20:05:39 <david-lyle> #startmeeting Horizon
20:05:39 <openstack> Meeting started Wed Jul  1 20:05:39 2015 UTC and is due to finish in 60 minutes.  The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:05:40 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:05:40 <ducttape_> is there a horizon thing today, or just talk of drinks?
20:05:43 <openstack> The meeting name has been set to 'horizon'
20:05:46 <david-lyle> got distracted
20:06:00 <TravT> its okay, we determined several important things.
20:06:00 <david-lyle> hello everyone
20:06:03 <TravT> tqtran is now a bot
20:06:09 <TravT> and we need a bar at the summit
20:06:10 <mwhagedorn> hello
20:06:20 <mrunge> good summary TravT
20:07:46 <david-lyle> #topic liberty priorities
20:07:58 <david-lyle> #link https://etherpad.openstack.org/p/YVR-horizon-liberty-priorities
20:09:33 <david-lyle> reOrg: I think there are very few bits left that haven't moved at least a few millimeters
20:10:04 <david-lyle> app structure also seems well on its way
20:10:19 <krotscheck> o/
20:10:55 <david-lyle> babel work still inflight
20:11:10 <david-lyle> patches are up, this is a high priority review
20:11:15 <tqtran2> its almost there, i think another day
20:11:21 <david-lyle> to unblock more work
20:11:24 <david-lyle> tqtran2: ok
20:11:57 <TravT> all but one of the launch instance NG testing patches have landed.
20:12:07 <david-lyle> cool
20:12:16 <david-lyle> great progress
20:12:41 <david-lyle> for the plugins work, I have a patch up to move sahara content to openstack_dashboard/contrib
20:13:06 <david-lyle> making that move helped determine some holes in testing pluggable content
20:13:19 <tqtran2> hehe yes
20:13:33 <tqtran2> the tests for plugins is quite bare
20:13:36 <david-lyle> so a couple fixes are included in that patch, a couple more were noted and will be addressed separately
20:14:06 <vgridnev> Horizon guys, hello. I am sahara contributor. I have small question about replacing sahara-horizon-plugin to contrib. Do we have some chance of merging several of current sahara changes before moving sahara-horizon-plugin to contrib? I would allow to avoid rebasing in all current sahara patches
20:14:56 <mrunge> make sense to me
20:15:06 <ducttape_> vgridnev - I had to do exactly that for designate, it wasn't that bad if you mess around with git patches
20:15:08 <david-lyle> mrunge: you didn't do the move
20:15:14 <david-lyle> :P
20:15:19 <david-lyle> either way
20:15:20 <mrunge> dansmith, right, I did not
20:15:32 <mrunge> sorry.  I meant david-lyle
20:15:55 <david-lyle> mostly path changes
20:16:10 <david-lyle> so either way will not be much work
20:16:19 <ducttape_> you can hand edit a git patch to correct the path changes, and then git apply your patch
20:16:53 <mrunge> it's more the question, if one patch needs to be changed, or several
20:17:12 <ducttape_> yep.  if it is going to move, the sooner the better I think
20:17:24 <mrunge> but: looking at speed of merging sahara patches, that'd delay contrib etc.
20:17:32 <david-lyle> if I wait for all sahara patches to land, I will get this patch merged in M I think
20:17:49 <mrunge> later ;-)
20:18:31 <TravT> we need to get a pattern established for contrib, so i wouldn't want to see it get to held up
20:18:46 <david-lyle> I don't mind rebasing sahara patches either
20:18:53 <mrunge> yupp
20:19:11 <david-lyle> any way, moving on session cleanup
20:19:51 <mrunge> linked patch is to last_activity session flag
20:19:53 <ducttape_> that is a minor patch on the etherpad, lhcheng was thinking of bigger changes
20:20:07 <david-lyle> there is one clean up patch, I posted a patch to devstack to require memcache, that was rebuked
20:20:13 <ducttape_> but that patch, is related
20:20:24 <david-lyle> mandatory memcache requires more work
20:20:31 <mrunge> prerequisite already landed
20:20:39 <ducttape_> memcached has really been horid in my dev setup with the login issue
20:21:18 <david-lyle> so I can either deliver all the platform patches for memcache or change to LocMemCache
20:21:28 <mrunge> yes, I have seen interesting effects with memcache, but not with a database session store
20:21:31 <david-lyle> neither is a great choice for production
20:22:01 <tsufiev> mrunge, interesting in good sense of this word?
20:22:04 <ducttape_> is there tunning for LocMemCache, and does it suffer the same django + reload = login again problem as memcached?
20:22:18 <mrunge> david-lyle, if you'd be delivering production setup, we would set up more security features
20:22:35 <david-lyle> ducttape_: not sure, I experienced the relogin with LocMemCache
20:22:49 <doug-fish> ducttape_: I wonder if you are doing something wrong in your dev setup. It memcache works well for me - no need to log in @ every server restart.
