notificationt sample test https://review.openstack.org/443686 | |
| 14:28:55 | bhagyashris | johnthetubaguy: Hi, could you please take some insights on https://review.openstack.org/#/c/409644/ if you have free time. | |
| 14:28:56 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Using max api version in notificationt sample test https://review.openstack.org/443686 | |
| 14:30:01 | openstackgerrit | Stephen Finucane proposed openstack/nova master: conf: Move 'floating_ips' opts into 'network' https://review.openstack.org/431039 | |
| 14:30:01 | openstackgerrit | Stephen Finucane proposed openstack/nova master: conf: Deprecate 'default_floating_pool' https://review.openstack.org/431038 | |
| 14:30:02 | openstackgerrit | Stephen Finucane proposed openstack/nova master: conf: Add neutron.default_floating_pool https://review.openstack.org/431037 | |
| 14:30:02 | openstackgerrit | Stephen Finucane proposed openstack/nova master: conf: Rename '[neutron] ovs_bridge' https://review.openstack.org/450266 | |
| 14:33:41 | sfinucan | sdague, johnthetubaguy: Renamed 'ovs_bridge' instead ^ :) | |
| 14:40:35 | sean-k-mooney | sfinucan: regarding the ovs bridge name | |
| 14:40:48 | sean-k-mooney | that is now passed by neutron to nova | |
| 14:41:05 | sfinucan | sean-k-mooney: Yeah, that option is the default in case it's not needed | |
| 14:41:07 | sean-k-mooney | is this config being kept for nova-networks | |
| 14:41:09 | sfinucan | hence the rename | |
| 14:41:16 | sfinucan | See https://review.openstack.org/450266 | |
| 14:41:47 | sean-k-mooney | sfinucan: yes but my question is when do we need it? | |
| 14:41:50 | sfinucan | I'm going on the assumption that there are still instances where that won't be passed by neutron, though I don't know how I could go find out | |
| 14:41:52 | johnthetubaguy | I think its to support old neutron versions, I guess. At somepoint we can probably pull that, if we got microversion like API system | |
| 14:41:54 | sfinucan | Aha, jinx :) | |
| 14:42:31 | sfinucan | Yeah, I wouldn't even know where to start in investigating that | |
| 14:42:47 | sfinucan | ...besides just handing it off to neutron guys | |
| 14:43:02 | sean-k-mooney | ya as far as i know the openvsiwtch,ovn and odl drivers now pass the bridge as of newton | |
| 14:43:31 | sean-k-mooney | ill take a quick look and confirm | |
| 14:45:04 | sfinucan | sean-k-mooney: Awesome. If so we could drop that option entirely instead | |
| 14:46:13 | sean-k-mooney | sfinucan: im not seeing it in odl at first glance but it would be a small change to enable. | |
| 14:46:52 | openstackgerrit | Matthew Booth proposed openstack/nova master: Improve the readability of check_instance_state https://review.openstack.org/440536 | |
| 14:53:12 | bauzas | mriedem: dansmith: gentle reminder for https://review.openstack.org/#/c/391060/ | |
| 14:53:19 | bauzas | mriedem: do you want me to log a bug against that ? | |
| 14:54:01 | dansmith | ack | |
| 14:59:00 | mriedem | i thought i asked for a bug before, but can't remember | |
| 14:59:06 | mriedem | it at least needs a release note i think | |
| 14:59:10 | mriedem | but it is a bug imo | |
| 15:00:48 | bauzas | mriedem: I provided a note | |
| 15:00:51 | jaypipes | edleafe: pkgutils.walk_packages() thing looks cool. feel free to push a patch to os-traits on top of my refactor for the future one. | |
| 15:00:57 | bauzas | mriedem: but I can also provide a bug report | |
| 15:01:06 | artom | derekh, heya, bit of a weird blast from the past for you | |
| 15:01:18 | sean-k-mooney | sfinucan: it looks like only the ovs ml2 driver passes the bridge https://github.com/openstack/neutron/blob/master/neutron/plugins/ml2/drivers/openvswitch/mech_driver/mech_openvswitch.py#L101-L115 | |
| 15:02:07 | mriedem | dansmith: sdague: am i reading this correctly that in grenade, we don't attach a volume on the old side and then detach it on the new side? https://github.com/openstack-dev/grenade/blob/master/projects/60_nova/resources.sh | |
| 15:02:17 | mriedem | i see we add a floating ip on the old side | |
| 15:02:28 | sean-k-mooney | sfinucan: we would have to modify odl/ovn to also set it to remove that option unfortunetly. | |
