[07:47] <kjackal> Good morning Juju world!
[10:29] <anrah_> Is there a way to log what states are active?
[10:30] <anrah_> oh, bus.get_states() seems to do that
[11:10] <erik_lonroth_> kjackal: I'm trying it again. For my presentation I had to fake it so I would still like to succeed with this.
[11:11] <kjackal> hi erik_lonroth_ help me remember is that for proxing bigtop?
[11:12] <erik_lonroth_> No, I think I managed to get through that. It was stuck on "configuring spark...."
[11:12] <erik_lonroth_> I'm deploying the bundle again as we speal.
[11:12] <erik_lonroth_> speak.
[11:12] <kjackal> cool, I am here to help
[11:13] <erik_lonroth_> kjackal: Its great! I'm very glad for your help.
[11:13] <erik_lonroth_> Are you working for Canonical today?
[11:13] <admcleod_> hehe
[11:15] <kjackal> erik_lonroth_: yeap!
[11:15] <kjackal> admcleod_: still here! :)
[11:16] <admcleod_> kjackal: syncharitiria!
[11:18] <kjackal> admcleod_ chur
[15:07] <lazyPower> anrah: Just as a belated follow up - charms.reactive get_states (when in a hook context) will dump a dictionary of the states on the unit.
[15:11] <boolman> I'm trying to wrap my head around this issue I have when deploying openstack, http://ix.io/xKB
[15:12] <boolman> the node clearly has IP's in the range, but for some reason it fails and cannot find a suitable IP
[16:30] <kwmonroe> erik_lonroth_: how did the spark deploy go?
[16:42] <Budgie^Smore> o/ juju world
[17:09] <kwmonroe> \o Budgie^Smore!
[17:17] <lazyPower> o/ Budgie^Smore
[21:22] <bdx> put your docker containers on lxdbr0 when 'docker.ready' https://gist.github.com/jamesbeedy/39829a14fdc64583dfda4a6be1812aea
[21:22] <bdx> :)
[21:23] <bdx> possibly something similar could be worked into layer-docker?
[21:28] <bdx> a simple bridge option ... I guess that all goes out the window when k8s and network plugins come into play though
[21:55] <magicaltrout> our new k8s cluster has 350TB  of ram \o/
[21:55] <tvansteenburgh> wat
[21:56] <tvansteenburgh> how many nodes
[21:56] <magicaltrout> quite a lot
[21:57] <magicaltrout> i dunno tvansteenburgh i'm waiting for it to come online but we got quoted the core and ram size of the whole cluster
[21:58] <magicaltrout> its DARPA's attempt to make machine learning automatic, machine learning for machine learning if you will.
[21:58] <tvansteenburgh> and you're gonna put cdk on this?
[21:59] <magicaltrout> yeah we have a small cluster going for test stuff so i was on the call an hour ago about getting our test stuff migrated out of our openstack testbed onto bare metal
[22:01] <tvansteenburgh> so how will this gianormous cluster be provisioned? MAAS?
[22:02] <magicaltrout> dunno about that yet, that would be my preference of course, have to sell it higher up as well though
[22:47] <Budgie^Smore> sounds really cool magicaltrout