16:00:04 <smcginnis> #startmeeting cinder 16:00:05 <openstack> Meeting started Wed Apr 4 16:00:04 2018 UTC and is due to finish in 60 minutes. The chair is smcginnis. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:06 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:08 <openstack> The meeting name has been set to 'cinder' 16:00:14 <Swanson> Hi 16:00:18 <smcginnis> Ping list: DuncanT diablo_rojo, diablo_rojo_phon, rajinir tbarron xyang xyang1 e0ne gouthamr thingee erlontpsilva patrickeast tommylikehu eharney geguileo smcginnis lhx_ lhx__ aspiers jgriffith moshele hwalsh felipemonteiro lpetrut 16:00:25 <smcginnis> #topic rollcall 16:00:30 <xyang> hi 16:00:30 <geguileo> hi! o/ 16:00:35 <eharney> hi 16:00:47 <ganso> hello 16:01:29 <gouthamr> o/ 16:01:40 <e0ne> hi 16:01:49 <smcginnis> OK, seems a bit light, but we can get going. 16:01:50 <smcginnis> #link https://etherpad.openstack.org/p/cinder-queens-meeting-agendas 16:02:05 <smcginnis> Looks like we are currently around line 50. 16:02:10 <smcginnis> #topic Announcements 16:02:29 <smcginnis> I guess the only announcement I have is I am running the meeting while Jay is on vacation. :) 16:02:39 <ganso> smcginnis: you're the vice-PTL xD 16:02:46 <smcginnis> Hah ;) 16:02:55 <smcginnis> #link https://etherpad.openstack.org/p/cinder-spec-review-tracking Spec review tracking 16:03:03 <smcginnis> Normal reminder of the tracking etherpad. 16:03:13 <smcginnis> Not sure if it's been updated recently, but still some good things to check on there. 16:03:37 <smcginnis> It would be great to get attention to any outstanding specs that we discussed at the PTG that we would like to get in Rocky. 16:03:44 <_alastor_> o/ 16:04:11 <smcginnis> There are also a few new drivers, but I last I looked, most of them carried over from past cycles appear to have decided not to continue with them. 16:04:32 <smcginnis> Oh, I guess one more announcement in case it wasn't seen on the ML. 16:04:42 <smcginnis> The next release has been officially named "Stein". 16:05:06 <smcginnis> And all I have to say about that is marketing folks better take full advantage of that release name in Germany. 16:05:36 <ganso> Stein: a large earthenware beer mug. 16:05:47 <smcginnis> #topic Check-in on HA development progress 16:05:58 <smcginnis> geguileo: I think this is the point where Jay prods you. :) 16:06:11 <e0ne> :) 16:06:33 <e0ne> we were just talking about it in the cinder channel 16:06:46 <smcginnis> Ah. 16:06:57 <smcginnis> Anything we should record for the meeting logs? 16:07:32 <tbarron> hi 16:07:33 <e0ne> I need to update a spec for HA scheduler with more details 16:07:51 <smcginnis> e0ne: Do you have the spec link handy? 16:08:08 <e0ne> I think, geguileo can confirm that backups-via-scheduler shouldn't be blocked by HA scheduler 16:08:15 <e0ne> geguileo: am I right? 16:08:35 <e0ne> smcginnis: here is it https://review.openstack.org/#/c/556529/ 16:08:48 <e0ne> I'll update it with more details later this week 16:08:56 <smcginnis> Thanks. So anyone interested, please watch for updates to that patch. ^^ 16:09:21 <smcginnis> e0ne, geguileo: Anything else we should bring up in the meeting about HA for now? 16:09:32 * geguileo hides 16:09:36 <smcginnis> :) 16:09:38 <geguileo> smcginnis: unfortunately not yet 16:09:46 <e0ne> nothing for me. I hope, we'll have more details and discussion next week 16:09:57 <smcginnis> OK, thanks! 16:10:11 <e0ne> np 16:10:13 <smcginnis> #topic Stable policy changes 16:10:24 <smcginnis> I know Jay wants to be here for this discussion. 16:10:35 <e0ne> smcginnis: are we allowed to backport fixes for Pike now? 16:10:38 <smcginnis> So for now, just a heads up for everyone on changes to stable. 