20:00:36 <r1chardj0n3s> #startmeeting horizon 20:00:37 <openstack> Meeting started Wed Dec 14 20:00:36 2016 UTC and is due to finish in 60 minutes. The chair is r1chardj0n3s. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:00:38 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:00:41 <openstack> The meeting name has been set to 'horizon' 20:00:47 <r1chardj0n3s> ducttape_ o/ 20:01:11 <lcastell> o/ 20:01:24 <betherly> o/ 20:01:40 <rhagarty> o/ 20:02:04 <rdopiera> o/ 20:02:09 <r1chardj0n3s> I'll start off with the usual 20:02:10 <r1chardj0n3s> #topic Priority patches for review 20:02:26 <r1chardj0n3s> #link https://review.openstack.org/#/q/starredby:r1chardj0n3s%20AND%20status:open is the current set, please focus attention here if you can 20:03:18 <r1chardj0n3s> if you have anything you believe should be getting attention on the priority list please let me know 20:03:40 <r1chardj0n3s> We've got about four weeks until feature freeze for Ocata 20:04:13 <ducttape_> ediardo - did your css change for the overflow make it yet ? 20:04:26 <ediardo> o/ 20:04:29 <r1chardj0n3s> At this point, it looks like we're going to have a lot of trivial fix patches in. 20:04:33 <ducttape_> that *might* be something to consider too 20:05:02 <ediardo> NOpe.. 20:05:27 <ying_zuo> I have a blueprint accepted for ocata. https://blueprints.launchpad.net/horizon/+spec/launch-instance-volume-quotas 20:05:35 <ducttape_> k. might need to touch base with hurgleburgler again. that bug is not great. thanks ediardo 20:05:59 <ying_zuo> Right now there's only one patchset. Should that be one the priority list? 20:06:11 <r1chardj0n3s> ying_zuo: seems entirely reasonable to me! added :-) 20:06:34 <ying_zuo> Thanks Richard 20:06:51 <r1chardj0n3s> ducttape_: I don't have any of your quotas work in the list - should I? 20:06:58 <ediardo> this could be a "trivial fix": https://review.openstack.org/#/c/409753/ 20:07:15 <ducttape_> the one quota thing, you have in the list r1chardj0n3s. I'll talk w u after this meeting if avail 20:07:21 <rdopiera> r1chardj0n3s: I'd like some attention on that django config patch, if possible 20:07:23 <ducttape_> my other stuff is not ready 20:07:43 <r1chardj0n3s> ducttape_: ok 20:08:00 <r1chardj0n3s> rdopiera: is it ready for review? 20:08:05 <rdopiera> r1chardj0n3s: yes 20:08:31 <rdopiera> r1chardj0n3s: https://review.openstack.org/#/c/404735/ 20:08:43 <r1chardj0n3s> ok, cool! could you please add the patchsets to a section in https://etherpad.openstack.org/p/horizon-ocata-priorities Ocata-2/3 20:08:49 <tsufiev> o/ 20:09:32 <rdopiera> added 20:09:41 <r1chardj0n3s> just the one patch? ok :-) 20:10:01 <rdopiera> r1chardj0n3s: there are some cleanup patches, but they are independent and not as important 20:10:13 <r1chardj0n3s> righto, that's what confused me 20:10:27 <r1chardj0n3s> I'm easily confused ;-) 20:10:30 <rdopiera> in fact, they all conflict with each other... 