=== andrewsmedina_ is now known as andrewsmedina [02:08] <_mup_> Bug #957620 was filed: A spec for how to cleanly stop agents. < https://launchpad.net/bugs/957620 > [02:08] <_mup_> Bug #957621 was filed: A spec for how to cleanly stop agents. < https://launchpad.net/bugs/957621 > [02:08] argh.. lbox [02:08] niemeyer, ^ [02:09] niemeyer, lbox should use the closest open milestone and shouldn't reduplicate a bug. ie mp before bug [02:10] * hazmat files a bug [02:10] it originally had the branch attached to both bugs [04:12] <_mup_> Bug #957675 was filed: juju debug-hooks doesn't use exit code as hook exit code < https://launchpad.net/bugs/957675 > === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [12:34] q === Leseb__ is now known as Leseb [12:43] m_3, we can still see you :P [12:43] ha! [12:43] just working through some config issues [13:41] <_mup_> Bug #957906 was filed: Environment settings explicitly set. < https://launchpad.net/bugs/957906 > [13:51] <_mup_> juju/hook-alias-expansion-redux r472 committed by kapil.thangavelu@canonical.com [13:51] <_mup_> address review comments [14:32] whoops.. wrong lbox subcommand [14:34] hmm... interesting lbox doesn't seem to invoke the post commit hooks, probably because of the temp dir usage defeats a locations setup === koolhead17|away is now known as koolhead17 [16:21] <_mup_> juju/repository-broken-charms r477 committed by kapil.thangavelu@canonical.com [16:21] <_mup_> rearrange charm exception hierarchy, charmnotfound is not a charm structural error, add catch all handler to repository for unexpected errors [16:25] hey all - so is juju always written lower case unless it starts a sentence? [16:26] yes [16:26] Juju or juju, never JuJu though [16:27] ok - I'll leave it as Juju then :-) [16:27] thanks jcastro [18:07] jamespage, we using jetty for anyting you've charmed? [18:58] jcastro or SpamapS: If you would like to have a look at Bug #944246 when you get a chance and grab the more frosty copy of my branch and have a look :D [18:58] thanks [18:58] <_mup_> Bug #944246: Charm Needed: Subway IRC client/server < https://launchpad.net/bugs/944246 > [18:58] I am frosty now! [18:58] looking! [19:14] bkerensa, hmm, lxc problems, I need to sort them [19:16] bkerensa, oh it's working, just getting it all started took a while [19:25] jcastro: So its working nice for you? [19:25] still waiting for it to come up [19:25] (it has to fetch the OS and stuff, clean container) [19:50] bkerensa: there's something else broken in the charm. I got a failure in lxc because 'make' wasn't available. [19:54] <_mup_> juju/ns-from-env r476 committed by kapil.thangavelu@canonical.com [19:54] <_mup_> merge trunk [19:54] <_mup_> juju/ns-from-env r477 committed by kapil.thangavelu@canonical.com [19:54] <_mup_> pep8 and minor refactor [20:21] SpamapS, my LXC isn't working at all [20:21] I blame you [20:22] SpamapS, do I still need to blow away something in /var/lib or whatever on occasion? [20:23] jcastro: the most common problem is precise mirrors being inconsistent with 'hash sum mismatch' [20:24] jcastro: check data-dir/units/master-customize.log [20:24] jcastro: that process needs to detect errors but does not.. :p [20:25] jcastro: if you see any problems reported in that log, destroy-environment and bootstrap again should fix it [20:25] jcastro: if you still have problems.. try killing /var/cache/lxc/* [20:26] hmm.. [20:27] i'm currently working on that lxc-wait issue, but its a bit time consuming [20:27] jcastro, if you have any logs those would be helfpul [20:33] hazmat: the problem I've hit twice has been apt-get failing. Its not detected during master-customize. [20:45] SpamapS, doing anything right now? [20:46] I have an idea === koolhead17 is now known as koolhead17|zzZZ [21:13] hazmat, hey so [21:14] how do you do shared environments in juju? [21:14] say marco and I wanted to share managing an environment? [21:16] SpamapS: Make isnt available by default? [21:16] =/ [21:17] jcastro: ^ what? Your the juju guy =s [21:19] much to learn about deploying in a way I can share with someone else [21:20] jcastro, create a new key and share along with the environment.yaml snippet [21:20] ie. ssh key explicitly specified by content in the environment.yaml snippet is probably the most self-contained way [21:21] hazmat, ok so from looking at how mims set ours up for another project [21:21] he just put all our ssh keys in authorized-keys: [21:21] and then we just share the snippet right? [21:22] jcastro, cool, that sounds nice as well, yes [21:26] hazmat, since OMG guy can't set up wordpress I'm going to do it in juju for him [21:36] hazmat, and that worked perfectly. [21:36] man, awesome [21:56] <_mup_> juju/force-upgrade r464 committed by kapil.thangavelu@canonical.com [21:56] <_mup_> per review only raise error if flags differ on upgrade, tolerate empty upgrade nodes (old format) [21:59] hazmat, still there? [22:05] jcastro, yup [22:07] nm found it, just looking for where juju puts the mysql root pw [22:07] we've got OMGs data and we've set up wordpress [22:08] hazmat: I've added a relation from WP -> HA, but HA's URL is showing the apache page [22:08] but the haproxy isn't doing what we think it should be doing [22:08] If I go directly to the WP URL it shows the WP install [22:09] marcoceppi, odd, is it pointing at the right address? it might be the wp only sets itself up for serving off the public address of the unit, ha is going to do the request via its private networking address [22:10] Ah, so I'll need to mod the apache conf to accept connections from all then? [22:10] That makes a lot of sense actually [22:10] did the charm change recently? [22:10] I thought this worked ootb [22:11] jcastro, it would if the proxy is using the public address, or the apache is setup for that vhost on both public/private addresses [22:12] jcastro, in particular it wouldn't be an issue anywhere public and private address are the same, but it would matter for ec2 [22:12] or openstack [22:13] hazmat, ah so that would be why when I've demoed it on my laptop it just worked [23:48] marcoceppi, d0od is here ^ [23:48] d0od: o/ [23:48] Howdy! [23:53] Hopefully we can put something in place for you shortly d0od [23:54] ok so picking an m1.small won't work [23:54] marcoceppi: Very much appreciated [23:54] going to have to go large [23:55] hazmat: is there any way to specify a specific unit's size? or is that still a global juju config? [23:55] marcoceppi, till next week yes [23:56] hah man [23:56] nice timing [23:56] hazmat: how well does juju take to rebooting units outside of Juju? [23:56] we're running oneiric with origin:PPA [23:57] marcoceppi, which provider? [23:57] ec2 [23:57] marcoceppi, jcastro it should work in theory, i haven't tested in practice though [23:58] jcastro: I'll bootstrap a quick juju setup on my ec2 and test [23:58] ok [23:58] everything running on the machine is upstartified [23:58] hazmat, so basically I thought m1.smalls would be enough [23:58] but we did the redirect and the load on the wordpress node was like 98 [23:58] jcastro, huh.. interesting [23:59] that would be a nice jujitsu tool [23:59] grow/shrink a unit [23:59] * marcoceppi *nods furiously* [23:59] juju add-unit --type=m1.xboxlarge wordpress [23:59] is what I want