| mwhudson | are there many charms for trusty yet? | 00:47 |
|---|---|---|
| mwhudson | or does that sort of thing wait until after trusty is out? | 00:48 |
| Kupo24z | Hey all, is it possible to select a MAAS node when running the first 'juju bootstrap' to setup the juju environment? I have pre-defined nodes and I'd like to make sure it provisions on the right one | 00:56 |
| davecheney | mwhudson: the latter | 01:08 |
| mwhudson | davecheney: is the cheat in the mean time branching all the precise repos to $localcharmrepo/trusty or however you do that? | 01:10 |
| rick_h_ | mwhudson: do you need them to run on trusty? You can just juju deploy precise/mysql for instance? | 01:11 |
| davecheney | mwhudson: correct | 01:11 |
| mwhudson | rick_h_: i am mostly interested in architectures that are only supported in trusty :-) | 01:11 |
| davecheney | rick_h_: that is the main rason | 01:12 |
| rick_h_ | mwhudson: if you download zip of a charm you can drop it on the gui and pick the series to use | 01:12 |
| rick_h_ | mwhudson: ah gotcha | 01:12 |
| mwhudson | well i guess armhf was in precise but i still want to be closer to the bleeding edge | 01:12 |
| mwhudson | rick_h_: ok, bit tedious when dealing with a gigantic bundle | 01:12 |
| mwhudson | (unless there is a trick for that too) | 01:12 |
| rick_h_ | mwhudson: definitely, no trick until we get fat bundles and the like | 01:13 |
| mwhudson | ah ok | 01:13 |
| === arosales_ is now known as arosales | ||
| === vladk|offline is now known as vladk | ||
| === CyberJacob|Away is now known as CyberJacob | ||
| === vladk is now known as vladk|offline | ||
| === CyberJacob is now known as CyberJacob|Away | ||
| === vladk|offline is now known as vladk | ||
| melmoth | hola juju people ! With juju 1.18.0 i have been told one can tell juju, when using openstack provider, to use a given neutron network for vms (so it does not die when the current tenant as several available networks). | 08:12 |
| melmoth | any idea actually how to tell juju wich net to use ? | 08:12 |
| melmoth | ahhh, most probably the new network setting that is in environment.yaml | 08:24 |
| frankban | rbasak: morning. the new quickstart package(in your ppa) seems to work very well | 08:31 |
| rbasak | frankban: great, thanks! I'll upload. | 09:02 |
| frankban | rbasak: cool thank you | 09:03 |
| rbasak | (done) | 09:08 |
| === Kyle____ is now known as Kyle | ||
| === Kyle is now known as Guest44217 | ||
| === Guest44217 is now known as Eliz | ||
| noodles775 | I'm having to `sudo pkill mongod` each time I re-bootstrap a local environment. Is this known, or the same issue as bug 1208430 ? | 11:35 |
| _mup_ | Bug #1208430: mongodb runs as root user <mongodb> <juju-core:Triaged> <juju-core (Ubuntu):Triaged> <https://launchpad.net/bugs/1208430> | 11:35 |
| melmoth | i m experiencing problem with juju 1.18.0 tyring to use it behind a proxy. | 13:58 |
| melmoth | https://pastebin.canonical.com/108082/ | 13:58 |
| melmoth | if anyone has an idea, you are welcome | 13:58 |
| === hatch__ is now known as hatch | ||
| mgz | melmoth: can you access the maas API directly using HTTP_PROXY envvar and wget say? doesn't seem like a juju related error | 14:16 |
| melmoth | i dont need a proxy to access MAAS, and the proxy had no access to MAAS directly | 14:17 |
| melmoth | (all my boxes, MAAS as well are vm in a hypervisor;hosted on some libvirt network). The proxy is a real proxy that my main hypervisor must use to access external boxes | 14:18 |
| melmoth | so if i try to access the maas box via the proxy, it will not work.. is it what is going on here ? | 14:18 |
| mgz | melmoth: printenv? | 14:20 |
| melmoth | mgz http://pastebin.com/UsBbLR5u | 14:21 |
| melmoth | (i m using juju directly from the box where maas is installed) | 14:21 |
