15:01:42 <ramishra> #startmeeting heat
15:01:43 <openstack> Meeting started Wed Oct 12 15:01:42 2016 UTC and is due to finish in 60 minutes.  The chair is ramishra. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:44 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:46 <openstack> The meeting name has been set to 'heat'
15:01:55 <ramishra> #topic roll call
15:02:04 <therve> heya
15:02:50 <spzala> hi
15:02:51 <ricolin> o/
15:03:11 <zaneb> \o
15:04:11 <duvarenkov> hi
15:04:55 <ramishra> #topic adding items to agenda
15:05:09 <ramishra> #link https://wiki.openstack.org/wiki/Meetings/HeatAgenda#Agenda_.282016-10-12_1500_UTC.29
15:06:20 <ramishra> #topic finalize heat summit schedule
15:06:48 <ramishra> I've created a draft schedule, not sure if it works for all of us.
15:07:00 <ramishra> #link https://wiki.openstack.org/wiki/Design_Summit/Ocata/Etherpads#Heat
15:07:50 <ramishra> I've not yet updated it in Cheddar that would update the summit schedule.
15:08:09 <therve> ramishra, Do we need a session on that maturity stuff?
15:08:37 <ramishra> There was only one fishbowl session that did not overlap with magnum and I've used that for the magnum session.
15:09:16 <ramishra> therve: zaneb suggested it, if we have something else to discuss.
15:09:36 <ramishra> may be the translation thing?
15:10:32 <therve> Maybe
15:10:34 <spzala> ramishra: is that in regards to heat translator we have?
15:10:44 <therve> spzala, Nope
15:10:44 <spzala> or something else?
15:10:53 <therve> Translation rules for properties
15:11:01 <spzala> therve: OK, thought so but wasn't sure
15:11:19 <zaneb> ramishra: can you remind me which of the sessions on that list are fishbowls vs. workrooms?
15:12:10 <ramishra> zaneb: the 2 performance related sessions and convergence-2 session are fishbowls
15:12:31 <ramishra> rest all are WR sessions
15:12:33 <zaneb> thanks
15:12:54 <ramishra> I don't know how it overlaps with tripleo sessions
15:13:24 <ramishra> some of us may be interested to attend some tripleo sessions.
15:13:39 <zaneb> no overlap with fishbowls by the looks of it
15:14:09 <zaneb> first two sessions on Friday conflict
15:14:20 <ramishra> I've kept the important ones (my opinion) on Thrusday;)
15:14:37 <ramishra> Most of the WR sessions on Friday
15:15:05 <zaneb> ramishra: this schedule looks good to me
15:15:37 <zaneb> sad to be missing the RPC versioning one, but there's nothing I would switch it with
15:16:49 <therve> Hum, not sure we can do it without you though
15:17:19 <ramishra> yeah I agree.
15:17:57 <ramishra> therve, zaneb: do we want to replace the maturity session with translation rules one?
15:18:26 <zaneb> ramishra: not sure there's much point if Peter is not going to be there
15:19:01 <therve> We can talk about translation during the meetup I guess
15:19:16 <ramishra> yep
15:20:03 <ramishra> so should I push this schedule?
15:20:28 <therve> sure
15:20:29 <ramishra> I assume we all agree that this is the best we can have.
15:21:09 <therve> Or that we can change the subjects anyway :)
15:21:59 <ramishra> ok, moving on
15:22:18 <ramishra> #topic stable releases
15:22:30 <ramishra> #link http://lists.openstack.org/pipermail/openstack-dev/2016-October/105302.html
15:23:02 <ramishra> It seems mitaka is moving to phase II and liberty to III support this Friday
15:23:37 <ramishra> So wanted to know, if we should have stable releases for them, I can see there are few patches still in the queue.
15:23:59 <ramishra> #link https://review.openstack.org/#/q/status:open+project:openstack/heat+branch:stable/mitaka
15:24:02 <ramishra> #link https://review.openstack.org/#/q/status:open+project:openstack/heat+branch:stable/liberty
15:24:59 <zaneb> we did releases for those pretty recently though, right?
15:25:04 <ramishra> We did a release few weeks back though.
15:25:16 <ramishra> zaneb: yep 4/5 weeks back
15:25:44 <duvarenkov> Sorry, I cant be present at meeting anymore today. Just wanted to draw some attention to patch about caching issue which I added to agenda. So, please share your thoughts about it and review :) I think, it is important one.
15:26:15 <zaneb> unless there's something critical in there I'd say there's probably no need then
15:26:17 <ramishra> If you see the discussion some projects think it's a good idea to have releases when the support phase changes.
15:26:29 <zaneb> I think the plan was to do more stable releases around o-1
15:28:20 <zaneb> ramishra: we've not been super-strict about enforcing the differences between phase 1/2/3 though
15:28:46 <zaneb> we should definitely try to get some of those open reviews merged
15:29:05 <ramishra> ok, sounds good, we can just leave it for o-1
15:29:58 <ramishra> #topic discuss caching issue
15:30:04 <therve> If we don't get the liberty one in, we can't release it AFAIU though
15:31:14 <ramishra> therve: I assume we can only do security fixes for liberty, right?
15:31:21 <therve> Yeah
15:32:11 <ramishra> IMO, doing a release does not hurt:)
15:32:28 <zaneb> therve: ah, because liberty is going to be closed?
