=== cmagina_ is now known as cmagina === natefinch-afk is now known as natefinch === mwenning is now known as mwenning-afk [08:01] jamespage, https://code.launchpad.net/~gnuoy/charms/trusty/cisco-vpp/amulet/+merge/277787 if you get a moment [08:48] gnuoy, merged [08:48] ta [09:43] gnuoy, https://code.launchpad.net/~james-page/charms/trusty/odl-controller/doc-updates/+merge/277797 [09:45] jamespage, don't you need to reference the config.yaml at line 44 of the diff? [09:45] gnuoy, yes pushed [09:45] ta [09:49] jamespage, approved [10:00] gnuoy, I'm working on refreshes to the bundles for liberty and the 15.10 charm release this morning - we should add that to the release checklist [10:01] ack, good catch [10:03] jamespage, done [10:23] gnuoy, mind if I do a trivial on all of our charms for "maintainer: OpenStack Charmers " [10:23] ? [10:23] jamespage, please do, thanks [10:55] gnuoy, done [10:55] gnuoy, I tied some pretty crappy summary and description bits at the same time [10:56] tided rather [10:57] sounds good, thanks [11:14] gnuoy, can you sanity check: [11:14] https://code.launchpad.net/~james-page/charms/bundles/openstack-base/bundle/+merge/277798 [11:14] and [11:14] https://code.launchpad.net/~james-page/charms/bundles/openstack-telemetry/bundle/+merge/277799 [11:14] I've just deployed telemetry in dellstack and run through the README ok [11:14] telemetry == base + ceilometer [11:17] 'num_units: 0' looks odd but since you've deployed with it is looks like it does no harm for subordinates [11:17] gnuoy, thats a gui-ism [11:17] jamespage, why are the mcast ports explicitly stated for a couple of services? [11:17] gnuoy, again I have not idea why the gui elected to add those... [11:18] haha [11:18] they could be dropped [11:20] gnuoy, charmstore does some verification of bundles [11:20] https://jujucharms.com/u/james-page/openstack-base [11:20] and https://jujucharms.com/u/james-page/openstack-telemetry [11:20] that validates the format and the charm revs... [11:20] the bundles look ok to me [11:23] gnuoy, ok [11:23] thanks === blahdeblah_ is now known as blahdeblah [16:10] hey folks - what's the difference between "destroy-environment local" and "destroy-environment local --force"? help tells me that force destroys "directly through the env provider", but i don't know what that means for local. === jamesmil_ is now known as jamesmillerio [18:30] kwmonroe: hi [18:30] hey asanjar1 [18:31] kwmonroe: thanks for v9 version. Maria's team could use you onsite for few hours to resolve the remaining issues.. do you have time? I have already asked arosales [18:32] asanjar1: is the caffetteria food any good? [18:32] (and yeah, i can be there around 1:45 this afternoon) [18:32] kwmonroe: I wounder if there is something unique with their environment [18:32] kwmonroe: no man, it SUCKS [18:33] kwmonroe: I am losing weight [18:33] okay let me ping here === natefinch is now known as natefinch-afk === Guest12059 is now known as med_ === ericsnow is now known as ericsnow-afk === _thumper_ is now known as thumper [20:01] anyone have a tldr on the current state of the jenkins charm? [20:01] it hasn't been touched for over a year so it's either really awesome or horrible. :) [20:05] huh wait a minute, it is updated, the jujucharms changelog is broken?! [20:06] jcastro - correct. its been receiving some attention. The changelog in the store has been silently sometimes-functional due to the v3->v4 transition [20:08] does juju support ipv6? I can't seem to juju ssh to a machine with ipv6 public-address the proxy-ssh netcat fails. http://pastebin.ubuntu.com/13334236/ [20:09] *doh* nevenmind. i was misinterpreting the message. works fine [20:15] jrwren woo! \o/ === natefinch-afk is now known as natefinch [20:21] lazypower: ack, filed a bug [20:26] any write amulet tests which execute on an ipv6 environment? :) [20:52] o/ jcastro, jenkins charm has action in the review queue to, to bring feature parity of the python rewrite back up to le ole bash precise jenkins charm. [20:52] too, even. [20:55] beisner: do you guys maintain it? you guys being the openstackers [20:56] jcastro, no. but we consume an old diverged fork, and would like to switch to consuming a converged, updated, works-for-everyone version, ie. the python rewrite + back-filled features. [20:56] jcastro, so, by virtue, contributing to the thing, testing the thing, making proposals of the thing, could be pretty close to "maintaining" ;-) [20:59] bash charms <3 [21:01] :) === natefinch is now known as natefinch-afk [22:36] hey, ive got a charm questino wrt coordination between two services. i have service A that will eventually get a relation to B. when it does, B will do some stuff on A's API. before it can, service A needs a complete relation to service 1 and 2 (say, DB and message queue) [22:37] is there a good way to interrogate that on B's side, short of adding some data in the relation settings exported by A that B can use as a test? [22:37] its been a while since ive juju'd, but i remember dealing with this often. wondering if a good practice has come up since [22:38] jamespage, maybe you know [22:40] adam_g: I think the normal way is that B doesn't set it's api details on the B-A relation till it has B-1 and B-2 relations up and whatever real interfaces for those up and ready [22:40] so, A will not get the relation config stuff it needs to start the api poking till B is really ready to act on it [22:41] juju now has nice reporting in status where a charm can say it is "blocked" for these sorts of situations [22:41] mgz, well, thats complicated by the fact that the API endpoint isn't advertised via relation, its published into a catalog (read: keystone) [22:42] mgz, ya, im looking at the status stuff now in the openstack charms. will have to go find the docs on that [22:44] adam_g: there really isn't a way to interogate b's relations, Juju doesn't set up that kind of dependency train, otherwise A becomes co-dependant to Bs architecture. Relation data is the best way to handle this [22:44] adam_g: https://jujucharms.com/docs/1.24/reference-status is some of it, there are some new tools for setting that from hooks [22:45] marcoceppi, yeah, was hoping to get around it without having to extend existing stuff [22:45] adam_g: if you can't set the api details directly, you can just have a api-ready: bool type thing on the relation config [22:45] mgz, thats what i was thinking, thanks