| mgz | melmoth: that's odd. what about /home/ubuntu/.profile and /home/ubuntu/.juju-proxy? running juju from inside of the maas... that it's using itself? seems shakey. | 14:26 |
| melmoth | i always did it that way. | 14:26 |
| melmoth | i mean, juju is just a client, it should be ran from...anywhere | 14:26 |
| melmoth | i could fire up yet a new vm just for it, but, i never had to, nor do i see the point | 14:27 |
| melmoth | lets have a look at the profile | 14:27 |
| melmoth | http://pastebin.ubuntu.com/7226462/ | 14:28 |
| mgz | well, I have no idea where your juju client is getting the proxy setting from then. | 14:29 |
| melmoth | well, from the nevironement most probably, i did set juju set http-proxy | 14:29 |
| melmoth | because i need the new machine to use the proxy | 14:29 |
| melmoth | that was the whole point of the exercice | 14:29 |
| melmoth | but i dont need communication from the juju box or the bootstrap node to MAAS to be proxyed | 14:30 |
| melmoth | hmmm | 14:30 |
| melmoth | may be i should set no-proxy for the maas box ? | 14:30 |
| mgz | well, you are telling the bootstrap node to use a proxy with that | 14:30 |
| mgz | affecting the local client when uploading a charm is a little odd | 14:30 |
| mgz | melmoth: you probably should | 14:31 |
| melmoth | i m just not sure what name to use or wich actual machine it s trying to cennect to | 14:31 |
| melmoth | let s try with 192.168.101.2 | 14:32 |
| melmoth | \o/ | 14:34 |
| melmoth | thanks mgz :-) | 14:34 |
| mgz | workies? | 14:34 |
| melmoth | looks like it did. At least i have a prompt. it s a subordinate so nothing happened, that s normal. let s continue deployment | 14:35 |
| melmoth | yeah, mysql seems to be under deployment | 14:37 |
| mgz | ace. | 14:40 |
| === alexlist is now known as alexlist` | ||
| cory_fu | So, I'm new to bzr. If I want to submit a (set of) change(s) for review on charm-helpers, do I just do bzr push lp:~johnsca/charm-helpers/name-for-branch in my local copy? | 16:06 |
| timrc | arosales, Hey I pushed a fix for: https://bugs.launchpad.net/charms/+source/jenkins/+bug/1272520 -- but it's pushed to a +junk branch so I can submit an MP | 16:08 |
| _mup_ | Bug #1272520: Unable to relate to jenkins-slave: hook failed: "master-relation-changed" <audit> <jenkins (Juju Charms Collection):Confirmed for lazypower> <https://launchpad.net/bugs/1272520> | 16:08 |
| * timrc really doesn't want any responsibilities associated with being a member of ~charmers, but this particular bug has bit us, and thought I would kill it with fire | 16:09 | |
| marcoceppi | cory_fu: yes, then do a bzr lp-propose lp:charm-helpers | 16:09 |
| lazyPower | timrc: right on. I'll take a look closer to EOB unless you need it landed immediately. | 16:09 |
| lazyPower | and follow marco's directive ^ | 16:09 |
| timrc | lazyPower, Nah, take your time. Not hugely critical. We are deploying from a branch atm... there is some weirdness around authentication too but I think that's separate | 16:09 |
| timrc | lazyPower, Right-o. Thanks | 16:10 |
| arosales | timrc: thanks for the mp. I think I had filled a abug against jenkins . . . | 16:10 |
| arosales | timrc: https://bugs.launchpad.net/charms/+source/jenkins/+bug/1272520 was the bug I was hitting did you see something similiar? | 16:11 |
| _mup_ | Bug #1272520: Unable to relate to jenkins-slave: hook failed: "master-relation-changed" <audit> <jenkins (Juju Charms Collection):Confirmed for lazypower> <https://launchpad.net/bugs/1272520> | 16:11 |
| timrc | arosales, Well I didn't see any logs, so I can't be 100% sure but my master-relation-changed problem was due to the fact that the master-relation-changed hook was expecting the slave-relation-joined hook (run by jenkins-slave) to set 'slaveaddress' which it wasn't | 16:12 |
| timrc | arosales, so master-relation-changed was assigning an empty string and passing that to hooks/addnode and causing breakage | 16:12 |
