20:00:19 <david-lyle> #startmeeting Horizon
20:00:20 <openstack> Meeting started Wed Oct  7 20:00:19 2015 UTC and is due to finish in 60 minutes.  The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:00:22 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:00:24 <openstack> The meeting name has been set to 'horizon'
20:00:29 <david-lyle> show yourselves
20:00:43 <david-lyle> :)
20:00:47 <neelashah> hi
20:00:51 <_ducttape_> o/
20:00:52 <barrett1> howdy
20:00:58 <tsufiev> hello
20:01:02 <tqtran> [=_=]/
20:01:07 <TravT_> o/
20:01:09 <jwy> hi
20:02:12 * david-lyle shakes bushes
20:02:34 <david-lyle> ok let's get rolling
20:02:54 <david-lyle> the agenda fro today
20:02:57 <mrunge> o/
20:02:57 <david-lyle> #link https://wiki.openstack.org/wiki/Meetings/Horizon#Agenda_for_October_7_2000_UTC
20:03:14 <david-lyle> #topic Liberty Status
20:03:32 <david-lyle> RC-2 has been opened and translations merged
20:03:40 <david-lyle> we also had two other patches land
20:03:47 <neillc> o/
20:03:53 <david-lyle> #link https://launchpad.net/horizon/+milestone/liberty-rc2
20:04:22 <david-lyle> mrunge: reported a third bug, but as there is no fix for master yet, and PKI is being phased out by keystone
20:04:33 <david-lyle> I don't think we should hold the release on it
20:04:42 <_ducttape_> can more docs be added to liberty?  thinking the changes in themeing, had 1:1 talk w hurgleburgler earlier
20:04:48 <david-lyle> #link https://bugs.launchpad.net/horizon/+bug/1503575
20:04:48 <openstack> Launchpad bug 1503575 in OpenStack Dashboard (Horizon) "switching projects via project switcher does not work" [High,New]
20:05:09 <david-lyle> _ducttape_: docs are not really possible this late
20:05:14 <david-lyle> because of translation
20:05:20 <_ducttape_> ugg, that stinks
20:05:22 <david-lyle> and string freezes
20:05:29 <mrunge> would be good, if someone could try to reproduce that bug at all.
20:05:37 <mrunge> I mean https://bugs.launchpad.net/horizon/+bug/1503575
20:05:48 <_ducttape_> we use fernets, and we do not see it
20:06:03 <mrunge> great
20:06:07 * _ducttape_ wonders if fernets should be plural ?
20:06:18 <david-lyle> sure internets is
20:06:33 <david-lyle> are
20:06:36 <david-lyle> hmmm
20:06:40 <mrunge> ferrets?
20:06:52 <david-lyle> token delivery by ferret
20:07:09 <_ducttape_> which is faster than keystone (truestory)
20:07:21 <david-lyle> deleting gets messy though
20:07:39 <lhcheng> mrunge: isn't that related to cookie overflow?
20:07:56 <_ducttape_> thats what I was thinking lhcheng, thats the same behavior
20:07:58 <mrunge> lhcheng, I had even database backed sessions
20:08:03 <david-lyle> ok, so I moved the above bug out of RC-2 and tagged as liberty-backport-potential
20:08:10 <mrunge> that was my first thought
20:08:54 <david-lyle> bpokorny: don't you use PKI?
20:09:06 <bpokorny> Yep
20:09:19 <david-lyle> have you seen the above issue?
20:09:37 <david-lyle> or would you have a chance to try and reproduce?
20:09:38 * bpokorny was just reading the above
20:09:43 <lhcheng> mrunge: try setting OPENSTACK_TOKEN_HASH_ENABLED=False
20:10:03 <bpokorny> I haven't seen that yet, but also haven't tried the latest master code in a while.
20:10:23 <lhcheng> if the keystonemiddleware don't have caching, that might be another cause..
20:10:28 <bpokorny> I'll try that out with master and see if it works ok.
20:10:38 <david-lyle> thanks bpokorny
20:10:41 <bpokorny> np
20:11:09 <david-lyle> RC-2 will be tagged tomorrow barring any show stoppers tonight
20:11:14 <mrunge> yes, thank you bpokorny
20:11:27 <david-lyle> and then only something catastrophic would cause an RC-3
20:11:36 <david-lyle> but final release is next week
20:11:56 <david-lyle> so Liberty is close to final for us
20:12:06 <david-lyle> questions/concerns about Liberty?
