12:00:41 <david-lyle> #startmeeting Horizon
12:00:42 <openstack> Meeting started Wed Apr  1 12:00:41 2015 UTC and is due to finish in 60 minutes.  The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:00:43 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
12:00:45 <openstack> The meeting name has been set to 'horizon'
12:00:59 <david-lyle> hello everyone
12:01:56 <david-lyle> time change lose everyone?
12:01:56 <robcresswell> o/
12:02:15 <robcresswell> Seems like it
12:02:39 <pkarikh> Hi!
12:02:49 <samueldmq> hello
12:02:56 <sambetts> o/
12:03:50 <rdopiera> hi
12:04:55 <david-lyle> We're in the middle of of the RC1 milestone.
12:05:25 <david-lyle> We had 6 FFEs, 4 merged, 1 is close and 1 has been deferred.
12:05:37 <david-lyle> hopefully the last one will merge this week
12:05:46 <david-lyle> https://launchpad.net/horizon/+milestone/kilo-rc1
12:06:27 <david-lyle> The Federated Identity one is mostly an enablement for work happening in django_openstack_auth
12:07:22 <david-lyle> We do have some bugs in RC1,  but almost all have a patch up
12:08:01 <david-lyle> This is time to make sure if you see a critical bug that should effect the release, to raise its visibility
12:08:51 <david-lyle> There is another bug that I have yet to file, that I started working on yesterday where django 1.7 breaks our selenium tests, so we are blocked from updating our requirements.txt
12:09:07 <tsufiev> david-lyle, hi! Speaking of important bugs I think that https://bugs.launchpad.net/horizon/+bug/1427819 definitely should be fixed as part of Kilo
12:09:08 <openstack> Launchpad bug 1427819 in OpenStack Dashboard (Horizon) "Downloading large objects from swift containers exhausts web-server memory" [High,In progress] - Assigned to Timur Sufiev (tsufiev-x)
12:09:08 <david-lyle> I'll file that this morning
12:10:52 <david-lyle> tsufiev: it's targeted for RC1 and is marked high. Looks like the check jobs are failing, but once they're not, I don't see a reason it can't merge
12:11:16 <tsufiev> david-lyle, I'll fix one failing test in a couple of hours
12:12:07 <david-lyle> So getting a stable RC should be the top priority right now.
12:12:27 <david-lyle> After RC1 is cut, master will be open for Liberty features
12:13:04 <david-lyle> Translations are the only thing targeted for RC-2
12:13:22 <david-lyle> any questions/comments on RC1?
12:13:44 <samueldmq> no, I think you're good for horizon rc1 :)
12:14:16 <david-lyle> Next general item, summit
12:15:07 <david-lyle> Looks like we should get all the sessions I requested, 3 fishbowl, 8 working sessions, and a full day of collaborator meetup
12:15:34 * david-lyle can't quite remember if I asked for full or half day, but believes full
12:15:34 <tsufiev> fishbowl?
12:15:41 <david-lyle> traditional sessions
12:15:59 <tsufiev> ah )
12:16:39 <david-lyle> we have a proposed topic list here https://etherpad.openstack.org/p/horizon-liberty-summit
12:17:03 <david-lyle> feel free to add/augment content there
12:18:16 <david-lyle> There is not a formal agenda posted today, but I know samueldmq had an item
12:18:28 <david-lyle> #topic hierarchical projects
12:18:31 <samueldmq> o/
12:18:38 <samueldmq> I have been working on adding support for hierarchical projects on Horizon
12:18:47 <samueldmq> I thought it was good to add a point to this meeting to let you know what is going on
12:19:13 <samueldmq> we are working together with Piet and his team, and we have  a fist implementation
12:19:30 <samueldmq> that will be probably change :)
12:19:32 <samueldmq> first, a blog post explaining what was implemented
12:19:40 <samueldmq> #link www.samueldmq.com/hierarchical-projects-on-horizon/
12:19:55 <samueldmq> also, we have a demo running on 150.165.15.68, with Horizon credentials admin/nomoresecrete
12:20:10 <samueldmq> this service will be kept until Saturday, feel free to try it out and give some feedback
12:21:22 <david-lyle> this is something we will be discussing at the summit, but it would be great if people checked it out now to become more familiar with the topic
12:21:38 <david-lyle> and provide any UX feedback
12:21:55 <samueldmq> david-lyle, ++ exactly
12:22:11 <david-lyle> so please take a look
12:22:23 <david-lyle> samueldmq: what's the best way to get feedback to you?