| arosales | timrc: sounds promissing that your branch may resolve this issue | 16:13 |
| arosales | timrc: thanks for taking the time to contribute the fix back into the juju community :-) | 16:14 |
| === JoshStrobl__ is now known as JoshStrobl | ||
| timrc | marcoceppi, Just to confirm my lp-propose should be: | 16:43 |
| timrc | Source: lp:~timrchavez/charm-helpers/jenkins-slave-fix-slave-relation-joined | 16:43 |
| timrc | Target: lp:charm-helpers | 16:43 |
| timrc | should result in* rather | 16:44 |
| marcoceppi | timrc: is this a fix in charm-helpers or the jenkins charm? | 16:44 |
| timrc | marcoceppi, it's a fix to a jenkins charm... when I attempted to propose my branch for merge against the actual charm I got a little hate :( | 16:45 |
| timrc | bzr: ERROR: exceptions.Exception: lp:~timrchavez/charm-helpers/jenkins-slave-fix-slave-relation-joined is not mergeable into lp:charms/jenkins-slave | 16:45 |
| marcoceppi | timrc: you should push your branch to lp:~timrchavez/charms/precise/jenkins/jenkins-slace-fix-slave-relation-joined | 16:46 |
| marcoceppi | timrc: you're trying to merge not only across projects, but across distros | 16:46 |
| timrc | marcoceppi, muhaha | 16:46 |
| timrc | if I could push to charms/ I would have.. | 16:47 |
| marcoceppi | timrc: you can | 16:47 |
| marcoceppi | timrc: it's charms in your namespace | 16:47 |
| marcoceppi | anyone can push to lp:~user/charms/... | 16:47 |
| timrc | marcoceppi, Ah, I had some problem... maybe my path was bad | 16:47 |
| timrc | marcoceppi, Ah, I attempted bzr push lp:~timrchavez/charms/precise/jenkins-slave-fix-slave-relation | 16:48 |
| marcoceppi | timrc: yeah, since charms is a distro, you have to put a branch name in there and jenkins becomes the project name | 16:48 |
| timrc | Makes sense. Thanks for working around my obtuseness and helping me :) | 16:50 |
| marcoceppi | timrc: np! once it's all pushed, you can just do `bzr lp-propose lp:charms/jenkins-slave` | 16:50 |
| timrc | lazyPower, Okay stumbled my way through :) https://code.launchpad.net/~timrchavez/charms/precise/jenkins-slave/jenkins-slave-fix-slave-relation-joined/+merge/214994 | 17:03 |
| === natefinch is now known as natefinch-afk | ||
| === roadmr is now known as roadmr_afk | ||
| === roadmr_afk is now known as roadmr | ||
| === vladk is now known as vladk|offline | ||
| tvansteenburgh | anyone ever successfully configured sticky sessions on the haproxy charm? | 18:24 |
| marcoceppi | tvansteenburgh: people in IS probably have | 18:32 |
| tvansteenburgh | marcoceppi: thanks, will check there | 18:34 |
| === CyberJacob|Away is now known as CyberJacob | ||
| Jeffrey_ | Is there anyway to start a service on the juju agent machine? Or is there a way to deploy to a specific machine number in juju status? | 19:03 |
| marcoceppi | Jeffrey_: yes, you can use deploy --to | 19:06 |
| Jeffrey_ | marcocappi: Then I get "Added charm "local:percise/rethinkdb-2" to the environment. ERROR cannot assign unit "rethinkdb/0" to machine 0: series does not match" and the machine type of number 0 is "series: percise" | 19:07 |
| marcoceppi | Jeffrey_: is this a local provoder? | 19:08 |
| Jeffrey_ | Yes from a directory on the juju client computer. | 19:08 |
| Jeffrey_ | Also if I add a charm it creates a new machine that stays in the pending state. How can I start it? | 19:11 |
| marcoceppi | Jeffrey_: the bootstrap node isn't a real cloud VM, it's your actual machine | 19:11 |
| marcoceppi | you'll need to add another machine | 19:11 |
| Jeffrey_ | marcoceppi: Ok so the bootstrap node can only be for the agent? | 19:11 |
| marcoceppi | Jeffrey_: only on local provider | 19:11 |
| marcoceppi | Jeffrey_: all other real clouds you can deploy --to 0 | 19:12 |
| Jeffrey_ | marcoceppi: I'm sorry I meant I am running a MaaS provider. | 19:12 |
| Jeffrey_ | I am running from a local repository | 19:12 |