20:10:35 <r1chardj0n3s> \o/ 20:10:52 <r1chardj0n3s> so, in a related note 20:10:57 <r1chardj0n3s> #topic swift ui-router update 20:11:22 <r1chardj0n3s> we have previously agreed that ui-router will be used where necessary for multi-level routing 20:11:32 <r1chardj0n3s> I've got the first patch up that uses it in the Swift UI 20:11:47 <r1chardj0n3s> It's still a little WIP, but I'd like feedback now if possible 20:11:53 <r1chardj0n3s> #link https://review.openstack.org/#/c/350523 has a big UI change, so feedback would be nice 20:12:12 <r1chardj0n3s> in short, I nuked the containers sidebar and that's a pretty big change :-) 20:12:47 <r1chardj0n3s> I'd like to know whether I need to try to revert that before continuing work on it 20:12:50 * ducttape_ would like to see where containers go 20:13:04 <r1chardj0n3s> ducttape_: it's ok, to simplify things I just delete them all 20:13:23 <ducttape_> rm -rf /openstack is super simple 20:13:30 <r1chardj0n3s> +2 +a 20:13:31 <ducttape_> ;) 20:13:41 <r1chardj0n3s> so, feedback on the patch would be really appreciated :-) 20:13:45 <r1chardj0n3s> moving on 20:13:49 <r1chardj0n3s> #topic bootstrap update 20:14:05 <r1chardj0n3s> We're upgrading our xstatic angular-bootstrap this week! 20:14:10 <r1chardj0n3s> This is going to break EVERYTHING!!! 20:14:15 <r1chardj0n3s> well, except mitaka and newton 20:14:18 <rdopiera> the end is near 20:14:20 <rdopiera> repent 20:14:30 <ediardo> damn 20:14:31 <r1chardj0n3s> but master will break, as will Ocata's beta1 and beta2 releases 20:14:42 <r1chardj0n3s> we have a plan for minimising the breakage 20:15:34 <r1chardj0n3s> In short, once https://review.openstack.org/#/c/409121/ merges the xstatic-angular-bootstrap 2.2.0.0 package will be released to pypi 20:16:33 <r1chardj0n3s> in openstack land, a patch to upper-constraints.txt will be automatically generated by the release of that file, pushing the pin "XStatic-Angular-Bootstrap===0.11.0.8" to "XStatic-Angular-Bootstrap===2.2.0.0" 20:16:43 <r1chardj0n3s> (yes, we're very, VERY out of date) 20:16:53 <r1chardj0n3s> once that pin changes, all the Ocata Horizons will break 20:17:10 <r1chardj0n3s> so we have a patch we're going to try to release as simlutaneously as possible 20:17:11 * ducttape_ thinks this is going to be very exciting in a terrible way 20:17:28 <tsufiev> Omg 20:17:47 <r1chardj0n3s> https://review.openstack.org/#/c/401245/ 20:17:56 <tsufiev> Should it really make such a big step? 20:17:59 <r1chardj0n3s> that patch ain't perfect, but it does get Horizon working again 20:18:20 <r1chardj0n3s> tsufiev: the backward incompatible changes were introduced in angular-bootstrap 0.13 :-) 20:18:28 <r1chardj0n3s> we've been out of date for a *very* long time 20:18:28 <ducttape_> as long as we are cut from any nearby milestone branches, lets forge ahead 20:19:06 <robcresswell> I'll take a look at the patch again if you like 20:19:09 <r1chardj0n3s> as I said, stable mitaka and newton *should* be unaffected by this, since we've put in a lot of effort to ensure they remain pinned to 0.11.0.8 20:19:10 <robcresswell> I actually thought it fixed most things 20:19:28 <r1chardj0n3s> robcresswell: I saw some styling artifacts, that's all 20:19:39 <robcresswell> Oh I can do styling 20:19:42 <robcresswell> Diana taught me 20:19:45 <r1chardj0n3s> :-) 20:19:57 <r1chardj0n3s> anyway, I don't want messing with that patch to delay it 20:20:42 <r1chardj0n3s> once this meeting is over I'll get on to hassling the release team about getting https://review.openstack.org/#/c/409121/ approved (I'm sure they have nothing better to do ;-) 20:21:18 <r1chardj0n3s> making https://review.openstack.org/#/c/401245/ depends-on the upper-constraints.txt patch that follows https://review.openstack.org/#/c/409121/ should minimise our breakage 20:21:32 <robcresswell> Sure 20:21:56 <r1chardj0n3s> (especially if it's pre-approved and depends-on ...) 