14:00:25 <tdurakov> #startmeeting Nova Live Migration 14:00:26 <openstack> Meeting started Tue Aug 30 14:00:25 2016 UTC and is due to finish in 60 minutes. The chair is tdurakov. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:27 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:29 <openstack> The meeting name has been set to 'nova_live_migration' 14:00:50 <tdurakov> hello everyone, agenda for the meeting: https://wiki.openstack.org/wiki/Meetings/NovaLiveMigration 14:00:55 <pkoniszewski> o/ 14:01:04 <luis5tb> o/ 14:01:12 <mdbooth> o/ 14:01:25 <tdurakov> let's wait a minute and then start 14:02:32 <tdurakov> so 14:03:15 <tdurakov> #topic newton-3 feature freeze 14:03:47 <tdurakov> do we have any patches except imagebackend, that requires attention and need review? 14:04:19 <tdurakov> imagebackend is the next topic) 14:04:35 * kashyap waves 14:05:17 <tdurakov> as I understand no, let's move on then 14:05:20 <johnthetubaguy> there was a stuck one the other day, let me take a quick peak 14:05:32 <tdurakov> johnthetubaguy: sure 14:05:45 <johnthetubaguy> oh, that is the imagecache one 14:05:55 <johnthetubaguy> carry on... :) 14:06:24 <tdurakov> #topic image backend 14:06:36 <tdurakov> mdbooth: anything except review request? 14:07:01 <mdbooth> So I don't think this is going to get there 14:07:05 <mdbooth> Just reviews 14:07:27 <mdbooth> johnthetubaguy: Which is the imagecache one? 14:07:42 <mdbooth> Is that the 1 with -1 from jay and +2 from somebody else? 14:07:48 <tdurakov> #action anyone to review imagebackend topic 14:08:00 <tdurakov> s/everyone 14:08:15 <mdbooth> It's really core review it needs at this point, tbh 14:08:25 <johnthetubaguy> I think the problem is we need that conflict resolving 14:08:35 <mdbooth> johnthetubaguy: Got a link? 14:08:45 <johnthetubaguy> https://review.openstack.org/#/c/334956 14:08:50 <johnthetubaguy> I mean the -1 from jay 14:08:59 <mdbooth> Ah, right. 14:09:08 <mdbooth> TBH, that mini-series isn't that important. 14:09:25 <johnthetubaguy> I think I pinged some image cache folks to take a peak, but they have not had chance 14:09:27 <mdbooth> I think it's correct, but I haven't made a fuss about it because it's not a dependency of anything. 14:09:37 <johnthetubaguy> ah, so maybe I am just getting distracted from the meat by that bit 14:09:54 <mdbooth> It was just a cleanup I did because I had my head in that code for other reasons and noticed that it was pretty obtuse. 14:09:58 <mdbooth> And I fixed it. 14:10:29 <tdurakov> mdbooth: tried to review this one: https://review.openstack.org/#/c/335966/16 tbh it's pretty complicated, tend to agree with dansmith about nested mocks 14:10:54 <mdbooth> Oh, I missed his comments. 14:11:02 <mdbooth> I actually wish he'd left a -1 :/ 14:12:42 <tdurakov> mdbooth: also related to your work https://review.openstack.org/#/c/355415/ 14:13:39 <tdurakov> I haven't asked Jens Rosenboom yet why he decided to leave this patch, as for me it's ok, and I think it could be merged before your refactoring 14:13:41 <mdbooth> I've already reviewed that. I was +1 iirc. 14:14:00 <mdbooth> Right. The refactor was the reason I was less fussed than mriedem, I think. 14:14:11 <mdbooth> But it's backportable. 