20:12:40 <david-lyle> #topic https://wiki.openstack.org/wiki/Horizon/WeeklyBugReport
20:12:59 <david-lyle> I think robcresswell is out and may not have this updated
20:13:05 * david-lyle looking
20:13:22 <tsufiev> yes, that's from the previous meeting
20:14:06 <david-lyle> 4/6 bugs seem to be fixed
20:14:27 <david-lyle> blueprints, not much progress
20:14:45 <hurgleburgler> robcresswell says sorry cause he's sick, he'll update the list tomorrow
20:14:45 <david-lyle> moving on
20:15:12 <david-lyle> #topic Summit Session Planning
20:15:32 <david-lyle> it's that time again to filter and finalize summit sessions
20:15:45 <matt-borland> :)
20:15:47 <david-lyle> the topic proposals are here:
20:15:48 <tsufiev> performance! performance!
20:15:51 <david-lyle> #link https://etherpad.openstack.org/p/horizon-mitaka-summit
20:16:12 <david-lyle> and the slots are
20:16:43 <david-lyle> #link https://mitakadesignsummit.sched.org/overview/type/horizon?iframe=yes&w=i:100;&sidebar=yes&bg=no
20:16:43 <david-lyle> iframe FTW
20:17:24 <david-lyle> I'd like to walk through and make sure we're not missing topics
20:17:45 <lhcheng> tsufiev: you'll be interested to attend this: https://etherpad.openstack.org/p/keystone-mitaka-summit-x-project
20:18:13 <tsufiev> lhcheng, thanks, will add it to my list
20:18:18 <david-lyle> that's grown a bit
20:19:45 <david-lyle> x-project?
20:20:05 <tsufiev> project X :)
20:20:21 <mrunge> sad, no cool name
20:20:26 <david-lyle> or a keystone session?
20:20:26 <r1chardj0n3s> o/
20:20:43 <lhcheng> cross project :)
20:21:13 <david-lyle> lhcheng: it's a cross-project session proposal?
20:21:14 <lhcheng> hmm time for a new project
20:21:20 * david-lyle still confused
20:21:41 <lhcheng> keystone cross project
20:21:48 <mrunge> there's a proposal with zaqar folks for a cross-project-session as well
20:22:08 <lhcheng> for tsufiev's requests of pagination for LDAP backend
20:22:22 <tsufiev> lhcheng, it's not just me I suppose
20:22:43 * tsufiev is flattered for being the single source of the whole cross-project session
20:23:25 <david-lyle> I don't see either of those on http://odsreg.openstack.org/
20:23:26 <lhcheng> tsufiev:  you're the first person I thought :P
20:23:37 <lhcheng> tsufiev: I'm sure others want it too
20:23:45 <mrunge> david-lyle, that is for sessions with > 2 projects
20:24:02 <mrunge> for zaqar, a session on wed. @4.40 would work
20:24:08 <mrunge> that'd be our slot
20:24:25 <mrunge> or wed @1.30 for zaqar, iirc.
20:24:36 <tsufiev> lhcheng, so far we decided that filtering + users limit on keystone side + horizon notification in case exceeding users were truncated would solve the issue w/o introducing pagination API
20:24:36 <david-lyle> cross-project overloaded
20:25:01 <TravT> mrunge: david-lyle: i think cross project is for things that all projects are interested in.
20:25:20 <tsufiev> hope another little keystone implementation detail won't ruin this brilliant plan
20:25:33 <david-lyle> TravT:  yes, but it's being used in another context here which was confusing me
20:25:51 <stevemar_> david-lyle: each project has X amount of fishbowl sessions
20:25:52 <TravT> yeah, i'm slightly confused about what mrunge is saying
20:26:06 <stevemar_> david-lyle: and cross-project sessions themselves are all fishbowl
20:26:12 <lhcheng> sorry to distract, let's move on to horizon sessions :)
20:26:12 <mrunge> should I try to rephrase?
20:26:32 <stevemar_> i decided to use one of keystone as another cross-project session, but one that focuses on keystone
20:26:50 <mrunge> I started a proposal on the bottom of https://etherpad.openstack.org/p/horizon-mitaka-summit
20:26:52 <david-lyle> stevemar_: I understand how the summit works, trying to decipher what level the keystone topic was
20:26:58 <mrunge> we just need to fill in topics
20:27:18 <david-lyle> stevemar_: ok
20:27:18 <david-lyle> makes sense
20:27:38 <david-lyle> will come and be argumentative^H^H^H^Hcollaborative
20:27:52 <david-lyle> :)
20:28:48 <hurgleburgler> ah, the nature of openstack :-D
20:30:01 <david-lyle> mrunge: we don't have a wed 1:30
20:30:15 <mrunge> david-lyle, yes, but zaqar has
20:30:16 <stevemar_> david-lyle: i really just wanted to kill the argument of pagination in keystone again
20:30:29 <TravT> stevemar_: can we have pagination?
