11:57:16 <david-lyle> #startmeeting Horizon
11:57:17 <openstack> Meeting started Wed Apr 29 11:57:16 2015 UTC and is due to finish in 60 minutes.  The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot.
11:57:19 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
11:57:21 <openstack> The meeting name has been set to 'horizon'
11:57:32 <sambetts> Hi all o/
12:00:48 <david-lyle> will let people trickle
12:00:49 <david-lyle> in
12:02:54 <david-lyle> alright, let get going
12:03:04 <mrunge> hey o/
12:03:08 <akrivoka> hi
12:03:16 <robcresswell> o/
12:03:36 <bradjones> \o
12:03:45 <doug-fish> hello
12:04:00 <david-lyle> #topic General
12:04:51 <david-lyle> So as most saw, I added doug-fish, robcresswell and TravT to Horizon core, welcome and thank you
12:05:27 <doug-fish> thank you! I'm happy to be part of the team.
12:05:44 <mrunge> david-lyle, I was a bit astonished, as in the past current cores were asked to vote for this kind of change
12:06:11 <david-lyle> I forewent the usual informal vote on the mailing list to help reduce the overall number of emails, I reached out to as many cores as I could ping
12:06:24 <mrunge> ah, I see
12:06:49 <mrunge> seriously, new cores, you are more than welcome from my side
12:07:11 <robcresswell> Thanks :)
12:07:29 <david-lyle> I think these additions are overdue, but I was trying to create a lot of turmoil at the end of the release cycle
12:07:46 <david-lyle> not to create
12:08:06 <david-lyle> I wanted to save turmoil for now :)
12:08:16 <doug-fish> well done!
12:09:05 <david-lyle> We published RC-2 for Horizon last week
12:09:26 <david-lyle> and I haven't seen an issue that would block it being used as the final release
12:09:30 <mrunge> do we expect a rc3?
12:09:45 <robcresswell> david-lyle: Is this not a blocker? https://bugs.launchpad.net/horizon/+bug/1447288
12:09:45 <openstack> Launchpad bug 1447288 in OpenStack Dashboard (Horizon) "create volume from snapshot using horizon error" [Critical,In progress] - Assigned to Masco Kaliyamoorthy (masco)
12:09:46 <mrunge> david-lyle, I was able to reproduce the nova-network issue in launch instance
12:10:25 <doug-fish> robcresswell: yeah, I was going to bring up the same bug
12:10:27 <david-lyle> robcresswell: that one is marked critical, but I'm not sure I would classify it that way
12:10:51 <david-lyle> to be honest, I'm not sure it was working in the last 6-12 months
12:11:06 <david-lyle> looking at the changes in the appropriate files
12:11:20 <david-lyle> We can fix on master and backport
12:11:21 <doug-fish> isn't the problem that volumes created from snapshots don't have the expected contents?
12:11:40 <mrunge> imho this is not fixed here: https://bugs.launchpad.net/horizon/+bug/1444421
12:11:40 <openstack> Launchpad bug 1444421 in OpenStack Dashboard (Horizon) kilo "Launch instance fails with nova network" [High,In progress] - Assigned to David Lyle (david-lyle)
12:13:54 <david-lyle> mrunge: sorry reading through comments
12:14:04 <david-lyle> do you have a trace to add?
12:14:23 <mrunge> david-lyle, I proposed a fix
12:14:34 <mrunge> https://review.openstack.org/178163
12:14:54 <mrunge> with that fix, the issue doesn't happen any more
12:17:38 <mrunge> at least, one can create instances with nova network enabled
12:18:42 <david-lyle> hmm, that might require an rc3
12:19:11 <mrunge> is a rc3 really required?
12:19:30 <david-lyle> to get that fix into kilo it would be
12:19:40 <mrunge> ah, ok
12:19:50 <david-lyle> not being able to launch an instance on nova-net seems like a huge regression
12:20:10 <david-lyle> I'm not sure why that step is even being called
12:20:17 <mrunge> yes, same here
12:20:21 <david-lyle> is it network profile stuff?
