[00:00] cool, ty [00:17] SpamapS: oh and about the mouse thing, i use photoshop and other mouse tools WAY too much not to be a mouse guy [00:19] actually on my desk i have a mouse, a blootooth multitouch track pad, ( for my left hand while mouse in the right in photoshop ) and i use my ipad off to the side connected with wifi to change brushes etc with adobe connect and then a bamboo pen tablet wireless i place infront of my keyboard when i use it , so many more "hand movement" tools i use too much but ive got my own kinda gestures and shortkuts i do with them [00:21] oh and my guitar hero usb controller at the side of my chair for when i cant rock out on the telecaster ( /me wants a new les paul :( ) [00:21] heh [00:22] imbrandon: perhaps consider Mavis Beacon? [00:22] ok, time to go do real world stuff [00:22] :) [00:24] i land one of those two i was telling ya about and i'll take a full real class and mavis :) just for you [00:32] hazmat, wasn't sure if you saw but Euca provided a possible patch for bug 907450 [00:32] <_mup_> Bug #907450: juju does not work with Walrus when s3-uri has a suffix < https://launchpad.net/bugs/907450 > [01:16] hasp, yeah.. i've been talking to brian [01:16] hasp, i'm at a conf at the moment, what we can do about the archives first thing next week [01:16] er. but i'll see ^ [01:45] cool [01:45] thanks man [01:45] hows the conference btw [02:56] I'm having some trouble trying to deploy-test my first charm and every piece of literature i find tells me the same command set that i'm using - however its not finding my in-dev charm. [02:57] lazyPower: what's your command? [02:57] http://paste.ubuntu.com/936340/ [02:58] i suspect i haven't got something setup correctly [02:58] lazyPower: quick fix, inside your repo folder, create an oneiric folder and put your charms in their [02:58] there* [02:59] well that caused it to barf with a different message, \o/ thanks marco [03:00] np, what's the new message? [03:01] http://paste.ubuntu.com/936345/ [03:02] well, that's interesting, pastebin your metadata.yaml [03:02] You can also install charm-tools, then in the terria folder run charm proof . [03:02] I did that but didnt understand the output [03:03] let me man it [03:04] i dont have a provides: statement [03:04] that'll do it [03:05] marcoceppi: oneiric? You mean precise!! [03:05] be bold, taste the beta [03:05] SpamapS: Just trying to get it working at this point - i'll worry about targeting control in the next iteration. [03:05] if i cant even deploy the service thats cart before the horse isnt it? :) [03:06] lazyPower: newer versions of juju will default to precise anyway [03:06] its specified in my environment.yaml [03:06] so it'll be looking for oneiric [03:06] lazyPower: if you're using the latest from the PPA, and you don't have 'default-series: oneiric' , then you will be spawning precise anyway [03:06] ahh ok [03:06] lazyPower: btw, the provides is just a suggestion. Anything W: or I: is not going to break your charm. [03:06] in the provides: statement can i be arbitrary with my definition or are there guidelines for it that i'm missing? [03:07] lazyPower: provides is a structured dictionary. It needs two levels deep under it. the first level is the name of the relation, the second is the interface [03:07] provides: [03:07] foo: [03:07] interface: bar [03:08] Ok i see, nice. [03:08] meaning "I provide something I call foo, and it can talk to things that require an interface: bar [03:09] Theres something wrong here - let me push this upstream to my github repo and maybe you can tell me why juju is barfing on the deploy statement. [03:09] lazyPower: but you should be able to deploy without a provides [03:09] proofing is coming back with only a w [03:09] can you pastebin the barf too? [03:09] certainly [03:09] ahh [03:10] marcoceppi: I noticed the gravatars are broken on omg [03:10] lazyPower: your structure should be /home/ubuntu/repo/oneiric/terraria/metadata.yaml [03:11] is there a charm command i need to run on ~/repo to make it charm friendly that I may have missed? [03:11] lazyPower: no [03:11] its just a dir [03:11] with the releases of ubuntu under it as dirs [03:11] then each charm