12:22:38 <samueldmq> also, the screen for the List Project table will change for sure with the new angularized tables
12:22:49 <samueldmq> david-lyle, maybe comments on the patches
12:22:59 <samueldmq> david-lyle, there links are in the blog post
12:23:07 <david-lyle> ok, at the bottom of the page
12:23:10 <david-lyle> good point
12:23:27 <samueldmq> #link https://review.openstack.org/#/c/119469/
12:23:34 <samueldmq> the patches start on this one ^
12:23:53 <david-lyle> for most projects this will not be a very invasive change, other than around quotas
12:23:55 <samueldmq> and we put them as dependent, so just need to look at dependencies on it
12:24:26 <david-lyle> for horizon this will effect a lot more and we should take the time to understand it and provide feedback
12:24:45 <samueldmq> david-lyle, in the blog post I tried to say how other projects can take advantage of this
12:25:04 <samueldmq> david-lyle, yes I agree we have to get the time to implement this on horizon
12:25:28 <samueldmq> david-lyle, we need to talk to folks from other projects to know how it should be for them
12:26:02 <david-lyle> ok, anything else samueldmq?
12:26:16 <samueldmq> david-lyle, no, that's all :)
12:26:18 <samueldmq> thanks
12:26:28 <david-lyle> Thanks!
12:26:44 <david-lyle> #topic Open Discussion
12:27:50 <tsufiev> david-lyle, I've been talking with my colleagues about recent failure of their tests for muranodashboard (Horizon plugin) due to Django 1.6->1.7 req...
12:28:21 <rdopiera> tsufiev: you might want to repeat
12:28:48 <rdopiera> or not...
12:28:51 <david-lyle> tsufiev: ok, any insights
12:29:03 <david-lyle> rdopiera: second IRC window
12:29:27 <tsufiev> rdopiera, they said failures were caused by the Django requirement being changed in devstack global-reqs
12:29:34 <tsufiev> david-lyle, ^^^
12:29:45 <tsufiev> missed the addressee )
12:29:54 <david-lyle> yes, but the problem is between horizon and selenium
12:30:17 <tsufiev> so it appeared weird to me: the change itself didn't make it to the horizon repo, yet it was accepted in devstack
12:30:47 <tsufiev> do we have some influence over the devstack changes, I mean any guarantees of Horizon being stable there?
12:31:05 <david-lyle> there is a missing STATIC_ROOT setting in horizon/test/settings, but beyond that Selenium can no longer find objects on the page
12:31:16 <david-lyle> the pages aren't rendering
12:31:29 <david-lyle> I worked on it part of the day yesterday and plan to resume today
12:31:51 <david-lyle> tsufiev: the running code should work, it's just selenium testing that is blocking us now
12:32:00 <david-lyle> and some of those pass too
12:32:14 <david-lyle> like the login test
12:32:24 <david-lyle> which is the only global tempest job horizon has
12:32:52 <tsufiev> david-lyle, yeah, thanks for quick response! I've just been thinking aloud about how we can prevent such situations in devstack in future to not harm the projects that depends on it...
12:33:16 <david-lyle> tsufiev: have you found devstack to be broken?
