20:00:32 <stevebaker> #startmeeting heat
20:00:33 <openstack> Meeting started Wed Sep  9 20:00:32 2015 UTC and is due to finish in 60 minutes.  The chair is stevebaker. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:00:34 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:00:37 <openstack> The meeting name has been set to 'heat'
20:00:41 <stevebaker> greetings everyone
20:00:45 <stevebaker> #topic rollcall
20:00:49 <dgonzalez> hi
20:00:55 <Drago> o/
20:00:55 <skraynev_> o/
20:00:57 <jasond> o/
20:01:25 <ryansb> o/
20:01:43 <stevebaker> zaneb: ping
20:02:13 <stevebaker> #topic Adding items to the agenda
20:02:15 <zaneb> o hai
20:02:16 <stevebaker> #link https://wiki.openstack.org/wiki/Meetings/HeatAgenda#Agenda_.282015-09-09_2000_UTC.29
20:03:20 <stevebaker> crickets
20:03:34 <stevebaker> #topic rc1 blueprint status
20:03:48 <skraynev_> #link https://launchpad.net/heat/+milestone/liberty-rc1
20:03:56 <stevebaker> skraynev_: beat me to it
20:04:30 <skraynev_> I just was prepared  ;)
20:04:33 <zaneb> he was just waiting there with his finger on the trigger
20:04:45 <skraynev_> yeah
20:04:57 <skraynev_> looks like this one is done https://blueprints.launchpad.net/heat/+spec/scaling-group-common ?
20:05:07 <skraynev_> zaneb ^ ?
20:05:10 <zaneb> so all the patches I proposed merged
20:05:12 <stevebaker> the blueprint feature freeze exception is entirely up to us, if it is in the rc1 list it has a FFE. Sing out if you have suggestions for any changes required
20:05:22 <zaneb> but I have other patches for that bp
20:05:43 <zaneb> never actually got around to posting the part where we create common scaling group code :)
20:05:52 <zaneb> that will be a mitaka thing
20:06:00 <zaneb> so I don't know what to do with the bp
20:06:03 <stevebaker> zaneb: shall we call that bp done for now?
20:06:08 <zaneb> I vote kick it to M
20:06:23 <zaneb> because what has landed is basically nothing but refactoring
20:06:40 <zaneb> there was like one or two improvements to the algo
20:06:42 <skraynev_> zaneb: maybe add comments with clarification current status and them mark as done
20:06:55 <skraynev_> and create new one in M
20:07:04 <skraynev_> with another description
20:07:20 <skraynev_> zaneb: ah see
20:07:31 <zaneb> skraynev_: we're running out of names for this one. I am not the first person to start on this ;)
20:07:39 <stevebaker> zaneb: It all landed after l3, so none of it will be in liberty unless it is backported to stable/liberty. I think it might be worth backporting anyway since it will make backporting scaling fixes easier during M
20:07:40 <skraynev_> in this case - moving bp to M is more reasonable
20:08:11 <zaneb> stevebaker: can we consider what I landed part of Oleksii's bp?
20:08:21 <zaneb> stevebaker: that's the only thing it's really needed for anyway
20:08:36 <zaneb> otherwise it's 98% refactoring
20:08:58 <stevebaker> zaneb: ok, what is that bp? its not in rc1 currently.
20:09:08 * zaneb looks
20:09:27 <skraynev_> zaneb: yeah. you have done solid platform for his bp
20:09:54 <stevebaker> zaneb: and is there any chance you could propose that whole series to stable/liberty? I'll check but I think approval is rubber-stamping
20:09:55 <zaneb> stevebaker: https://blueprints.launchpad.net/heat/+spec/resource-group-batching
20:10:31 <zaneb> stevebaker: I guess. can you go over what the release process is now? I thought we didn't open for Mitaka until RC?
20:10:43 <zaneb> so clearly I am behind the times
20:11:24 <stevebaker> oh, you're right. I see no stable/liberty
20:11:30 <stevebaker> well, that makes it easy
20:11:49 <stevebaker> zaneb: stand down!