20:30:33 <TravT> :p
20:30:36 <_ducttape_> so then keystone will support it?
20:30:39 <_ducttape_> ;)
20:30:41 <mrunge> oh yes please stevemar_
20:30:42 <tsufiev> lol )
20:30:45 <TravT> great!
20:31:04 <_ducttape_> stevemar_ said this would happen, he promised us all
20:31:15 <mrunge> or we'd come up with a ugly solution and blame keystone
20:31:17 <r1chardj0n3s> this is going well, I think ;)
20:31:20 <mrunge> ;-)
20:31:45 <david-lyle> do we have an idea on the fishbowls?
20:31:58 <david-lyle> zaqar, ceilometer, etc might be a good fishbowl
20:32:19 <mrunge> go with it?
20:32:20 <doug-fish> should we solicit input from operators at a fishbowl?
20:32:26 <mrunge> or how many do we have?
20:32:33 <TravT> doug-fish: ++ give them an opportunity to burn effigies
20:32:40 <mrunge> operators would be great for a fishbowl
20:32:40 <david-lyle> 2 fishbowls
20:32:53 <david-lyle> I worry operators are oversubscribed
20:33:04 <david-lyle> but if they will attend, I like having the session
20:33:38 <mrunge> so, e.g thu 9:50 operators?
20:33:46 <mrunge> just for planing purposes?
20:34:26 <doug-fish> I don't know how to predict/encourage them to come
20:34:32 <david-lyle> that may work, I don't see an op track for then
20:34:45 <david-lyle> looks like tues and wed for operators
20:37:56 <david-lyle> best meeting ever, everyone typing in etherpad
20:38:23 <david-lyle> reality, angular and the future is a topic I want early on
20:38:26 <neillc> +1
20:38:33 <r1chardj0n3s> sorry david-lyle :
20:38:38 <doug-fish> should Wed @ 5:30 include a discussion of all client side impl as well?
20:39:03 <david-lyle> r1chardj0n3s: I meant in a good way
20:39:22 <david-lyle> otherwise I randomly pick topics
20:40:14 <r1chardj0n3s> cross-domain tokens and federation stuff - does our horizon/keystone stuff need a whole session?
20:40:16 <david-lyle> doug-fish: not sure I understand the scope of all
20:40:50 <david-lyle> I think we could cover it on Fri if we're short on slots
20:41:07 <david-lyle> K2K is the big hole right now
20:41:09 <doug-fish> The end game for "all" might go as far as serving up the single static html and javascript and running completely client side
20:41:21 <doug-fish> hmm, that's a good one to talk about
20:41:51 <tqtran> single page application might be the term you're fishing for
20:41:54 <david-lyle> topic for operator session feedback is angular too
20:42:04 <david-lyle> *a topic
20:42:17 <tqtran> doug-fish: no pun intended on "fishing" lol
20:42:21 <doug-fish> :-)
20:42:22 <stevemar_> oh damn, i didn't realize i was barging in on the meeting, my bad.
20:42:24 <r1chardj0n3s> I guess it might be good for us to revisit the question posed in Paris and see if the answer we came up with there is still the one we agree on?
20:42:35 <david-lyle> stevemar_: always welcome
20:42:36 <stevemar_> no pagination for anyone!
20:42:39 <doug-fish> stevemar_: now you're doing it again  :-)
20:42:45 <mrunge> r1chardj0n3s, what was the question?
20:42:53 <r1chardj0n3s> do we move to pure JS
20:42:56 <mrunge> we all know, 42 is the answer...
20:43:02 <r1chardj0n3s> :)
20:43:16 <stevemar_> j/k we can chat about it at the summit :) i have a plan doug-fish
20:43:25 <doug-fish> is K2K important for others?
20:43:32 <r1chardj0n3s> stevemar_ has a solution, everyone! woo!
20:43:51 <david-lyle> doug-fish: I think others in the community, maybe not well represented in the current meeting
20:43:53 <TravT> stevemar_ will solve pagination!