12:21:01 <mrunge> this method is called from launch instance wizard initialization step
12:22:25 <mrunge> https://github.com/openstack/horizon/blob/master/openstack_dashboard/dashboards/project/instances/workflows/create_instance.py#L731
12:22:48 * mrunge assumes, permission check is wrong here
12:22:58 <mrunge> https://github.com/openstack/horizon/blob/master/openstack_dashboard/dashboards/project/instances/workflows/create_instance.py#L727
12:23:16 <mrunge> as this issue only occurs, when you have admin credentials
12:25:55 <david-lyle> so you have a network service in your catalog for admin for some reason
12:26:08 <mrunge> no I don't
12:26:49 <mrunge> if I'm listing keystone endpoints (even with admin credentials), I don't have a network service listed
12:28:22 <david-lyle> hmm, we're calling a templatetag for the permissions check for some reason
12:29:13 <david-lyle> but that hasn't changed since 2012
12:29:40 <mrunge> interesting
12:30:54 <david-lyle> and has_perms in d-o-a hasn't changed since 2013
12:31:04 <david-lyle> not sure what changed here
12:31:24 <mrunge> hmm, then I'll double check again here
12:31:42 <mrunge> and since nobody else could reproduce this
12:31:45 <mrunge> ...
12:32:14 <david-lyle> I set up devstack with nova-net and couldn't reproduce the first time either, but I'd like to know why you're seeing it
12:32:33 <david-lyle> if you can reconfirm, I'll set up the test env again and try again
12:32:52 <david-lyle> really don't want to ship with that broken
12:32:59 <david-lyle> ok
12:33:06 <mrunge> my backend is juno, but that shoulnd't harm anyways
12:33:25 <david-lyle> shouldn't think so
12:33:46 <mrunge> thanks
12:33:53 <david-lyle> that was the live bug triage portion of the meeting :)
12:34:01 <mrunge> uhm, I have something else to discuss
12:34:06 <david-lyle> sure
12:34:21 <mrunge> what happens with old bugs in launch instance workflow?
12:34:41 <mrunge> i.e. if I discover an issue in old laucnh instance in juno?
12:34:58 <mrunge> fixes are supposed to be in kilo or liberty first, then to be backported
12:35:14 <david-lyle> mrunge: until we're comfortable making the angular workflow the default, the old workflow is still the default
12:35:27 <mrunge> ok, no change yet
12:35:33 <doug-fish> I'd say we need to fix in master until we actually remove the old workflow and not allow it as an option
12:35:34 <david-lyle> I don't see us removing that code in L even if we switch
12:35:49 <mrunge> but sooner or later, that'd be the case
12:36:05 <david-lyle> if we do remove the old workflow, we would need to patch the stable branches directly
12:36:18 <mrunge> yes, makes sense
12:36:20 <david-lyle> rather than cherry pick
12:37:00 <mrunge> ok, thanks!
12:37:03 <david-lyle> sure
12:37:29 <david-lyle> doug-fish: the cinder bug is good for a backport to the first stable follow on to Kilo
12:37:47 <doug-fish> david-lyle: yes, agreed
12:38:09 <david-lyle> it's broken for sure, but at this point it's not destructive and I can't really determine that it's a regression
12:38:28 <david-lyle> final release is tomorrow
12:38:40 <david-lyle> so the bar is pretty high
12:38:45 <doug-fish> yeah understood
12:38:59 <david-lyle> then it's all liberty all the time
12:39:03 <doug-fish> it's important; but getting it in the first fix pack seems appropriate
12:39:05 <doug-fish> :-)
12:39:26 <david-lyle> more common items
12:39:52 <david-lyle> reminder that TC voting closes later today, eligible voters should have received an email
12:40:20 <david-lyle> this is your chance to pick over half of the TC membership for a year
12:42:08 <david-lyle> #topic Summit planning
12:42:34 <david-lyle> #link https://etherpad.openstack.org/p/horizon-liberty-summit
12:43:10 <david-lyle> we need to start voting on topics
12:43:29 <david-lyle> type and overall interest
12:43:59 <david-lyle> copied from the top of that etherpad
12:44:04 <david-lyle> 3 Fishbowls (Large crowd.)
12:44:04 <david-lyle> 8 Working sessions (Small room. 40 minutes)
12:44:04 <david-lyle> 2 Half day contributor meetups (All day Friday)
12:44:25 <david-lyle> so we should have plenty of time
12:44:35 <david-lyle> help prioritize the items
12:45:06 <david-lyle> Next week we would like to have the schedule set
12:45:17 <david-lyle> so people have time to prepare
12:45:25 <robcresswell> I'd really like to get some of the Angular sessions in. This is a good time to introduce people to the process, existing reusable components, and take away that "fear factor" from a new approach.
