[11:14] <Spaulding> hello folks! is there any explanation how to do ssl termination on haproxy juju charm?
[13:58] <stokachu_> arosales, can you do some deploys of spark-processing from the edge channel of conjure-up?
[14:16] <xilet> Possibly stupid question, but how do I deploy multiple copies of a charm to the same juju (physical) system, I keep getting ' application already exists' when trying to deploy a second copy of a local charm
[14:17] <jrwren> xilet: did you give it an alt name? Most charms aren't written to support this, but if it is your charm and it supports this, it should be possible?
[14:19] <stokachu> xilet, juju add-unit
[14:19] <stokachu> xilet, you can see juju help add-unit
[14:19] <stokachu> if you want to deploy to same system under multiple containers
[14:29] <xilet> thanks
[14:29] <xilet> Was exactly what I was missing
[16:48] <cory_fu> kwmonroe: Crap.  http://pastebin.ubuntu.com/23955241/
[16:49] <cory_fu> kwmonroe: 2.1-beta5's autoload-credentials will create a cred for LXD but it's missing a field and can't actually be used
[16:50] <kwmonroe> wait cory_fu, didn't this work yesterday?
[16:50] <cory_fu> kwmonroe: I didn't actually check that the credential it created would bootstrap because it looked the same as what you manually created.  I didn't notice the missing field
[16:51] <kwmonroe> ha!
[16:51] <kwmonroe> i sense "cat" in your future.
[16:51] <cory_fu> :)
[17:13] <kwmonroe> cory_fuuuuudge:  https://travis-ci.org/juju-solutions/layer-cwr/builds/199679703
[17:14] <icey> is it possible to export a bundle from the cli rather than the GUI?
[17:15] <kwmonroe> icey: i've always had to fire up the gui to do it.  if you find a way, lmk.  that would be super useful to me.
[17:15] <icey> kwmonroe: that's what I was thinking :-/
[17:15] <icey> wonder if I can hack up something to grab it from the gui with a cli command :-P
[17:16] <kwmonroe> icey: munging "juju status --format=yaml" to get the right fields might be worth looking at too
[17:17] <icey> kwmonroe: in essence, what I want to do is get the current non-controller model, create a new model with a matching deploy, and destroy the old model :)
[17:17] <icey> yaml output may be perfect but we'll see
[17:19] <kwmonroe> icey: i feel like you're on a complicated path to eseentially "rename" a model, but by golly, go for it because i need a cli export too ;)
[17:19] <icey> well, I want to re-deploy local charms into the new model ;-)
[17:19] <icey> kwmonroe: not just rename it
[17:19] <kwmonroe> heh... suuuuure..
[17:20] <icey> kwmonroe: 3/4 of my current model is local: charms :)
[17:39] <catbus1> stokachu: where is conjure0 interface defined?
[18:12] <catbus1> brctl shows conjureup0 isn't attached to any interface.
[18:15] <catbus1> that's the conjureup0 created in openstack with novakvm on maas cloud
[18:15] <catbus1> the conjureup0 in openstack with novalxd on localhost works.
[18:38] <catbus1> stokachu: i found where conjureup0 is created. nm.
[18:56] <kwmonroe> cory_fu: travis is busted, but https://github.com/juju-solutions/layer-cwr/pull/59.  just merge it.  do it.
[18:57] <cory_fu> kwmonroe: Gotta wait for travis to pass.  ;)
[18:57] <kwmonroe> one sec, just need to remove .travis.yml right quick
[18:58] <cory_fu> :)
[18:58] <cory_fu> I wonder what's going on wit Travis, tho.  It's also messed up for charms.reactive
[18:59] <cory_fu> kwmonroe: Merged
[18:59] <kwmonroe> methinks something went south with a charm-tools dep.  like maybe python-libcharmstore used to be available, but now it's not, and that's breaking the env setup.
[18:59] <kwmonroe> thx for the merge
[19:21] <stokachu> catbus1, its in our spells lxd profile
[19:54] <catbus1> stokachu: I don't find it. I see conjureup0 gets created in the conjure-up init bridge.start, but I don't find which physical interface is attached to it.
[19:54] <catbus1> I can't find  it.
[19:55] <stokachu> catbus1, it's attached to same interface lxdbr0 is
[19:55] <catbus1> stokachu: is it still the same if using maas as the cloud?
[19:55] <stokachu> yea should be
[19:56] <stokachu> it'll just nat just like lxdbr0
[19:57] <catbus1> stokachu: I don't see lxdbr0 on the maas/conjure-up node.
[19:58] <stokachu> catbus1, oh right, sorry
[19:58] <stokachu> catbus1, maas different ballgame
[19:58] <stokachu> conjureup0 is only useful with novalxd and localhost
[19:58] <stokachu> maas is all up to you to setup
[19:58] <stokachu> conjureup0 just gets added automatically b/c we can't determine early enough what type of provider you'll be deploying too
[19:58] <stokachu> so it's a kind of prepare for every scenario
[19:59] <catbus1> so with maas as a cloud, should I manually attach the physical interface to the conjureup0?
[20:00] <catbus1> so that maas/conjure-up node can ssh to the openstack instance on a different physical server.
[20:00] <stokachu> catbus1, yea that's up to you to configure
[20:01] <catbus1> ok
[21:36] <arosales> stokachu: just landed from Gent, but will get some deploys from edge on spark processing when I get back into the home office
[21:38] <stokachu> arosales, cool
[22:28] <skay> just started seeing this exception when I try to deploy postgresql today https://paste.ubuntu.com/23956963/