20:43:59 <stevemar_> \o/
20:43:59 <r1chardj0n3s> \o/
20:44:02 <david-lyle> and world hunger
20:44:10 <stevemar_> no promises on that one david-lyle
20:44:22 <mrunge> boo
20:44:31 <david-lyle> by feeding pages to ...
20:45:19 <david-lyle> ok, I think the grand question should be discussed again
20:45:29 <david-lyle> single page or not to single page
20:45:47 <tqtran> i think we should devote one of the working session to covering plugins, how we are going to handle testing, translation, and the future angular work in it
20:45:47 <r1chardj0n3s> looks like that later Wed session is the go there
20:46:01 <mrunge> yupp
20:46:26 <r1chardj0n3s> I think someone who is on the angular front should try to write a plugin :)
20:46:45 <tqtran> considering that our path encourage people to develop their own plugins, we need to provide them the path of least resistance
20:46:53 <tqtran> r1chardj0n3s: already did :)
20:46:59 <r1chardj0n3s> oh?
20:47:14 <tqtran> and it works beautiful i must say, i have a demo line up at the summit, be on the look out!
20:47:20 <r1chardj0n3s> ok
20:47:26 <mrunge> great!
20:47:31 <tqtran> it also incorporated the workflow extenstion model we talked about at the midcycle
20:47:52 * tsufiev again says one word: Performance
20:48:04 <mrunge> wait, tsufiev :D
20:49:19 * tsufiev looking anxiously as the slots are being filled with shiny new things that work perfectly on devstack installations :/
20:49:21 <r1chardj0n3s> no, don't wait, we've only got 10 minutes of meeting left :)
20:49:40 <david-lyle> tsufiev: these are just placeholders
20:49:45 <david-lyle> nothing finalized
20:49:55 <david-lyle> the point is to make people speak up
20:50:17 <_ducttape_> tqtran - ping me with your plugin session, would love to attend
20:51:02 <tsufiev> david-lyle, ok, just consider this as a 'scale is coming' omen :)
20:51:36 <r1chardj0n3s> tsufiev: scale has already arrived in some places and Horizon doesn't cope - I'd be interested :)
20:51:40 <david-lyle> mrunge: ceilo has a conflict at 11:50, but maybe we could convince someone to attend
20:51:57 <mrunge> david-lyle, damn.
20:52:01 <david-lyle> tsufiev: that's actually why I wanted to go down the JS road to begin with
20:52:12 <david-lyle> scale and performance
20:52:13 <mrunge> david-lyle, I'm not sure, if zaqar would be the better option anyways...
20:52:20 <david-lyle> neither are being realized yet
20:52:52 <TravT> re: zaqar... i have had a few emails about how zaqar / searchlight might work together with horizon. was hoping to handle that discussion at the summit.
20:52:55 <r1chardj0n3s> the scaling issues I run into are actually terrible performance of the APIs we're calling, so it's not going to be fixed with JS :)
20:53:16 <_ducttape_> +1 r1chardj0n3s
20:53:26 <mrunge> r1chardj0n3s, TravT will fix that for us with searchlight
20:53:34 <r1chardj0n3s> mrunge: lolk
20:53:47 <lhcheng> r1chardj0n3s: +1, the landing pages take looooong time to load
20:53:48 <doug-fish> r1chardj0n3s: still it might be more palatable if multiple calls are being made in parallel, and you have a screen to look at
20:53:49 <david-lyle> r1chardj0n3s: true, but gracefully requesting more data could be and representing a usable page without all the data is too
20:53:59 <tsufiev> r1chardj0n3s, partly this can be fixed by improving APIs
20:54:15 <_ducttape_> david-lyle - that sounds like a DOS attack to me
20:54:19 <tsufiev> the problem is that they seem weren't designed with a Horizon in mind
20:54:24 <mrunge> ... but what if apis don't support pagination?
20:54:29 <r1chardj0n3s> just saying: the solution isn't just a simple "move to JS" :)
20:54:37 <mrunge> +1 r1chardj0n3s
20:54:39 <david-lyle> r1chardj0n3s: I fully agree
20:54:40 <_ducttape_> +2 r1chardj0n3s
20:54:44 <hurgleburgler> +1
20:55:04 <tsufiev> that is going to be a long tiresome war...
20:55:06 <david-lyle> and at this point we have work on things that don't benefit at all from it happening
20:55:11 <mrunge> so, what else did we had for sessions?
