14:00:15 <tdurakov> #startmeeting Nova Live Migration
14:00:17 <openstack> Meeting started Tue Jan 24 14:00:15 2017 UTC and is due to finish in 60 minutes.  The chair is tdurakov. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:21 <openstack> The meeting name has been set to 'nova_live_migration'
14:00:29 <tdurakov> hi everyone
14:00:31 <scsnow> hi
14:00:34 <davidgiluk> hi
14:00:34 <siva_krishnan> hi
14:01:23 <mdbooth> Hi
14:01:35 <markus_z> o/
14:01:52 <tdurakov> agenda is the same https://wiki.openstack.org/wiki/Meetings/NovaLiveMigration
14:01:57 <tdurakov> so let's start
14:02:00 <tdurakov> #topic ci
14:02:45 <tdurakov> so I've asked  mriedem for help reviewing the serial console hook patch
14:02:56 <tdurakov> will ping him today
14:03:10 <tdurakov> since I think everything is ok there
14:03:11 <markus_z> I think I answered his questions in IRC and in the review itself.
14:04:03 <tdurakov> once that merged(hope it will be this week) we could move to https://review.openstack.org/#/c/421207/2
14:04:16 <tdurakov> as it goes next in a queue
14:05:18 <markus_z> I can take a look at it in the next days if you like.
14:05:50 <tdurakov> markus_z: review is very welcome
14:06:12 <scsnow> this is actually required for https://review.openstack.org/#/c/396218/
14:06:38 <tdurakov> right, as a test coverage
14:07:06 <markus_z> Right, I know the context of that. I will probably be Thursday when I can review them both.
14:07:13 <scsnow> in merge conflict right now as I don't want to rebase markus_z's patch once again. let's wait until it's merged.
14:07:44 <tdurakov> agree
14:07:49 <tdurakov> anything else on ci?
14:08:01 <tdurakov> pkoniszewski: are you around?
14:08:30 <tdurakov> does anyone know current status for the grenade job?
14:08:48 <markus_z> no, sorry, I'm out of touch
14:09:22 <tdurakov> will try to catch Pawel to figure that out, is there smth left before making the job voting
14:09:23 <tdurakov> ok
14:09:27 <tdurakov> let's move on
14:09:40 <tdurakov> #topic bugs
14:10:18 <tdurakov> I've found one that mostly affects liberty code, but is still valid for the master branch as well
14:11:10 <tdurakov> the thing is that we have disk-over-commit param for the lm(it's still valid for prev microversions)
14:11:35 <tdurakov> so, once anyone passed that param compute tries to verify disk space on dest node
14:12:10 <tdurakov> the problem here that it's could be negative, due to disk-over-commit > 1.0 in nova.conf
14:12:28 <tdurakov> that leads comparison to be done against negative value
14:12:40 <tdurakov> will file bug for that and prepare fix today
14:12:51 <tdurakov> any new bugs?
14:14:47 <tdurakov> ok, let's move on
14:14:55 <tdurakov> #topic pike specs
14:15:10 <tdurakov> please share your proposals here: https://etherpad.openstack.org/p/pike-live-migration
14:16:27 <tdurakov> next topic
14:16:35 <tdurakov> #topic open discussion
14:16:49 <tdurakov> do you have anything to bring up?
14:16:59 * mdbooth has been looking at this list https://etherpad.openstack.org/p/ocata-nova-priorities-tracking
14:17:09 <mdbooth> Is it current?
14:17:20 <mdbooth> The live migration bit, that is.
14:17:37 <mdbooth> Lines 93-113
14:18:25 <mdbooth> In particular, it seems that this change has been around forever: https://review.openstack.org/#/c/244489/
14:18:28 <davidgiluk> mdbooth: Do you know if that postcopy bug vs networking was fixed?
14:18:42 <mdbooth> davidgiluk: I haven't seen anything.
14:19:07 <tdurakov> mdbooth: yeah, the one with claims is a long story
14:19:16 <mdbooth> davidgiluk: Somebody else was working on it, but I can't remember who.
14:19:33 <siva_krishnan> I have patch for post copy bug
14:19:53 <siva_krishnan> davidgiluk: https://review.openstack.org/#/c/413021/9
14:19:59 <mdbooth> pkoniszewski: You're the most recent committer on the claims patch. Have you 'formally' taken it?
14:20:05 <davidgiluk> siva_krishnan: Ah great
14:20:29 <siva_krishnan> reviews are welcome :)
14:22:13 <tdurakov> siva_krishnan: what's the point in additional callback in a driver's method signature?
14:22:49 <mdbooth> siva_krishnan: I'm happy to look at those. The main problem I had when I was working on it was that nobody seemed to understand the network calls involved sufficient to change them.
14:22:52 <mdbooth> Certainly not me.
14:23:49 <mdbooth> tdurakov: I assume it's because network setup can be called in 1 of 2 places?
14:23:50 <siva_krishnan> tudurakov extracted network setup functionality and moved it right after post copy switch happens
14:24:49 <mdbooth> siva_krishnan: You could just call it immediately though, no?
14:25:04 <mdbooth> Anyway, we can have that discussion in review.
14:25:38 <mdbooth> With the claims patch, it seems that a bunch of people want that as it's a real bug.
14:25:59 <mdbooth> I'm guessing the lack of traction is mainly down to the complexity of claims?
14:26:15 <siva_krishnan> mdbooth: ok
14:26:57 <tdurakov> mdbooth, siva_krishnan, need to spend more time on that one, thanks for working on that, will leave a comment in review
14:27:42 <tdurakov> mdbooth: I assume you are right on that, I'm going to walk through it once again this week
14:28:08 <mdbooth> I've been looking at it, but I barely understand claims.
14:28:35 <mdbooth> I think I found a problem with it beyond just maintenance issues.
14:28:53 <mdbooth> But I won't be surprised to discover I missed something.
14:30:04 <pkoniszewski> mdbooth: sorry, was afk
14:30:20 <pkoniszewski> mdbooth: yes, I've taken it
14:30:41 <johnthetubaguy> need a callback in there, because thats how the virt driver calls code thats owned by the compute manager, generally speaking
14:30:58 <johnthetubaguy> just followed that success and failure callback patterns
14:31:07 <tdurakov> johnthetubaguy: is it about post-copy?
14:31:12 <johnthetubaguy> yeah
14:31:15 <tdurakov> ok
14:31:16 <johnthetubaguy> the callback
14:31:28 <johnthetubaguy> that mdbooth was mentioning
14:32:10 <tdurakov> pkoniszewski: could you please share update on grenade too?
14:32:40 <pkoniszewski> tdurakov: sure
14:32:58 <pkoniszewski> the grenade job is fixed and I haven't observed the issue for a while
14:33:11 <pkoniszewski> we still have patches up for review to enable back and forth live migrations in grenade job
14:33:43 <pkoniszewski> https://review.openstack.org/#/c/382456/ - Clark Boylan left -1 there and never came back
14:34:03 <pkoniszewski> he was asked multiple times to take another look
14:35:02 <tdurakov> will ping him or other infra folks for feedback
14:35:25 <tdurakov> afair I've left a comment with the alternative approach on that
14:35:35 <tdurakov> we just need to figure out which one to use
14:36:03 <tdurakov> ok
14:36:14 <tdurakov> anything else on lm?
14:37:12 <tdurakov> ok, thanks everyone for joining
14:37:18 <tdurakov> #endmeeting