17:00:11 <ildikov> #startmeeting cinder-nova-api-changes
17:00:12 <openstack> Meeting started Thu Mar 30 17:00:11 2017 UTC and is due to finish in 60 minutes.  The chair is ildikov. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:13 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:16 <openstack> The meeting name has been set to 'cinder_nova_api_changes'
17:00:21 <jungleboyj> o/
17:00:27 <smcginnis> o/
17:00:28 <ildikov> DuncanT ameade cFouts johnthetubaguy jaypipes takashin alaski e0ne jgriffith tbarron andrearosa hemna erlon mriedem gouthamr ebalduf patrickeast smcginnis diablo_rojo gsilvis  xyang1 raj_singh lyarwood breitz jungleboyj
17:00:42 <hemna> \o
17:00:58 <lyarwood> o/
17:01:04 <jungleboyj> hemna:  Look at you and your right handedness.
17:01:19 <ildikov> hi all :)
17:01:21 <hemna> well I am right handed
17:01:55 <ildikov> let's wait a bit to see whether we have more people joining
17:01:59 <johnthetubaguy> so daylight savings has just screwed me on this timing
17:02:13 <jungleboyj> johnthetubaguy:  So the UK is suffering this week?
17:02:15 <johnthetubaguy> so I am only around for a little bit
17:02:19 <ildikov> johnthetubaguy: I was wondering about that just today
17:02:20 <johnthetubaguy> yeah, its our week of pain
17:02:30 <johnthetubaguy> although the weather is quite nice today
17:02:38 * jungleboyj looks at johnthetubaguy sympathetically.
17:02:43 <ildikov> johnthetubaguy: is it today or this slot does not work for you in general?
17:02:51 <jungleboyj> We have had clouds most of the last two weeks.
17:03:02 <lyarwood> johnthetubaguy: I think this time is set now, the US moves to DST ahead of us moving to BST
17:03:31 <johnthetubaguy> ildikov: its a thursday thing really, its usally bad now
17:03:43 <johnthetubaguy> lyarwood: yeah
17:04:02 <ildikov> johnthetubaguy: ok, we can look into switch back to the Monday slot we used earlier or smth like
17:04:18 <johnthetubaguy> I normally have a rehersal to get to, so I try stop work early on a thursday about now (6pm)
17:04:29 <johnthetubaguy> I mean, maybe, I am not required as such now
17:04:41 <johnthetubaguy> the spec is merged, I just have stuff to review
17:04:44 <johnthetubaguy> (in theory)
17:04:55 <ildikov> johnthetubaguy: well, we have more Nova items now and I think we don't have Matt today either
17:05:18 <johnthetubaguy> yeah, matt is out today, being welcomed into his new company
17:05:28 <ildikov> johnthetubaguy: I think the next one now is lyarwood's detach refactor, I think that review is still open
17:05:37 <smcginnis> Yay for the spec being merged.
17:05:41 <johnthetubaguy> yeah, so whats the hit list for the next week
17:05:49 <johnthetubaguy> smcginnis: +1 on the yay :)
17:05:50 <ildikov> I rebased jgriffith's detach PoC on top of that
17:05:59 <jungleboyj> smcginnis: +@
17:06:00 <lyarwood> ildikov: yeah the review is open, I updated it this afternoon and it now has a merge conflict so I'll address this after the call
17:06:04 <jungleboyj> +2 even
17:06:06 <ildikov> smcginnis: johnthetubaguy: +1 :)
17:06:21 <smcginnis> Gotta celebrate our victories where we can. :)
17:06:22 <ildikov> lyarwood: ok, thanks
17:06:31 <ildikov> :)
17:07:02 <lyarwood> smcginnis: is your detach change ready to review btw?
17:07:14 <lyarwood> ah sorry
17:07:21 <lyarwood> jgriffith: ^
17:07:57 <ildikov> lyarwood: it would be great to get review on that to see whether the direction is what we want
17:08:18 <smcginnis> lyarwood: Do you have a link to your patch?
17:08:23 <smcginnis> ...handy
17:08:30 <lyarwood> ildikov: yeah I couldn't tell if it was still WIP tbh, there are a few nits in there that I think we will need to address
17:08:44 <ildikov> lyarwood: I think the nova/virt/block_device.py changes are messed up, that was me, I was clearly not really sure at what I'm doing there...
17:08:53 <lyarwood> smcginnis: https://review.openstack.org/#/c/439520/
17:09:14 <lyarwood> moving forward I'd like to keep the nova patches against the https://review.openstack.org/#/q/topic:bp/cinder-new-attach-apis topic
17:09:25 <smcginnis> lyarwood: Good call.