20:55:18 <david-lyle> which goes back to the existential crisis
20:55:31 <david-lyle> yes, save for Tokyo
20:55:41 <_ducttape_> I think tqtran's demo of ng plugin would be very helpful, my vote
20:55:43 <TravT> on all of the above, i definitely would like to have discussion on searchlight.
20:56:27 <mrunge> should we make plugins a session?
20:56:46 <_ducttape_> sounds reasonable, there are a lot of new plugins in the last yr
20:56:51 <r1chardj0n3s> TravT: searchlight and zaqar in the same session? fishbowl or work?
20:56:52 <david-lyle> I did
20:56:56 <david-lyle> mrunge: ^^
20:57:03 <doug-fish> Maybe a session on Domain/Federation/K2K and maybe HMT - to sequence the work if nothing else
20:57:03 <david-lyle> 4:40 Wed
20:57:11 <mrunge> right david-lyle
20:57:25 <_ducttape_> can tqtran have a slice of that time then david-lyle ?
20:57:31 <TravT> r1chardj0n3s: i don't care when exactly it happens. might even be best in contributors meetup.
20:57:40 <r1chardj0n3s> oh, "performance and scale" is already a session, searchlight can go there
20:57:49 <TravT> ok
20:57:49 <david-lyle> _ducttape_: nobody's claimed any time yet
20:57:51 <david-lyle> just topic
20:57:53 <david-lyle> s
20:58:06 <_ducttape_> 15 minutes of fame:  how to write an ng plugin
20:58:14 * _ducttape_ should be in marketing
20:58:15 <mrunge> doug-fish, yes, that was my next thought: we had many keystone topics
20:58:20 <david-lyle> do we want an all things keystone session then doug-fish?
20:58:30 <doug-fish> sounds good to me
20:58:31 <r1chardj0n3s> oh, when can we schedule the replacement of our testing environment with pytest please can we pleeeease? :)
20:58:47 <doug-fish> I'd like to include stevemar_'s work on pagination from that session
20:58:54 <doug-fish> s/include/exclude
20:59:15 <TravT> david-lyle: FYI, i know i can't make the wed 4:40 and i was asked to go to a monasca / ceilometer session at 5:30 Wed.
20:59:17 <mrunge> r1chardj0n3s, just added, feel free to correct
20:59:32 <tsufiev> r1chardj0n3s, what is the gain of using pytest?
20:59:55 <r1chardj0n3s> tsufiev: I should write a BP :)
21:00:00 <tsufiev> ok )
21:00:21 <david-lyle> r1chardj0n3s: you're signing up to rewrite all 1600 tests?
21:00:31 <mrunge> awesome r1chardj0n3s
21:00:32 <TravT> cheers to r1chardj0n3s!
21:00:33 <david-lyle> O_o
21:00:35 <neillc> uh oh
21:00:40 <r1chardj0n3s> david-lyle: it's the only possible way to understand them ;)
21:00:43 <TravT> his minion seems worried, though
21:00:50 <hurgleburgler> ٩(͡๏̮͡๏)۶
21:00:50 <david-lyle> haha
21:01:04 <mrunge> so, all slots are now filled. did we forget something?
21:01:10 <TravT> beer
21:01:16 <r1chardj0n3s> that's 3:30 Thur
21:01:21 <mrunge> and more beer, TravT
21:01:30 <tsufiev> TravT, take beer to the sessions ;)?
21:01:31 <hurgleburgler> Asahi!
21:01:34 <tqtran> _ducttape_: https://www.openstack.org/summit/tokyo-2015/schedule/main-conference heres the session, sorry someone was talking to me in the office
21:01:36 <TravT> will there be a pre-summit meetup?
21:01:37 <david-lyle> time's up. I think I have to finalize the sessions in the next week
21:01:40 <doug-fish> would we be interested in an informal meetup before the summit starts?
21:01:42 <tqtran> got a bit distracted
21:01:43 <TravT> monday night or something?
21:01:53 <mrunge> doug-fish, +1 /me!
21:02:02 <mrunge> sure!
21:02:08 <tsufiev> +1
21:02:11 <doug-fish> let's take it to the horizon channel
21:02:16 <mrunge> :D
21:02:38 <david-lyle> there's actually not another meeting schduled in here
21:02:50 <david-lyle> but let's go to #horiozn
21:02:58 <david-lyle> #horizon
21:02:58 <TravT> so, doug-fish will schedule a meetup
21:03:05 <david-lyle> #endmeeting