20:00:48 <david-lyle> #startmeeting HorizonDrivers
20:00:49 <openstack> Meeting started Wed Jun 15 20:00:48 2016 UTC and is due to finish in 60 minutes.  The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:00:50 <hurgleburgler> ( ಠ‿ಠ)┘
20:00:50 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:00:52 <openstack> The meeting name has been set to 'horizondrivers'
20:00:55 <tyr__> o/
20:00:58 <TravT> o/
20:00:58 <tsufiev> o/
20:00:59 <bpokorny> o/
20:01:23 <david-lyle> Rob is off holidaying, so I'm here for your entertainment
20:01:33 <r1chardj0n3s> o/
20:01:51 <david-lyle> The agenda is not set, and the morning Horizon meeting was cancelled
20:02:05 <david-lyle> Couple of reminders
20:02:15 <david-lyle> The midcycle is coming up quickly
20:02:57 <david-lyle> https://wiki.openstack.org/wiki/Sprints/HorizonNewtonSprint
20:03:02 <david-lyle> #link https://wiki.openstack.org/wiki/Sprints/HorizonNewtonSprint
20:03:17 <david-lyle> please register if you plan to attend
20:03:31 <david-lyle> newer users of the wiki are not able to get accounts to edit
20:03:37 <TravT> i'm surprised clu isn't on there.
20:04:00 <david-lyle> so if you want to register but can't, please read the instructions on the page for alternative methods
20:04:24 <david-lyle> the etherpad for topics is linked from there as well
20:04:39 <david-lyle> TravT, that's clu_ to you
20:05:57 <david-lyle> Not sure of other general items other than npm_run_test continues to plague us
20:06:49 <david-lyle> There have not been many new blueprints proposed, does anyone have one they want/need reviewed?
20:07:17 <TravT> i do have a question
20:07:30 * david-lyle will fake an answer
20:07:30 <TravT> where are we on library upgrades and all the xstatic fun?
20:07:40 <david-lyle> well
20:07:42 <r1chardj0n3s> ah
20:07:58 <david-lyle> our gate is now all separate from integrated tempest
20:08:27 <david-lyle> and we're relying on upper-constraints, so are we ready to roll r1chardj0n3s ?
20:09:00 <r1chardj0n3s> I need to review where we're up to in the "plan" and see what the next step is
20:09:17 <david-lyle> I think we're at the break things part
20:09:30 <david-lyle> wait, that's where we're always at
20:09:34 <r1chardj0n3s> I was about to give a longer answer but I think david-lyle summarises well ;-)
20:09:40 <david-lyle> well, it's easy to find
20:09:53 <r1chardj0n3s> yeah, we should be able to attempt an automated release
20:09:57 <hurgleburgler> how exciting ]o/
20:10:00 <hurgleburgler> \o/ too
20:10:13 <r1chardj0n3s> I will work with my infra / releases folks to try to make that happen
20:10:14 <TravT> okay, just was thinking that we'd want to have library upgrades all done before newton 3
20:10:29 <david-lyle> yes, sooner is better
20:10:49 <r1chardj0n3s> that's now #1 on my horizon priorities
20:10:56 <david-lyle> the tempest part was ready a couple of weeks ago, so I'm not sure what we're stuck on
20:11:10 <r1chardj0n3s> I think we're stuck on me
20:11:13 <david-lyle> maybe rob's vacationing brain knows
20:11:39 <r1chardj0n3s> I got distracted after the tempest change went through, sorry.
20:11:45 <david-lyle> no worries
20:12:09 <david-lyle> so library soup is on the menu for the next few weeks
20:12:12 <david-lyle> delightful
20:12:17 <r1chardj0n3s> \o/
20:12:35 <tsufiev> do we plan to roll angular 1.4 in newton?
20:12:40 <tsufiev> just curious
20:12:49 <TravT> tsufiev, yes
20:12:53 <david-lyle> I'm going to be that guy
20:12:59 <tsufiev> nice :)
20:12:59 <TravT> robcresswell already did some work on that
20:13:11 <david-lyle> we need to make sure that the supported distros are going to line up
20:13:26 <r1chardj0n3s> yeah
20:13:38 <david-lyle> or we deliverable unrunnable code
20:13:43 <TravT> https://review.openstack.org/#/c/308251/
20:13:53 <david-lyle> 16.04 should have shiny new versions
20:14:17 <TravT> and david-lyle's scripts nixed his review fixing a few things
20:14:18 <TravT> https://review.openstack.org/#/c/309004/
20:14:25 <david-lyle> but we should pull mrunge in at some point to consult
20:14:43 <mrunge> sorry, in a meeting right now
20:15:04 <david-lyle> mrunge, we'll need you another day, just reminding people to talk to you about library versions
20:15:11 <david-lyle> mrunge, go about your business
20:15:13 <david-lyle> :)
20:15:17 <mrunge> :D
20:15:18 <mrunge> ack
20:15:24 <hurgleburgler> lol
20:16:13 <david-lyle> TravT, I aim to nix things
20:16:34 <david-lyle> I started going through the abandons of blocked, but figure people would undo those
20:16:41 <r1chardj0n3s> so what's the version of red hat we should be checking for our library versions
20:16:58 <david-lyle> r1chardj0n3s, I honestly don't know any more
20:17:19 <david-lyle> we should just consult with matthias another time
20:17:22 <r1chardj0n3s> I guess I'll just try to figure out what the latest release is ;-)
20:17:26 <r1chardj0n3s> yeah maybe
20:19:13 <david-lyle> any other topics?