20:22:59 <mrunge> tsufiev, well, you'll see all kinds of issues, and you don't have a clue what's going on
20:23:10 <lhcheng> not sure if LocMemCache would work if django runs on apache, thought it was only for running development server (django command line)
20:23:24 <david-lyle> lhcheng: hmm, I could test
20:23:27 <david-lyle> not sure
20:23:29 <ducttape_> that would be even better
20:23:41 <ducttape_> if you are running in apache, then setup a real db too
20:23:57 <lhcheng> apache will run horizon in multiple threads, and LocMemcache would only cache per thread.
20:24:00 <david-lyle> other option is sqlite again
20:24:01 <mrunge> doug-fish, tsufiev when session cookie becomes too big, it becomes truncated. login issues, etc included then
20:24:16 <tsufiev> mrunge, hmm, weird... in mirantis openstack horizon works well with Memcached
20:24:21 <mrunge> I have seen that only with memcache
20:24:25 <doug-fish> right - but are you getting too big for memcached?
20:24:34 <doug-fish> that's a big cookie.
20:24:37 <david-lyle> mrunge: memcache should be able to handle larger session size
20:24:38 <lhcheng> I still think we should move the service catalog out of session data into the django cache :)
20:24:55 <david-lyle> lhcheng: but was still need a cache
20:24:56 <mrunge> let's move that to #openstack-horizon
20:25:01 <ducttape_> I thought memcached was like 4G or something with decent size.   I know the session stuff works fine with memcached
20:25:27 <lhcheng> service catalog doesn't change, it doesn't have to be read/write every page refresh
20:25:28 <tsufiev> mrunge, perhaps it was just some misconfiguration issue?
20:25:41 <david-lyle> lhcheng: but you have to store it somehow
20:25:48 <mrunge> tsufiev, maybe?
20:25:54 <mrunge> probably?
20:26:04 <david-lyle> I can revisit the devstack patch
20:26:15 <doug-fish> lhcheng: I would like to know more about the benefit - I feel like I don't understand the reason for moving it
20:26:19 <lhcheng> david-lyle: yeah, in the django cache. either db, memcache, or not at all (cache it per request)
20:26:49 <tsufiev> mrunge, sorry, no exact idea, just guessing
20:27:04 <mrunge> tsufiev, sure
20:27:11 <david-lyle> right fallback is per request API call to keystone
20:27:22 <david-lyle> let's move on
20:27:32 <david-lyle> and discuss in #horizon later
20:27:38 <lhcheng> sure
20:27:47 <david-lyle> jscs/eslint changes are well underway
20:28:03 <david-lyle> moving the remaining bits not previously moved
20:28:26 <david-lyle> but the tool has switched to eslint to provide a common tool for openstack
20:28:36 <david-lyle> and jshint and jscs are gone
20:29:00 <david-lyle> auto-collecting is underway as well
20:29:17 <david-lyle> patches up and ready for review I believe
20:29:28 <TravT> yes, several are landed
20:30:11 <TravT> the sooner we get through these lint changes and re-orgs that better.
20:30:19 <TravT> s/that/the
20:30:34 <david-lyle> same is generally true for all the patches :)
20:30:50 <david-lyle> every one has their sacred cow
20:30:52 <TravT> well, these all are causing a number of rebases and creating a moving target
20:31:00 <TravT> harder to build on.
20:31:38 * ducttape_ can hear dave saying same is generally true for all the patches
20:32:03 <TravT> IMO this is a bit worse.
20:32:08 <krotscheck> WIth that in mind, I'm quite capable of going through and hitting some of the low-hanging fruit on the linting, I just have zero insight on which files have been claimed yet.
20:32:11 <david-lyle> your sacred cow
20:32:14 <david-lyle> :P
20:32:21 <david-lyle> moving on
20:32:24 <ducttape_> thats perspective TravT ;)
20:33:08 <TravT> it is. it is hard to tell people what to build on for angular at the moment.
20:33:09 <david-lyle> agenda for today is found at https://wiki.openstack.org/wiki/Meetings/Horizon
20:33:21 <david-lyle> TravT: try plugins
20:33:24 <david-lyle> :D
20:33:32 <david-lyle> just saying there's lots of work
20:33:38 <david-lyle> most of it's important
20:33:48 <TravT> anyway, there's a reason all the things in top priority are listed in top priority
20:34:19 <TravT> so, i guess we're not really disagreeing on anything.