20:12:00 * zaneb stands down
20:12:37 <zaneb> so, uh, probably a good time to remind folks not to land new features unless they have FFE from now on
20:12:55 <stevebaker> It would be preferable to land the rest of the BPs as early as possible, so cores please work off the rc1 milestone page when deciding what to review
20:13:09 <stevebaker> then review High rc1 bugs which are In Progress
20:13:40 <skraynev_> https://blueprints.launchpad.net/heat/+spec/convergence-simulator-tests - what about this one ?
20:13:55 <skraynev_> I think that patch with all tests was merged
20:14:30 <stevebaker> sweet, its in
20:14:51 <skraynev_> also this one is implemented too https://blueprints.launchpad.net/heat/+spec/get-attr-all-attributes
20:14:56 <skraynev_> already marked it
20:15:32 <skraynev_> I will ask Kanagaraj about https://blueprints.launchpad.net/heat/+spec/heat-resource-type-search tomorrow
20:15:39 <stevebaker> I got an update from shardy, https://blueprints.launchpad.net/heat/+spec/nested-validation will likely be close to the wire, and last to be ready for review
20:16:22 <stevebaker> Its also fairly low-risk, but we need to live with the data format "forever" so it would be good to get that right
20:17:09 <zaneb> stevebaker: FYI I moved my blueprint to future and ochuprykov's to liberty-rc1 and marked it complete
20:17:23 <stevebaker> zaneb: so did I
20:17:43 <stevebaker> #topic rc1 bug status
20:19:06 <stevebaker> shardy is still working on https://bugs.launchpad.net/python-heatclient/+bug/1224828, I think the current state needs some reviews
20:19:07 <openstack> Launchpad bug 1224828 in python-heatclient "heat stack-update wants all parameters and the template all over again" [High,In progress] - Assigned to Winnie Tsang (wtsang)
20:21:30 <stevebaker> not much to say here yet, please fix all the bugs, review all the fixes :)
20:21:50 <stevebaker> #topic state of the gate
20:22:45 <stevebaker> apart from the usual gate fun, we're being more impacted by races in our own tests
20:23:06 <stevebaker> #link https://bugs.launchpad.net/heat/+bugs?field.tag=gate-failure
20:23:54 <stevebaker> that list could probably do with some curating, if you know of a bug affecting our gate then feel free to tag it with gate-failure
20:24:39 <stevebaker> and by all means have a go at fixing any of ^ which are still an issue
20:25:13 <skraynev_> stevebaker: added this one https://bugs.launchpad.net/heat/+bug/1493700
20:25:14 <openstack> Launchpad bug 1493700 in heat "Functional test test_hook_pre_update_replace is failed" [Medium,New]
20:25:45 <stevebaker> skraynev_: thanks
20:26:17 <stevebaker> also, the ERROR engine log is very noisy with expected failures http://logs.openstack.org/24/218324/2/check/gate-heat-dsvm-functional-orig-mysql/3bce010/logs/screen-h-eng.txt.gz?level=ERROR
20:26:21 <skraynev_> stevebaker: I just follow therve advice about creating new bug per new fail on our gate ;)
20:26:35 <stevebaker> skraynev_: perfect
20:27:30 <skraynev_> stevebaker: do you have idea how to hide such expected raises ?
20:27:37 <stevebaker> I'm thinking a functional test run shouldn't log *anything* to ERROR, and we should eventually enforce that
20:27:42 <zaneb> stevebaker: hmm, that should be a bug. ERROR is supposed to be reserved for something the admin needs to pay attention to
20:28:00 <zaneb> ResourceFailure is definitely not ERROR
20:28:04 <zaneb> it's INFO at best
20:28:08 <stevebaker> skraynev_: Don't log.error ResourceFailure or WaidConditionTimeout
20:28:49 <stevebaker> ok, I'll raise a bug
20:29:54 <stevebaker> any finally on gate issues, it would be nice if the convergence job became voting during rc1.