12:34:52 <tsufiev> david-lyle, not exactly. Just the tests of apps relying on Horizon from devstack were broken
12:35:23 <david-lyle> ok, just trying to figure out the scope of the breakage
12:36:00 <david-lyle> the whole selenium test driven framework is proving very fragile
12:36:18 <tsufiev> because all dependencies in devstack are taken directly from global-reqs without Horizon check jobs such as here [the commit from proposal bot for Dajngo<1.8 should be here]
12:36:39 <tsufiev> david-lyle, yeah, totally true!
12:37:10 <tsufiev> my colleague is working on integration tests, recently they were broken again for some reason nobody understands why
12:37:12 <david-lyle> second severe break in 2 months
12:37:35 <david-lyle> tsufiev: yeah I spent some time looking into that one too
12:37:44 <david-lyle> no code changed on our part
12:37:55 <david-lyle> I have a bug for that one
12:38:40 <david-lyle> https://bugs.launchpad.net/horizon/+bug/1436903
12:38:41 <openstack> Launchpad bug 1436903 in OpenStack Dashboard (Horizon) "integration tests failing blocking gate" [Critical,Confirmed] - Assigned to David Lyle (david-lyle)
12:38:58 <tsufiev> hehe, seems that Horizon needs a person to dedicate himself to integration tests exclusively
12:39:12 <david-lyle> So we still don't have those voting in the gate
12:39:43 <tsufiev> Julie Pichon might have been such a person :(
12:42:16 <david-lyle> ok https://bugs.launchpad.net/horizon/+bug/1439169 is now filed
12:42:18 <openstack> Launchpad bug 1439169 in OpenStack Dashboard (Horizon) "selenium tests fail with Django 1.7" [Critical,Confirmed] - Assigned to David Lyle (david-lyle)
12:43:03 <david-lyle> Any other topics?
12:43:08 <rdopiera> I have two things
12:44:00 <rdopiera> One is about the xstatic packages again -- motivated by the yesterday's gate problems, I finally went and fixed xstatic packages publishing, so we now do that the same way as any other openstack projects
12:44:21 <rdopiera> I proposed a patch to the documentation explaining the new process here: https://review.openstack.org/#/c/169285/
12:44:38 <rdopiera> That should let us avoid such problems in the future.
12:45:09 <rdopiera> (the downside is that only cores can tag)
12:45:32 <david-lyle> well the problem we faced was the package was named incorrectly
12:45:47 <tsufiev> rdopiera, will vote with 2 arms for this :)
12:46:17 <rdopiera> david-lyle: that's not what caused the whole gate problems though -- changing a once released package was the mistake
12:46:36 <david-lyle> rdopiera: that is true
12:47:00 <rdopiera> hopefully the warning there will make people think twice at least :)
12:47:10 <david-lyle> but I think that was actually advice given by infra, would have to confirm that though
12:47:26 <rdopiera> david-lyle: they thought that it was already handled by other patches
12:47:40 <david-lyle> ok, probably poor communication
12:47:50 <david-lyle> the cause of so many issues
12:48:06 <david-lyle> I do appreciate your update of the docs
12:48:15 <rdopiera> anyways, we should be better off with a better defined process that is the same as the rest of openstack
12:48:28 <david-lyle> I agree
12:49:08 <david-lyle> rdopiera: second item?
12:49:26 <rdopiera> the second thing is more personal -- I'm leaving Red Hat in a month, and I don't know where I will land
12:49:27 <david-lyle> and why is clippy on review.o.o now?
12:49:41 <tsufiev> david-lyle, 1st april joke :)?
12:49:47 <rdopiera> I just thought it would be prudent to tell you guys, so that there is no surprise
12:50:14 <david-lyle> rdopiera: ok, thanks for the heads up
12:50:28 <tsufiev> rdopiera, are you going to continue with OpenStack/Horizon outside of RH?
12:50:50 <rdopiera> tsufiev: I'd love to, but I can't promise that
12:54:53 <david-lyle> seems like we are out of topics and almost time.
12:55:13 <david-lyle> So I'll close.
12:55:20 <david-lyle> Have a great week!
12:55:23 <tsufiev> bye!
12:55:24 <david-lyle> #endmeeting