20:34:32 <ducttape_> because someone spent time on the etherpad?  ;)
20:34:50 <david-lyle> #topic https://blueprints.launchpad.net/horizon/+spec/context-sensitive-help (ddomingo)
20:35:01 <TravT> because that's what was discussed in the room at the summit as the blockers to moving ahead. :)
20:35:05 <david-lyle> ddomingo doesn't appeat to be on
20:35:08 <ducttape_> this would be awesome, to have the help thing.  tell me more
20:35:08 <mrunge> ah, it was me to add that
20:35:25 <mrunge> ddomingo couldn't attend this meeting, he just asked for comments
20:35:40 <mrunge> he said, he will show up in next meeting
20:35:47 <ducttape_> I think it would help, kinda surprised its not already in horizon
20:36:08 <mrunge> david-lyle, and this is a hint/request to do a blueprint review
20:36:26 <david-lyle> yeah we saw something similar in Atlanta and no drive to push it
20:36:35 <david-lyle> I have no issues with the concept
20:36:45 <david-lyle> but we should review the details
20:36:53 <david-lyle> and comment on the blueprint
20:37:00 <ducttape_> plugin help would be a consideration too, but other seems like nothing but acceptance
20:37:06 <mrunge> don said, he would be willing to implement it
20:37:40 <david-lyle> mrunge: excellent
20:37:42 <mrunge> but he surely could use some ideas from us
20:39:28 <ducttape_> I think it would be good to have some extensions on panels etc with an added url attribute for help stuff.  if it's there, display help link.   right?
20:39:58 <ducttape_> maybe some base level horizon class used by forms / table views etc
20:39:59 <mrunge> ducttape_, makes sense to me
20:41:05 <david-lyle> make sense
20:41:20 * ducttape_ thinks no one wants to talk about help content.  even fewer want to implement it.
20:41:43 <mrunge> there was help content added to new launch instance, too
20:41:46 <TravT> ^ducttape_ i disagree. we put in some really nice help in the new launchy instance
20:42:13 <tqtran2> yes, i was thinking how this would fit into what we  have for help panels already
20:42:20 <mrunge> TravT, ducttape_ one can discuss, where this should be stored
20:42:23 <ducttape_> cool, I didn't see anyone chiming in.  is there docs etc on how to add that for new work?
20:42:30 <tqtran2> and if it would make sense to help that effort rather than adding more to panels we are going to deprecate
20:43:03 <mrunge> I would love to reuse help content for generating docs
20:43:08 <david-lyle> tqtran2: a general mechanism to source would reusable
20:43:26 <david-lyle> and the existing code is long from dead
20:43:28 <ducttape_> yep, help content should not be terribly sensitive to how the page is rendered
20:43:43 <tqtran2> i have to read the bp more
20:44:05 <david-lyle> #topic MidCycle meetup logistics
20:44:10 <david-lyle> TravT:
20:44:30 <TravT> yeah, I sent out a message to ML that we will host at HP in Fort Collins, CO
20:44:38 <TravT> see: https://etherpad.openstack.org/p/horizon-liberty-midcycle
20:44:59 <TravT> one question I do need to get answered is if we are going to want on site lunch catering or will be going out.
20:45:39 <TravT> and we'll need final numbers for who all will be attending.
20:46:07 <doug-fish> How long does it take to go out? (I don't know Fort Collins)
20:46:24 <TravT> there are lots of restaurants just across the street.
20:46:31 <ducttape_> there is some limited nearby stuff, but it's fairly chain-oriented stuff
20:46:37 <TravT> there is an on site cafeteria that i do not recommend
20:46:43 <doug-fish> :-)
20:46:45 <ducttape_> lol
20:47:14 <doug-fish> my suggestion is that we go out to break up the day a bit
20:47:27 <TravT> it does make for a nice break to do that.
20:47:35 <ducttape_> I think I could get my company to pay for a $1.50 costco hotdog meal for everyone, about 1 mile away
20:47:42 <TravT> we can still have morning and afternoon snacks, I think.
20:48:31 <TravT> #startvote (cater or go out for lunch)? cater, out
20:48:32 <openstack> Only the meeting chair may start a vote.
20:48:41 <TravT> david-lyle, can you start vote?
20:48:47 <hurgleburgler> denied!
20:48:56 * TravT crawls back into cave
20:49:00 <hurgleburgler> ಠ◡ಠ
20:49:02 <TravT> :)
20:49:10 <david-lyle> #startvote cater or go out for lunch ? cater out
20:49:11 <openstack> Begin voting on: cater or go out for lunch ? Valid vote options are cater, out.
20:49:12 <openstack> Vote using '#vote OPTION'. Only your last vote counts.
20:49:15 <tqtran2> ducttape_: costco hotdog, yummy
20:49:22 * tqtran2 licking fingers.
20:49:29 <doug-fish> #vote out
20:49:53 <tqtran2> #vote out
20:50:11 <hurgleburgler> #vote out
20:50:12 <ducttape_> #vote out
20:50:15 * david-lyle should have added a don't care option
20:50:27 <tqtran2> clu's not here, tsk tsk
20:50:37 <esp> #vote out
20:50:40 <TravT> #vote out
20:50:49 <david-lyle> #vote out
20:50:56 * david-lyle but doesn't really care
20:51:01 <david-lyle> any more
20:51:02 <david-lyle> ?