17:09:32 <ildikov> lyarwood: please add comments to the patch or hunt jgriffith and/or me down on IRC if there are more complicated parts
17:09:58 <lyarwood> ildikov: will do thanks
17:10:08 <ildikov> lyarwood: me thanks
17:10:47 <lyarwood> johnthetubaguy: what was next area we wanted to look at in nova? (/me opens the spec)
17:10:58 <ildikov> johnthetubaguy: there are small changes up to deprecate Cinder v2 and remove check_dteach
17:11:01 <johnthetubaguy> to be honest, attach is last
17:11:06 <johnthetubaguy> everything else comes first
17:11:11 <ildikov> johnthetubaguy: reviews appreciated, both are small changes
17:11:24 <lyarwood> johnthetubaguy: kk
17:11:28 <johnthetubaguy> ...but its tempting to get attach POC-ed so we can get the tests passing on that
17:11:51 <johnthetubaguy> like a DNM patch with attach, or something, to prove detach actually works
17:11:57 <ildikov> johnthetubaguy: lyarwood: I had a short chat this week with mdbooth about the lock we were talking about on the PTG a bit
17:12:10 <ildikov> we can chat about that at some point
17:12:14 <johnthetubaguy> ildikov: we should agree next steps for this week first
17:12:57 <ildikov> I will look up the attach PoC as we had a patch up for that, but I think it's either abandoned or just seriously outdated
17:13:08 <lyarwood> do we need check_detach removed before we land the v3 detach calls?
17:13:20 <ildikov> johnthetubaguy: sure
17:13:37 <ildikov> lyarwood: I think it's not crucial for those
17:13:46 <lyarwood> ildikov: kk just checking
17:14:02 <johnthetubaguy> its a good question
17:14:03 <ildikov> lyarwood: it's a simple cleanup which does not conflict with those as far as I can remember
17:14:17 <ildikov> lyarwood: nah, totally valid question, thanks for bringing it up
17:14:20 <johnthetubaguy> if we are happy with that, it might be good getting it landed
17:14:34 <ildikov> I was just thinking out loud
17:14:37 <johnthetubaguy> for the sake of simplicity if nothing else
17:14:49 <johnthetubaguy> you got the link for that one?
17:14:52 <ildikov> johnthetubaguy: https://review.openstack.org/#/c/446671/
17:15:00 <ildikov> lyarwood: ^^
17:15:28 <lyarwood> thanks
17:15:31 <johnthetubaguy> did we revert the one we landed or something?
17:15:46 <ildikov> johnthetubaguy: revert?
17:16:37 <johnthetubaguy> so I thought we had already merged a version of this patch, for some reason
17:16:45 <johnthetubaguy> I think I got two patches mixed up in my head
17:16:48 <lyarwood> check_attach
17:16:51 <lyarwood> I think
17:16:54 <smcginnis> There was something just after the PTG.
17:16:58 <ildikov> johnthetubaguy: that was check_attach
17:17:06 <johnthetubaguy> oh right... doh!
17:17:14 <ildikov> johnthetubaguy: we have check_detach too, which is just as unnecessary IMHO :)
17:17:56 <johnthetubaguy> well, we like giving folks nice error messages
17:18:13 <johnthetubaguy> the general concept doesn't seem required
17:18:21 <ildikov> johnthetubaguy: I don't think that's mutually exclusive :)
17:18:24 <johnthetubaguy> so my food is ready, so I kinda have to go eat it
17:18:27 <johnthetubaguy> ildikov: agreed
17:18:30 <ildikov> or at least I tried not to mess it up too much
17:18:52 <ildikov> johnthetubaguy: ok, then I would say let's go with the detach related changes
17:18:57 <ildikov> for next week
17:19:00 <johnthetubaguy> lyarwood: I think we just ignore that patch, unless it really helps detach look cleaner I guess?