14:14:28 <tdurakov> I'l ask Jens first, and if he don't mind update commit message 14:16:07 <tdurakov> ok, anything else? 14:16:27 <mriedem> so on https://review.openstack.org/#/c/355415/ 14:16:33 <mriedem> is anyone going to update it? 14:16:49 <mriedem> tdurakov: are you going to add a unit test? 14:17:04 <tdurakov> mriedem: yes 14:17:10 <mriedem> ok 14:17:19 <tdurakov> let's move on 14:17:24 <tdurakov> #topic CI 14:17:56 <tdurakov> live-migration job is pretty stable 14:17:59 <tdurakov> but 14:18:18 <tdurakov> I've spend some time trying to recheck/gather stats for nfs too 14:18:20 <tdurakov> https://review.openstack.org/#/c/329466/ 14:18:57 <tdurakov> so it looks like it's become unstable when nfs is enabled 14:19:22 <tdurakov> at the same time it's not reproducible on my local env 14:20:08 <tdurakov> I want to spend more time on that, or maybe consider enable ceph first 14:20:43 <mriedem> tdurakov: is it failing in any new way? 14:20:49 <mriedem> we could check just that change in logstash 14:21:00 <mriedem> i would agree with probably doing ceph before NFS 14:21:08 <mriedem> we don't test NFS anywhere else in upstream CI 14:21:11 <tdurakov> mriedem: yes, something new 14:21:33 <tdurakov> mriedem: acked, will try to do that 14:22:32 <tdurakov> let's move on 14:22:48 <tdurakov> #topic networking 14:23:13 <tdurakov> any updates on post-copy issues 14:23:47 <kashyap> tdurakov: Is there a specific bug you're referring to? 14:25:03 <kashyap> I think this - https://bugs.launchpad.net/bugs/1605016 14:25:03 <openstack> Launchpad bug 1605016 in OpenStack Compute (nova) "Post copy live migration interrupts network connectivity" [High,In progress] 14:25:04 <tdurakov> kashyap: 14:25:04 <tdurakov> https://bugs.launchpad.net/nova/+bugs?field.searchtext=post+copy&search=Search&field.status%3Alist=NEW&field.status%3Alist=INCOMPLETE_WITH_RESPONSE&field.status%3Alist=INCOMPLETE_WITHOUT_RESPONSE&field.status%3Alist=CONFIRMED&field.status%3Alist=TRIAGED&field.status%3Alist=INPROGRESS&field.status%3Alist=FIXCOMMITTED&field.assignee=&field.bug_reporter=&field. 14:25:04 <tdurakov> omit_dupes=on&field.has_patch=&field.has_no_package= 14:25:09 <tdurakov> ops 14:25:15 <tdurakov> https://bugs.launchpad.net/nova/+bug/1605016 14:25:25 <tdurakov> yeah, this one 14:26:03 <tdurakov> moved it to confirmed, as there is no assignee 14:27:21 <kashyap> tdurakov: Dave Gilbert mentioned that Luis (not sure of last name) was working on something (a spec) to address it? 14:27:26 <johnthetubaguy> hmm, I thought someone was working on that at one point 14:27:45 <johnthetubaguy> ah, I see they got unassigned 14:27:49 <luis5tb> I worked on the post-copy patch, but not on the fix 14:28:05 <johnthetubaguy> there is a wider thing in Neutron, tracking both port bindings (source and destination) 14:28:11 <johnthetubaguy> but we could fix it sooner 14:28:16 <luis5tb> I see it said it is in progress, did someone else started working on it? 14:28:34 <johnthetubaguy> comments say they have stopped working on it 14:28:52 <tdurakov> luis5tb: I've status already 14:29:12 <johnthetubaguy> I think its all about when we updating the port binding to point at the destination host, we do that today in post-live-migrate, we could do that sooner for post copy, to reduce the downtime 14:29:31 <johnthetubaguy> the proper fix is further away (tracking both source and destination port bindings) 14:30:00 <johnthetubaguy> I am hoping to get OSIC folks to help work on the nova bits for that next cycle, at least potentially 14:30:24 <johnthetubaguy> I added notes on a potential quick fix in the bug 14:31:23 <tdurakov> johnthetubaguy: thank you 14:31:59 <tdurakov> any update on setup_network_on_hosts for neutron api? 14:33:43 <tdurakov> so, let's move on 14:33:54 <tdurakov> #topic open discussion 14:34:30 <johnthetubaguy> there were comments on that patch 14:34:37 <johnthetubaguy> I don't remember seeing if there were resolved now 14:35:45 <johnthetubaguy> ah, it seems https://review.openstack.org/#/c/275073/ is ready to review again 14:37:05 <tdurakov> yep, reviews are welcome 14:37:13 <tdurakov> anything else? 14:37:59 <tdurakov> thanks everyone 14:38:04 <tdurakov> #endmeeting