12:45:43 <david-lyle> and attendees can plan their schedule
12:45:59 <david-lyle> robcresswell: +1, maybe one working is intro to angular dev
12:46:24 <mrunge> yes, that makes sense
12:46:38 <robcresswell> Thats what me and tqtran are planning out. I'm working on docs now too; plan to have them merged by summit.
12:46:45 <david-lyle> robcresswell: is that on the etherpad?
12:46:49 <mrunge> there are new folks coming in, asking the same questions again and again...
12:47:02 <david-lyle> I see it on the etherpad
12:47:16 <robcresswell> Just under tqtran, at the bottom
12:47:19 <david-lyle> lin3 49
12:47:32 <robcresswell> Or under TravT. There is some duplication there.
12:47:39 <david-lyle> ah, Thai has more details
12:47:43 <david-lyle> always is
12:47:47 <david-lyle> :)
12:48:11 <david-lyle> after I write release notes today, I'll try to pare down the duplication at least
12:48:46 <robcresswell> But yes, plan is to get solid docs up, I'm aiming for first pass on Thursday, then the team will iterate over. At summit, if there is session time available, we will talk people through the process, explain docs, answer questions.
12:49:21 <robcresswell> Hopefully move away from this sub-team idea, and introduce wider community to the angular process.
12:49:35 <david-lyle> robcresswell: agreed
12:50:51 <david-lyle> I will say, we needed concrete progress this release, lest we go 18 months with nothing tangible from the efforts
12:51:07 <david-lyle> but we needed the effort to broaden
12:51:21 <david-lyle> hopefully we've learned from the process
12:52:03 <david-lyle> doug-fish: did you mark that HMT and Domain tokens were resolved in Kilo?
12:52:14 <doug-fish> hmm
12:52:16 <david-lyle> on the etherpad?
12:52:18 <doug-fish> I did
12:52:22 <doug-fish> maybe that was an overstatement?
12:52:32 <david-lyle> tell me how that turned out
12:52:34 <david-lyle> :)
12:52:45 <david-lyle> those two are still moving targets
12:53:01 <doug-fish> I've focused the resolved parts a bit more
12:53:01 <david-lyle> better
12:53:09 <david-lyle> glass half full
12:53:38 <david-lyle> ok, running out of time, vote on sessions you would like to see
12:53:46 <david-lyle> #Open Discussion
12:53:49 <doug-fish> It's kind of hard to vote right now
12:53:55 <doug-fish> there are overlaps
12:53:56 <david-lyle> #topic Open Discussion
12:54:05 <robcresswell> mrunge: I agree with your comment on etherpad. (I think it is yours). Plan is to get docs merged into repo, then will update relevant contributing docs, wikis etc.
12:54:06 <david-lyle> doug-fish: sure, will work on cleaning up
12:54:19 <mrunge> robcresswell, yes, it was mine
12:54:32 <mrunge> doug-fish, go ahead, condense topics
12:54:44 <mrunge> making easier to vote for topics
12:54:45 <david-lyle> oh yeah
12:55:00 <david-lyle> feel free
12:55:07 <david-lyle> nothing sacred
12:55:50 <david-lyle> try not to delete, if something seems off-topic or low priority just move to the bottom of the page
12:56:10 <doug-fish> I'm cutting + pasting to try to bring similar discussions together
12:56:40 <robcresswell> Seems like a good time to ask. Are there any rules/guidelines/advice for new cores, to make sure we're following the correct process?
12:57:04 <doug-fish> rules?!
12:57:12 <robcresswell> haha
12:57:38 <david-lyle> I typically send an email
12:57:42 <david-lyle> I will do that too
12:57:54 <robcresswell> Cool. Thankyou.
12:58:03 <david-lyle> I haven't done the access rights bits yet
12:58:12 <david-lyle> so I'm buying time
12:58:56 <david-lyle> ok, time's up. Thanks everyone for a great Kilo release. Looks like RC-2 will stick as final. Lots of hard work and reviews made this happen.
12:59:01 <david-lyle> #endmeeting