17:19:14 <johnthetubaguy> lyarwood: I feel too far from the code to tell from a quick glance right now
17:19:35 <ildikov> and we can sync up if you would run out of reviews before the next meeting
17:19:54 <johnthetubaguy> yeah, sounds like we have things to get on with for this week
17:20:11 <johnthetubaguy> #link https://review.openstack.org/#/c/439520
17:20:12 <lyarwood> johnthetubaguy: yeah would be nice to sort out at some point this cycle tbh
17:20:22 <johnthetubaguy> #link https://review.openstack.org/#/c/438750
17:20:28 <ildikov> lyarwood: +1
17:20:43 <johnthetubaguy> lyarwood: probably best just to decide based on how the detach patch looks thats on top of yours
17:21:10 <ildikov> lyarwood: small change, should not hurt, would be great not to carry it for multiple releases
17:21:12 <johnthetubaguy> the tidy up might help that land quicker, so thats all good
17:21:34 <lyarwood> ildikov: so I'm also happy to look into the attach POC btw if you don't have time
17:21:59 * johnthetubaguy runs away to eat his dinner
17:22:00 <ildikov> johnthetubaguy: with lyarwood's change it's a change at one place for detach
17:22:08 <ildikov> johnthetubaguy: tnx for joining
17:22:12 <ildikov> johnthetubaguy: enjoy
17:22:14 <johnthetubaguy> yeah, it might be independent now
17:22:15 <johnthetubaguy> thats cool
17:24:00 <ildikov> lyarwood: https://review.openstack.org/#/c/330285/
17:24:17 <ildikov> lyarwood: I think that's the one that contained changes to make attach work too
17:24:47 <ildikov> lyarwood: not updated for a while, so I think it has attachment_id magic in it too, etc. so don't expect nice and shiny code
17:24:59 <ildikov> lyarwood: it worked at a time, but things have changed since then
17:25:13 <lyarwood> ildikov: yeah I think it had both in there and used attachment_uuid
17:25:17 <ildikov> who gets there first can update it and clean up I would say
17:25:18 <lyarwood> ildikov: so we will need to split it out
17:25:23 <lyarwood> ildikov: kk
17:25:42 <ildikov> agreed, I focused on the detach first so it can be reviewed now
17:26:27 <ildikov> we also have a ton of use cases to look into and I think having smaller patches for those would be good
17:27:45 <lyarwood> ildikov: yup agreed and we should be able to test these as we go with an attach POC in place
17:29:00 <ildikov> lyarwood: yep that sounds good
17:30:40 <ildikov> I kind of out of topics for today
17:30:53 <lyarwood> same, nothing else from me.
17:30:55 <ildikov> I think the next steps are clear for now
17:31:05 <ildikov> we have some changes on the Cinder side
17:31:16 <ildikov> like switching from instance-id to server-id
17:31:38 <ildikov> but that will not affect detach so we're good
17:32:05 <lyarwood> is that something new for attach?
17:32:28 <ildikov> attachment-create will change because of that
17:32:49 <ildikov> but attachment-update needs only an attachment-id for instance, so it's not a major change
17:32:55 <lyarwood> kk thanks, I missed that in the logs from last week
17:33:04 <ildikov> we're trying to clean the API up on the Cinder side before it gets used
17:33:17 <ildikov> lyarwood: it came up this week
17:33:39 <ildikov> lyarwood: patches are already up and we try to get them in quickly so when the PoC activity gets there it should be baked
17:34:16 <lyarwood> ah understood, thanks
17:34:37 <stvnoyes> git status
17:34:51 <ildikov> sure, np
17:35:12 <ildikov> stvnoyes: I hope you're checking a related code change :)
17:35:17 <stvnoyes> sorry, stray text ^
17:35:29 <jungleboyj> Somebody get that man some status!
17:35:35 <ildikov> stvnoyes: no worries, we've all been there :)
17:35:53 <stvnoyes> not the first time for me :-)
17:36:17 <ildikov> stvnoyes: haha, still could be worse :)
17:36:32 <smcginnis> Your branch is up-to-date with 'origin/master'.
17:36:38 <smcginnis> nothing to commit, working directory clean
17:36:43 <bmace> lol
17:36:52 <ildikov> smcginnis: lol
17:37:02 <stvnoyes> well I am happy to added some entertainment to the meeting ;-)
17:37:09 <smcginnis> ;)
17:37:21 <ildikov> stvnoyes: that's always welcomed as far as I'm concerned :)
17:37:22 <jungleboyj> smcginnis:  Thank you!
17:37:54 <ildikov> ok, now that we had some fun too I really don't have more for today
17:38:25 <ildikov> anyone with any questions, comments?
17:38:35 <jungleboyj> Sounds good.  Thanks ildikov
17:38:56 <ildikov> ok, then let's progress on both sides as agreed
17:39:02 <lyarwood> thanks ildikov
17:39:32 <ildikov> I will chat with johnthetubaguy on the meeting slot and let you know if we need a change
17:39:51 <ildikov> see you on the channels and here next week I think still at this slot
17:39:59 <ildikov> thank you all!
17:40:12 <ildikov> #endmeeting