| Jeffrey_ | Charm is from a local repository | 19:13 |
| marcoceppi | Jeffrey_: show me your juju status | 19:13 |
| Jeffrey_ | marcoceppi: | 19:14 |
| Jeffrey_ | http://pastebin.ubuntu.com/7227684/ | 19:14 |
| Jeffrey_ | marcoceppi: Even when I just do a normal deploy the new machine that is allocated for the charm doesn't start. Is there any way for force it? | 19:17 |
| marcoceppi | Jeffrey_: something else is wrong, it should start up eventually | 19:18 |
| Jeffrey_ | marcoceppi: This is the status: http://pastebin.ubuntu.com/7227701/ | 19:19 |
| Jeffrey_ | macroceppi: I think I figured it out. | 19:22 |
| === mwhudson- is now known as mwhudson | ||
| avoine | I'm thinking about changing some configuration variables for the django charm like: requirements_pip_files | 21:06 |
| avoine | in your opinion what would be better:? | 21:06 |
| avoine | 1) having a configuration option with a default but ignoring if the results fail | 21:06 |
| avoine | 2) a configuration variable with empty default and triggering errors if the file is not there | 21:06 |
| marcoceppi | avoine: the latter, in my opinion | 21:12 |
| tvansteenburgh | `juju deploy local:trusty/haproxy --repository=.` | 21:23 |
| tvansteenburgh | is the --repository=. needed there? | 21:23 |
| tvansteenburgh | i'm hacking on haproxy and my code changes aren't getting deployed to new units | 21:24 |
| tvansteenburgh | not using --repository and wondering if i need to | 21:24 |
| tvansteenburgh | marcoceppi: ^ | 21:25 |
| marcoceppi | tvansteenburgh: pretty sure you always need to specify a repository | 21:25 |
| tvansteenburgh | or else? | 21:25 |
| tvansteenburgh | maybe it's pulling from lp everytime and i haven't noticed | 21:26 |
| marcoceppi | no, it wont do that | 21:26 |
| tvansteenburgh | yeah i dunno what it's doing then | 21:27 |
| marcoceppi | tvansteenburgh: you can try with the --debug flag | 21:36 |
| marcoceppi | see what's happening | 21:36 |
| tvansteenburgh | well --repository=. fixed my immediate prob, so i'm moving on | 21:36 |
| arosales | mbruzek: marcoceppi does charm-help-test work for you? | 22:25 |
| arosales | on charm-tools 1.2.10 its not giving me the help | 22:25 |
| arosales | sorry I meant to state, charm-help test | 22:26 |
| arosales | charm help test, also isn't working for me. | 22:26 |
| mbruzek | no I get a usage when I try those commands | 22:27 |
| arosales | http://paste.ubuntu.com/7228396/ | 22:27 |
| arosales | I'll file a bug | 22:27 |
| arosales | mbruzek: thanks for confirming I was hoping to look at what the valid outputs are for charm tests | 22:28 |
| mbruzek | arosales, does "charm test" not work for you? It appears that the help is broken for me | 22:28 |
| arosales | mbruzek: 'charm test' does work for me | 22:29 |
| arosales | mbruzek: just that 'charm help test' does not or as the usage suggests 'charm-help test' also doesn't work | 22:29 |
| mbruzek | sure | 22:29 |
| mbruzek | That is a bug | 22:29 |
| marcoceppi | avoine: charm test -h | 22:32 |
| marcoceppi | should give you output | 22:32 |
| arosales | https://bugs.launchpad.net/charm-tools/+bug/1305337 | 22:32 |
| _mup_ | Bug #1305337: charm help test does not work on 1.2.10 <Juju Charm Tools:New> <https://launchpad.net/bugs/1305337> | 22:32 |
| arosales | marcoceppi that does work | 22:32 |
| arosales | perhaps it just the usage that needs updating then | 22:33 |
| marcoceppi | arosales: charm help <cmd> doesn't ever work? | 22:33 |
| marcoceppi | i guess it could | 22:33 |
| arosales | it didn't work for me in the commands I tried it with, list, test, get, etc. | 22:33 |
| arosales | cory_fu: fyi http://paste.ubuntu.com/7228419/ | 22:35 |
| arosales | which I finally got to give me from charm test -h | 22:35 |
| arosales | cory_fu: pass, fail, skip, and timout, so were hitting timeout | 22:35 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!