16:10:41 <smcginnis> #link https://review.openstack.org/#/c/552733 16:10:51 <e0ne> looks like 'yes', but I'm not sure yet 16:10:57 <eharney> so i do think we need to adjust things a bit, or at least solidify the thinking here 16:11:07 <smcginnis> e0ne: The overall policy has been relaxed, so we are now allowed to backport a little more than we used to. 16:11:22 <smcginnis> Part of it is up to the projects to decide. 16:11:30 <smcginnis> It is "allowed". 16:11:52 <smcginnis> eharney: I think that's the part Jay wanted to be here for. 16:12:06 <eharney> we've tended to be a bit too conservative with this, but we're trending in the right direction 16:12:07 <e0ne> so, we need to go the decision what should we backport to stable/pike or to driverfixes/pike 16:12:16 <smcginnis> Personally, I think we can relax things and allow backports for things that are low risk. 16:12:17 <eharney> no 16:12:24 <eharney> driverfixes/pike won't exist for now, right? 16:12:42 <smcginnis> eharney, e0ne: Right. Jay has a patch out there to create one, but I think that can be abandoned now. 16:12:43 <e0ne> eharney: that was my question 16:12:45 <eharney> so anything that we would have been landing in driverfixes needs to go into stble/pike instead 16:12:54 <eharney> stable* 16:12:55 <e0ne> eharney: +1 16:12:58 <smcginnis> We should be able to backport driver fixes. 16:13:22 <smcginnis> And now we should probably go through and make sure changes that went into driverfixes/ocata get over to stable/ocata. 16:13:44 <smcginnis> So for now, just watch and block any attempts to backport features or major changes. 16:13:51 <smcginnis> But I think we can allow other things. 16:14:02 <e0ne> smcginnis: it could be quite challenging 16:14:09 <smcginnis> And we should probably discuss next week when Jay is back to make sure things are clear. 16:14:14 <eharney> my wireless just died... maybe missed some things 16:14:19 <eharney> what's quite challenging? 16:14:19 <ganso> smcginnis: how about CI? 16:14:20 <smcginnis> e0ne: Getting ocata changes into stable? 16:14:42 <smcginnis> ganso: Our current policy doesn't require running against stable patches. 16:15:01 <smcginnis> But it is allowed and maybe encouraged if you can and you want to make sure nothing breaks. 16:15:04 <tbarron> so that's in line with what I'm doing atm: allowing fixes that *would* have gone to driverfixes to go to stable/{ocata,pike} 16:15:27 <tbarron> and thinking about maintaining them for other patches 16:15:28 <smcginnis> tbarron: Great! Yeah, I think it would be good for us to be consistent about what we are allowing. 16:15:36 <e0ne> smcginnis: yes. there shouldn't be merge conflicts, but I know how could it be painful for downstream maintainers to do such "backports" 16:15:46 <eharney> we will be considering more non-driver fixes, too, let's not just get stuck on the drivers here 16:15:51 <smcginnis> e0ne: Yeah, true. 16:16:00 <smcginnis> eharney: True as well. 16:16:07 <tbarron> eharney: agree, that's just the immediate and easiest case I think 16:16:12 <smcginnis> This does open things up to more than just driver fixes. 16:16:41 <eharney> which is really needed, because for a long time our stable branches have been too short lived and too conservative to really serve what people need from them 16:16:55 <tbarron> +1000 16:17:02 <e0ne> eharney: +1 16:18:26 <smcginnis> Part of the reasoning for this was we were EOLing branches right about the time downstream consumers were actually picking them up and finding issues. 16:18:33 <eharney> yep 16:18:47 <e0ne> +1 16:18:49 <Swanson> No real world user has ever gotten past Newton. 16:19:14 <smcginnis> Swanson: Hah, talking to ops folks, that's mostly a true statement. 16:19:31 <smcginnis> #topic Open discussion 16:19:39 <smcginnis> Anyone else have anything for today? 16:20:37 <smcginnis> I'll take that as a no. :) 16:21:02 <e0ne> see you next week 16:21:09 <Swanson> toodles 16:21:12 <smcginnis> Thanks everyone. 16:21:17 <smcginnis> #endmeeting