| 15:02:32 | artom | derekh, remember https://bugs.launchpad.net/nova/+bug/1622545 ? Do you by any chance have similar dumps for older openstack versions? The one you provided in comment #6 has been very useful for our downstream testing, and we'd like to backport the fix all the way back to Juno | |
| 15:02:32 | openstack | Launchpad bug 1622545 in OpenStack Compute (nova) "archive_deleted_rows isn't archiving instances" [Medium,Fix released] - Assigned to Dan Smith (danms) | |
| 15:03:16 | mriedem | johnthetubaguy: i've left comments in your cinder v3 spec https://review.openstack.org/#/c/373203/ | |
| 15:03:17 | sean-k-mooney | sfinucan: or rather there ml2 drivers not ovn/odl themselves | |
| 15:03:39 | dansmith | mriedem: did you check cinder? might be in there | |
| 15:03:56 | mriedem | ooo looking | |
| 15:04:07 | derekh | artom: Hi, afraid not, this was a redeploy and the data from the old installation is long gone | |
| 15:04:14 | dansmith | mriedem: well, we have a BFV | |
| 15:04:27 | artom | derekh, figured. Oh well, it was worth a shot - thanks :) | |
| 15:04:33 | mriedem | ah https://github.com/openstack-dev/grenade/blob/master/projects/70_cinder/resources.sh#L130 | |
| 15:04:37 | sdague | mriedem: we have a bvf on the cinder side, but not an explicit attach after iirc | |
| 15:04:58 | mriedem | ok, | |
| 15:05:12 | mriedem | for johnthetubaguy's cinder v3 spec i think we want a grenade scenario where we attach on the old side and detach on the new side, | |
| 15:05:13 | derekh | artom: its always worth a shot, sorry I don't have anything useful | |
| 15:05:24 | mriedem | because the detach flow is going to depend on how we attached the volume in the first place | |
| 15:07:45 | mriedem | johnthetubaguy: on https://review.openstack.org/#/c/373203/ my main concern was around the swap volume steps setting bdm.attachment_id, | |
| 15:07:57 | mriedem | which would trigger the new detach flow before we actually wanted that i think | |
| 15:08:11 | mriedem | johnthetubaguy: otherwise just some questions and clarifications but i'd be +2 | |
| 15:08:14 | johnthetubaguy | mriedem: yeah, +1 the grenade these, that sounds good | |
| 15:08:42 | johnthetubaguy | mriedem: oh, thats the new swap flow, so it only applies when attachement_id is already set in the BDM right? | |
| 15:08:44 | mriedem | and i think on the shared connections section you could be more clear that we're not dealing with that in this spec, and why we are not of course | |
| 15:09:34 | mriedem | johnthetubaguy: it doesn't say that https://review.openstack.org/#/c/373203/21/specs/pike/approved/cinder-new-attach-apis.rst@265 | |
| 15:09:53 | johnthetubaguy | mriedem: yeah, it should say that | |
| 15:10:27 | johnthetubaguy | mriedem: basically, like all the other ones, keep the old flow if the attachment_id is not there | |
| 15:11:01 | mriedem | ok i didn't see that in the swap volume part, really threw me off | |
| 15:11:16 | mriedem | feel free to point out where i'm blind, otherwise we should probably clarify | |
| 15:11:22 | sfinucan | sean-k-mooney: Hmm, OK can you leave that in the review (if you haven't done so already)_ | |
| 15:11:46 | sfinucan | Not sure if that's worth doing. It might be. We should probably add a TODO in the code in any case | |
| 15:11:50 | mriedem | sdague: dansmith: johnthetubaguy: also, fyi, i'm traveling to CA for orientation this week, leaving in less than an hour, getting back home thursday night. | |
| 15:12:00 | mriedem | so i probably won't be around much the next few days | |
| 15:12:02 | sdague | ok, cool | |
| 15:12:04 | sdague | enjoy | |
| 15:12:12 | dansmith | mriedem: ack | |
| 15:12:19 | mriedem | i will not enjoy tucking my shirt in | |
| 15:12:25 | mriedem | or you know, wearing pants | |
| 15:12:30 | dansmith | heh | |
| 15:19:13 | johnthetubaguy | mriedem: good luck with orientation! | |
| 15:19:31 | johnthetubaguy | mriedem: sounds like a very good clarification we need in the spec | |
| 15:19:38 | sdague | don't forget your compass | |
| 15:22:08 | bauzas | mriedem: North is in that direction => | |
| 15:22:42 | sean-k-mooney | sfinucan: sure, i was planning on requing it as part of the migration to os-vif objects so that would be a logical time to implement it. | |