20:21:26 <david-lyle> I can always dig into the stale bp list if people are wanting
20:21:27 * tsufiev again reminds everybody to look into dsvm-integration job failure videos
20:21:40 <TravT> link tufiev?
20:21:42 <david-lyle> now there's a fun date night
20:21:53 <TravT> lol
20:21:54 <tsufiev> it's fun and very educating
20:22:07 <david-lyle> TravT, they are part of the review logs
20:22:19 <david-lyle> so depends on the failed review
20:22:23 <tsufiev> TravT, inside test_report dir for every failed dsvm job
20:22:32 <david-lyle> s/review/job failure/
20:22:51 <TravT> okay, cool. thanks for that. i saw some chatter on that, but didn't digest it
20:23:12 <tsufiev> so far I've managed to see that most of failures are caused by dropdowns somehow not showing up when clicked
20:23:31 <hurgleburgler> uh oh
20:23:47 <david-lyle> because they're themed?!?!?!
20:23:50 <tsufiev> :))
20:24:09 <hurgleburgler> ಠ▃ಠ
20:24:17 <hurgleburgler> *runs away* ε=┌( >_<)┘
20:24:33 <tsufiev> my primary 'suspect' is that some delayed JS initialization is done on page at the same moment when test clicks the dropdown
20:24:46 <tsufiev> and it interferes with showing the dropdown content
20:24:49 <david-lyle> sleep(1)
20:24:54 <david-lyle> fixes everything
20:25:04 <TravT> i usually prefer nap(15)
20:25:08 <tsufiev> david-lyle, that's the EVIL way
20:25:17 <tsufiev> we're Jedi, not the Sith
20:25:17 <david-lyle> vacay(15)
20:26:02 <tsufiev> although there only ~2 people with integration tests knowledge, which aligns with Sith lore :/
20:26:32 <TravT> tsufiev, don't resist the power of the dark side
20:27:00 * tsufiev hums Imperial March
20:28:22 <david-lyle> other fun and excitement? or should we let tsufiev sleep and r1chardj0n3s wake up?
20:28:31 <tsufiev> any ideas why npm-run-test still bothering us, by the way?
20:28:34 <r1chardj0n3s> don't wanna
20:28:50 <r1chardj0n3s> I thought we had an npm mirror now
20:29:09 <r1chardj0n3s> but we're still seeing bizarro network issues in install part of tests
20:29:24 <tsufiev> and 40 mins timeouts as well
20:29:39 <r1chardj0n3s> yeah, like the test runner just decides to not exit
20:29:47 <david-lyle> krotscheck, around?
20:30:05 <krotscheck> Only if you're willing to fill out a Turing-test survey afterwards :)
20:30:06 <david-lyle> npm-run-test question
20:30:27 * krotscheck reads scrollback
20:30:32 <david-lyle> why does it not work? explained to melike I'm a horizon developer
20:30:58 <krotscheck> Righto.
20:31:03 <krotscheck> "I don't know"
20:31:21 <david-lyle> I already had that answer on my own :)
20:31:23 <krotscheck> We've observed it hanging at the end of both test runs, and the only thing that happens there is the generation of the coverage report.
20:31:31 <krotscheck> s/both/either/
20:31:49 <tsufiev> should we disable coverage reports temporarily?
20:32:01 <krotscheck> Of course, there's a lot of things happening during that, including (I believe) releasing the karma ports.
20:32:13 <krotscheck> Releasing the browsers, etc.
20:32:35 <krotscheck> Our original hypothesis whas that the system's # of open files is a problem.
20:32:47 <krotscheck> But we landed that patch, and it didn't do anything.
20:33:28 <krotscheck> (The reason we came up with that hypothesis is because it was stuck on popen_wait
20:34:35 <david-lyle> what can we do to help that won't get in your way?
20:35:21 <krotscheck> david-lyle: Well, I let matt-borland know monday that I can't really continue investigating it, as I've got 2 weeks to go before a 2 week vacation, and all this App Eco WG and JavaScript SDK work is sucking down my time.
20:35:43 <krotscheck> david-lyle: So, having someone take over would be nice.
20:35:57 <david-lyle> ok, so we won't step on your toes flailing at it for a bit
20:36:06 <krotscheck> This post will go a long way towards getting a test environment set up: https://krotscheck.net/2016/06/01/how-to-simulate-an-openstack-infra-slave.html
20:36:31 <krotscheck> Which is what I used to simulate the issue, but there's the unfortunate bug that vagrant doesn't play well with xenial yet.
20:36:40 <krotscheck> No you wont.
20:37:19 <david-lyle> ok, thanks for your efforts and the link
20:39:36 <david-lyle> anyone else, or should we wrap up?
20:40:36 <david-lyle> sounds like we're done. Thanks everyone.
20:40:39 <TravT> thanks
20:40:41 <david-lyle> #endmeeting