15:32:50 <zaneb> then yeah, we should merge as many of the liberty patches as we can and release that again
15:32:51 <therve> zaneb, Well, phase 3
15:34:01 <zaneb> therve: yeah, I was confused by that. so they don't delete branches any more but allow security patches indefinitely?
15:34:19 * zaneb must have missed that memo
15:34:45 <ramishra> I think we've time till Friday to do whatever we want:)
15:35:09 <therve> zaneb, No, but the deletion is another concern IIRC
15:37:32 <ramishra> so, we should merege as many patches for liberty and cut a release and leave mitaka for o-1, is that what we agree on?
15:38:25 <zaneb> I'd be+1 for that
15:38:35 <therve> +1
15:39:39 <ramishra> ok, we can move on to the next topic, though duvarenkov has already left
15:39:55 <ramishra> #link https://review.openstack.org/#/c/366696/
15:40:42 <therve> So, I don't understand why resource names matter
15:40:45 <therve> Are they prefixed?
15:40:48 <therve> Aren't
15:41:11 <ramishra> prefixed with?
15:41:32 <therve> Stack ID?
15:42:10 <ramishra> I don't think so.
15:42:46 <ramishra> I think it's just the finder methods (with the arguments)
15:43:40 <ramishra> I don't like the idea of resource plugins cleaning the cache.
15:44:20 <ramishra> Btw, I'm for getting rid of the finder chache thing completely.
15:44:31 <zaneb> I don't like the idea of this cache existing at all, particularly if there's no invalidation and data routinely leaks between stacks
15:45:02 <ramishra> +1, or disabled by default?
15:45:15 <zaneb> I believe it's disable by default already?
15:45:16 <therve> Well it's disabled, as global caching is disabled
15:46:02 <zaneb> so that's something
15:46:21 <therve> I'm -1 on the current patch though, it's bad
15:46:43 <zaneb> but honestly, you can't have something like this not used by developers or in the gate and expect it to work (and indeed it apparently doesn't)
15:46:59 <therve> It's used in the gate
15:47:25 <ramishra> At the gate global cahce is enabled.
15:48:03 * zaneb is not looking forward to debugging that
15:49:17 <ramishra> I added that cache thing, so I'm to be blamed:(
15:50:16 <therve> Well you're the PTL, so you're to be blamed for everything
15:50:19 <therve> :)
15:50:36 <zaneb> I'm sure it seemed like a good idea at the time ;)
15:51:11 <ramishra> yeah, including poor attendance in meetings:)
15:51:48 <therve> Heh
15:52:23 <therve> Let's move on, 8mins
15:52:29 <ramishra> yeah, I added that for the RESOLVE translation thing, though we now do it before CREATE/UPDATE only, that has changed a lot.
15:52:32 <ramishra> yep
15:52:59 <ramishra> #topic memory usage
15:53:10 <zaneb> oh boy
15:53:32 <therve> zaneb, So it sounds like that update change is more complicated than expected
15:53:39 <therve> zaneb, Can you remove it for now?
15:53:48 <zaneb> therve: sigh, yes
15:53:52 <ramishra> probably update the team on what all we've been doing.
15:54:13 <therve> #link https://review.openstack.org/#/q/topic:bug/1626675
15:54:29 <therve> 3 patches in the queue, though one that doesn't work for unknow reasons
15:54:46 <therve> The other 2 that improve triple memory usage quite a bit (especially for the output one)
15:54:47 <zaneb> therve: tbh though, just improving the memory usage on create doesn't really help us though. at some point people will have to update, and if that hoses their undercloud it's just as bad
15:55:23 <therve> zaneb, Agreed, although we'll have a new release by then! :)
15:55:58 <therve> I haven't tested update yet, but I'm worried for sure
15:56:07 <ramishra> therve: for the rpc patch for outputs, you've used stack-show, is that for easy backporting?
15:56:12 <zaneb> therve: there's lots of reasons to do a stack update on the overcloud long before a new release of the undercloud
15:56:17 <zaneb> e.g. scaling up
15:56:23 <therve> ramishra, I don't know another method
15:57:01 <therve> zaneb, Yeah I know...
15:57:34 <ramishra> I think adding a argument 'resolve_outputs=False' to list_outputs would have worked too, no?
15:57:43 <therve> ramishra, Yeah that would have
15:57:49 <ramishra> though we would not be able to backport it
15:58:13 <zaneb> so just to confirm... we don't think that Heat has gotten worse post cwolferh's fix, just that TripleO has expanded to the point where they're even worse off than before that?
15:58:18 <therve> Yeah, and I don' t think it makes a big difference
15:58:52 <therve> zaneb, That's my (poor) explanation yes
15:59:05 <therve> Someone needs to bisect tht to confirm
15:59:20 <zaneb> I believe stevebaker did that
15:59:36 <ramishra> yeah, I had the same impression.
16:00:01 <inc0> hey guys...:) can we take over channel?
16:00:29 <ramishra> we still have 2 more mins, right?
16:00:33 <therve> No :)
16:00:40 <ricolin> no
16:00:44 <zaneb> ramishra: we can continue in #heat if necessary
16:00:48 <ramishra> ok, then thanks all.
16:00:56 <ramishra> #endmeeting heat