20:29:56 <stevebaker> http://logs.openstack.org/24/218324/2/check/gate-heat-dsvm-functional-convg-mysql/1c55852/console.html
20:30:02 <stevebaker> Is anyone looking at those failures?
20:30:30 <zaneb> stevebaker: I don't think we can not log. it's tricky in resources, because we can't really distinguish between exceptions from the client and real coding errors
20:30:51 <stevebaker> zaneb: how about WARNING with no stack trace?
20:31:13 <zaneb> stevebaker: shouldn't be warning either though
20:31:23 <zaneb> I say INFO
20:31:43 <skraynev_> stevebaker: I suppose that HP guys constantly repair these tests
20:31:47 <zaneb> and the users will warn the admins if it's actually a warning ;)
20:32:17 <stevebaker> yeah, india and asalkeld are asleep, so this is the wrong week to ask that question
20:32:31 <skraynev_> stevebaker: I have seen one bug about some fails in convergence gate..
20:33:21 <stevebaker> #topic design summit pre pre planning
20:34:02 <stevebaker> We have 4 fishbowls, 8 workrooms and a half day meetup
20:34:21 <skraynev_> stevebaker: wow cool.
20:34:31 <zaneb> morning or afternoon?
20:34:41 <skraynev_> how about creating etherpad with session candidates ?
20:34:42 <stevebaker> friday morning meetup I think
20:34:49 <zaneb> excellent
20:35:05 <skraynev_> zaneb: +1 for the morning :)
20:36:22 <stevebaker> skraynev_: maybe just start thinking about sessions. I don't want to do too much before the new PTL, and I'd like us to focus on Liberty for just a bit longer ;)
20:37:13 <skraynev_> stevebaker: ok. I just suggested :)
20:37:30 <stevebaker> #topic Open Discussion
20:38:23 <stevebaker> #chair zaneb
20:38:23 <skraynev_> stevebaker: but yeah. it's really the best time to think enough about topics for sessions
20:38:25 <openstack> Current chairs: stevebaker zaneb
20:38:25 <stevebaker> #chair
20:38:26 <openstack> Current chairs: stevebaker zaneb
20:38:46 <zaneb> stevebaker: so, weirdly, the log level is already INFO: http://git.openstack.org/cgit/openstack/heat/tree/heat/engine/resource.py#n606
20:38:58 <skraynev_> I wanted to ask more review for rich-network stuff
20:39:12 <skraynev_> https://review.openstack.org/#/q/status:open+project:openstack/heat+branch:master+topic:rich-network,n,z
20:39:32 <stevebaker> skraynev_: +1, it would be good to land that so we can start testing it
20:39:46 <stevebaker> zaneb: I need to do school run, can you take over?
20:39:52 <zaneb> np
20:40:01 <stevebaker> back in 10 minutes
20:41:43 <skraynev_> nothing else from me ;)
20:41:45 <dgonzalez> There was a discussion about multi-cloud stacks on the mailing list yesterday. Is there already somebody working on this? Its a feature i am interested in and i would like to help with if possible
20:42:36 <zaneb> dgonzalez: I don't believe anyone is working on it yet
20:43:39 <dgonzalez> zaneb: Ok, so any idea how i could start with this? I guess i would have to take a look at how keystone federation works...
20:44:44 <zaneb> dgonzalez: as I said in the thread, you need to add an optional auth_url to the context, and find a way to get the right token
20:44:54 <zaneb> that should be it
20:44:58 <zaneb> SMOP ;)
20:45:36 <dgonzalez> zaneb: ok, that should be doable ;)
20:47:55 <zaneb> dgonzalez: I'll create a blueprint and post it to the list. then hopefully folks can use that to collaborate on specs and code, so that you're not all stepping on each other's toes
20:49:44 <dgonzalez> zaneb: ok thanks :)
20:51:41 <zaneb> discussion seems to have slackened off
20:51:48 <zaneb> shall we adjourn until next week?
20:52:05 <zaneb> #endmeeting