20:21:59 <robcresswell> (I think I've solved microversions l support locally btw) 20:22:07 <robcresswell> -l 20:22:21 <r1chardj0n3s> One more thing, then microversions ;-) 20:22:27 <rdopiera> invalid option -l 20:22:37 <r1chardj0n3s> #topic Integration tests are dead, long live integration tests! 20:22:43 <tsufiev> Amen 20:23:17 <r1chardj0n3s> #link https://review.openstack.org/#/c/410489/ is where they went away 20:23:46 <r1chardj0n3s> our task now is to get some new tempest suite tests added in a nice, orderly, well-thought-out manner 20:24:07 <david-lyle> that doesn't sound like us 20:24:11 <lcastell> :) 20:25:02 <r1chardj0n3s> I'm hoping to see a BP that we can use to create an initial small set of tempest additions at high level checking basic panel existence/functionality 20:25:08 * r1chardj0n3s tags lcastell 20:25:09 <r1chardj0n3s> :-) 20:25:15 <lcastell> o/ 20:25:48 * lcastell tags david-lyle 20:25:53 <r1chardj0n3s> I'd like to see breadth of tests, rather than depth... 20:26:14 <r1chardj0n3s> (it'd be super nice for the new suite to even check the angular panels ;-) 20:26:24 * lcastell taking notes 20:26:43 <david-lyle> let's slow down on the unicorn farm 20:26:54 <david-lyle> we'll get some tests going :) 20:27:06 <r1chardj0n3s> dammit, you said there were plenty of unicorns! 20:27:16 <robcresswell> Unicorn farm? 20:27:35 <r1chardj0n3s> speaking of unicorns 20:27:36 <r1chardj0n3s> #topic Microversions 20:27:41 * r1chardj0n3s tags robcresswell 20:27:42 <david-lyle> haha 20:28:20 <robcresswell> Sigh 20:28:37 <robcresswell> So we finished the blueprint for this, amongst much raging 20:28:57 <robcresswell> I've got a patch that should go up tomorrow that does it in the least bad way possible I think 20:29:17 <r1chardj0n3s> (so confident, much microversions, wow) 20:29:18 <robcresswell> And I've asked rhagarty to hold fire on his implementation to save us both time 20:29:42 <robcresswell> Someone else has also been implementing them in isolation and silence, but *shrug* 20:30:14 <robcresswell> r1chardj0n3s: Ha, I'm quite confident in the patch, the implementation is only about 80 lines 20:30:52 <robcresswell> I've used it for locked status on instances, and cinder consistency groups 20:31:01 <robcresswell> As examples :) 20:31:26 <robcresswell> That's all from me anyways 20:31:32 <rhagarty> robcresswell: thanks for driving 20:31:58 <r1chardj0n3s> #link https://blueprints.launchpad.net/horizon/+spec/microversion-support is the BP, for reference 20:32:54 <robcresswell> Thanks r1chardj0n3s 20:32:58 <r1chardj0n3s> I look forward to your patch robcresswell! 20:33:37 <r1chardj0n3s> robcresswell: on the xstatic front, I believe there's some compatibility work needed for D3 as well? 20:33:52 <robcresswell> r1chardj0n3s: Yeah, there is 20:34:00 <r1chardj0n3s> OK, I have someone to work on that then 20:34:11 <robcresswell> Oh good 20:34:24 <robcresswell> I was about to say, kinda losing motivation on doing all our updates 20:34:51 <rdopiera> we should rotate that position 20:34:52 <robcresswell> Would be nice if someone else could drive a couple :) 20:34:53 <r1chardj0n3s> yep, I have helpers now though 20:35:00 <robcresswell> Awesome 20:35:42 <r1chardj0n3s> that's all I have for the agenda today - does anyone have anything they'd like to bring up? 20:35:52 <ying_zuo> robcresswell: are you working on something for the instance locked status? 