20:51:10 <david-lyle> #endvote
20:51:11 <openstack> Voted on "cater or go out for lunch ?" Results are
20:51:12 <openstack> out (7): esp, ducttape_, doug-fish, TravT, david-lyle, hurgleburgler, tqtran2
20:51:24 <david-lyle> so catered lunch then
20:51:24 <TravT> ok. We'll do that, but still look to have some morning / afternoon snacks delivered.
20:51:28 <esp> +1 democracy
20:51:28 <david-lyle> :P
20:51:33 <hurgleburgler> and a bar
20:51:40 <david-lyle> oh yeah catered bar
20:52:06 <ducttape_> whats the hp policy on bringing your own keg ?
20:52:15 <doug-fish> one concern on this document:  https://docs.google.com/spreadsheets/d/1w0eI6SPCA2IrOyHiEYC2uDO3fbYGzahZRUQSva0UD3Y/edit#gid=1788841692 ....
20:52:16 <ducttape_> just asking, thanks
20:52:18 <david-lyle> TravT: do you want the attendance confirmations on the etherpad or google doc?
20:52:45 <doug-fish> I was planning to be there July 21-23 which is Tues - Thurs, but I see and agenda for Wed - Fri
20:52:51 <david-lyle> we just need to direct people somewhere
20:53:07 <TravT> yeah, let's just do the spreadsheet.
20:53:09 <david-lyle> 21-23
20:53:14 <TravT> because a number of people already filled that in.
20:53:24 <david-lyle> agenda dates are wrong
20:53:25 <tqtran2> right, 21-23 is what i was looking at as well
20:53:36 <david-lyle> s/dates/days/
20:53:54 <david-lyle> anything else midcycle related?
20:54:12 <TravT> i think that is the main thing for today.  just needed to get that info to the people with money
20:54:22 <tqtran2> lol
20:54:24 <david-lyle> #topic Open Discussion
20:54:29 <ducttape_> I have a blueprint I added this week, have the code pretty much done.  https://blueprints.launchpad.net/horizon/+spec/lbaas-vip-fip-associate
20:54:30 <david-lyle> thanks to the people with money
20:55:08 <david-lyle> ducttape_: started looking at, glad amotoki is reviewing as well
20:55:10 <tsufiev> would like to remind you guys about the curvature blueprint https://blueprints.launchpad.net/horizon/+spec/curvature-network-topology
20:55:18 <ducttape_> not sure how frequently bps get reviewed.... thought I'd spam everyone
20:55:28 <hurgleburgler> Also, I've made much progress on the css reorg work … but its gonna be dependent on landing: https://review.openstack.org/#/c/188162/ and https://review.openstack.org/#/c/195371/
20:55:53 <tsufiev> sorry for totally ignoring the priorities, but seems I found way to angularize topology graph to make it reusable
20:56:13 <ducttape_> tsufiev - that sounds awesome
20:56:51 <david-lyle> tsufiev: that sounds great
20:56:52 <ducttape_> and the priorities are not the same throughout the entire group, don't worry about that
20:56:52 <tsufiev> ducttape_, planning to upload first version of commit the end of this week or the beginning of next week
20:56:55 <david-lyle> will take another look
20:57:11 <tsufiev> david-lyle, do I need a separate blueprint for this activity?
20:57:12 <david-lyle> hurgleburgler: been meaning to look at your patches as well
20:57:21 <hurgleburgler> Thanks
20:57:55 <david-lyle> tsufiev: I don't think so unless curvature lands before your work for some reason
20:58:31 <hurgleburgler> One of them is having jenkins issues, I should have that sorted out today
20:58:38 <tsufiev> david-lyle, okay. Well, anyways I'm basing upon curvature commit, so it needs to be at least in mergeable state
20:58:41 <TravT> what is the status on curvature in general?
20:59:23 <tsufiev> TravT, it works well as is, although the code is a bit weird (most likely due to traslation from coffescript)
20:59:43 <TravT> this reminds me, didn't we need a d3 library upgrade?
20:59:45 <TravT> did that get done?
20:59:49 <ducttape_> what is heat using?  seems similar, like they could converge
21:00:32 <tsufiev> ducttape_, I think they could, once I roll out topology graph angular widget, someone may try use it for heat stacks
21:00:41 <david-lyle> TravT: d3 is all
21:00:50 <david-lyle> I mean ducttape_
21:01:26 <david-lyle> looks like we're out of time
21:01:49 <david-lyle> thanks everyone. please fill in your midcycle attendance + of -
21:01:54 <david-lyle> #endmeeting