20:36:07 <ying_zuo> I also have a patch for it. 20:36:16 <robcresswell> ying_zuo: Just adding proper microversion support :) 20:36:32 <robcresswell> Well, consistent I should say 20:36:48 <ying_zuo> I think that's what Cindy mentioned on the patch 20:36:53 <robcresswell> It won't change the functionality, but might require a rebase depending on who merges first 20:37:17 <ying_zuo> https://review.openstack.org/#/c/331874/ 20:37:50 <ying_zuo> I think my patch will need to rebase to yours? 20:37:55 <robcresswell> Ah right yes, the resulting work is in the blueprint r1chardj0n3s put up earlier 20:38:01 <robcresswell> And my patch should go up tomorrow 20:38:10 <tsufiev> I was about to ask if anyone thinks horizon needs profiling of angular pages, https://review.openstack.org/#/c/358574/ 20:38:17 <ying_zuo> cool. thank you 20:38:23 <robcresswell> tsufiev: That would be good 20:38:35 <robcresswell> ying_zuo: No problem 20:38:39 <r1chardj0n3s> tsufiev: yes! thank you for reminding me of that patch! 20:38:53 <tsufiev> Because in terms of ui it may not provide a competitor to Chrome network profiler 20:39:16 <r1chardj0n3s> cores, I'd really like https://review.openstack.org/#/c/395873/ to not be in my sad patches list :-) 20:39:32 <tsufiev> And on the other hand, angular api calls are usually very fine-grained 20:39:39 <r1chardj0n3s> tsufiev: but it can go deeper than Chrome, yeah? it can profile the services as well? 20:39:41 * rdopiera -1s it 20:40:11 <tsufiev> r1chardj0n3s: it can, but the UI as I said is inferior 20:40:44 <r1chardj0n3s> tsufiev: regardless of the inferiority of the UI, being able to see what's going on behind the server AJAX handler will be invaluable 20:40:44 <tsufiev> I mean, it needs to be taught how to show async calls 20:40:54 <r1chardj0n3s> righto 20:40:55 <tsufiev> Ok 20:41:29 <tsufiev> I'll try to improve it a bit then ) 20:41:32 <r1chardj0n3s> especially if we can see *why* the nova API call is taking 30 seconds to respod - all Chrome can tell us is that it does :-) 20:41:57 * r1chardj0n3s likes unicorns, please make more of them 20:42:08 <tsufiev> Hmm, that would require support from the other part of profiling team 20:42:19 <r1chardj0n3s> unicorns, dammit!! 20:42:36 <r1chardj0n3s> ;-) 20:43:10 <rdopiera> https://s-media-cache-ak0.pinimg.com/236x/1d/da/6b/1dda6b78a44783f963717043125e3da9.jpg 20:43:12 * robcresswell is still confused 20:43:24 <robcresswell> Fat deranged unicorn? 20:43:31 <tsufiev> Bloated unicorn ) 20:43:35 <rdopiera> it has thick bones 20:43:36 <tsufiev> Like bloatware 20:43:39 <hurgleburgler> \o/ 20:44:06 <ducttape_> husky 20:44:09 <tsufiev> hurgleburgler: were you summoned by a unicorn ;)? 20:44:18 <hurgleburgler> Apparently so 20:44:21 <hurgleburgler> :-D 20:44:27 <hurgleburgler> a fat deranged one. 20:44:51 <robcresswell> That should be the Horizon logo 20:44:55 <r1chardj0n3s> ok, if there's no other business, I'll let y'all get back to your reviewing 20:44:57 <rdopiera> too late 20:45:05 <r1chardj0n3s> robcresswell: omfg don't talk about the war 20:45:23 <robcresswell> Lol 20:45:25 <r1chardj0n3s> #endmeeting