gets a dir [03:11] with metadata.yaml, revision, and the hooks dir [03:11] i thought that - just analyzing the situation [03:12] what's your metadata.yaml look like? [03:12] lazyPower: the 'charm create' tool in charm-tools definitely creates a nice skeleton charm, but it doesn't manage the repo for you [03:12] err, if you're going to push the whole charm to github, that works too [03:12] incoming [03:12] http://paste.ubuntu.com/936352/ [03:12] oh that reminds me... [03:12] I should push the latest charm-tools into precise before the release [03:14] i used the bitlbee charm as a guideline, and the documentation to write it - it all appears valid to me. [03:15] It appears fine [03:18] ok the latest charm revisions are up on github as well. [03:18] i suspect this is something funky going on with my setup not so much the charm itself unless i'm just horribly daft. [03:18] cool, I'll try a quick deploy [03:21] lazyPower: easy fix, metadata.yaml "name" needs to be lowercase [03:22] the name needs to match the directory name of the charm [03:22] oh... [03:22] derp! [03:22] <3 [03:23] well on the bright side i'm finally getting around to pulling the latest precise updates on this server [03:23] heh [03:23] so the charm trouble had purpse then [03:25] marcoceppi: lets put that in charm proof! [03:25] thats an E: for sure [03:25] nice [03:26] just deployed successfully [03:26] lazyPower: ^5 [03:26] that is hot sauce, gentlemen, hi5 indeed. [03:28] marcoceppi: about done w/ a patch [03:30] interesting, is it common to see nothing in the deub log when you have juju-log commands in the install hook? [03:30] *debug [03:34] lazyPower: you should see stuff in the debug-log as soon as the unit starts its agent [03:35] lazyPower: IIRC, there is a bug where debug logging isn't turned on until after the install hook runs.. but I thought that was fixed [03:35] ooohhh - so i think what i'm experiencing is the nice lag in AWS on spinning up a new instance from the controller then [03:35] makes sense. [03:35] i had a machine that was pending - so it just assumed things were deployed. [03:36] marcoceppi: https://code.launchpad.net/~clint-fewbar/charm-tools/name-must-match-dir/+merge/102617 [03:36] lazyPower: you also have to wait for juju to install itself.. :) [03:36] lazyPower: in my experience, on m1.small's, it takes about 2 minutes [03:36] well i'm refreshing on my AWS panel - and no new instance spun up [03:36] lazyPower: still "pending" ? [03:37] lazyPower: you should also see any logs from the provisioning agent [03:37] lazyPower: I have to go afk for a bit.. but I'll be back to check on you soon. :) [03:37] Thanks :) [03:55] * marcoceppi reviews [03:58] SpamapS: The only thing I would say, shouldn't it be higher than a warn? as it's a blocker by Juju standards [04:22] awesome. I think i was a victim of AWS Roulette on my last controller. Just deployed a brand new controller and issued a deploy for my charm and the debug log is logging up a storm. [04:23] This is exactly what i needed guys, thanks for the patience. [04:54] marcoceppi: its not though, you can be submitting it to lp:..... and have it in "foo-bar-baz-bang" [04:54] marcoceppi: its just a helpful warning === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [12:08] SpamapS: cool [12:09] SpamapS: approved === VivekVC is now known as Guest44342 === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [14:37] Hi, we've got some scripts that wait for our charm to be deployed before proceeding. It involves polling 'juju status' waiting for the agent-state to transition to 'started'. When deploying to ec2 we see occassional error such as: "ERROR Connection to the other side was lost in a non-clean fashion: Connection lost.". Is this just a problem with the ssh layer that we should live with or problem juju should solve? === AlanChicken is now known as AlanBell [16:30] hello! We have a community contributor testing out the MAAS installation and juju deployment but he's running into a timeout error bootstrapping the environment. The error he gets is: Failure: juju.errors.ProviderInteractionError: Unexpected TimeoutError [16:30] interacting with provider: User timeout caused connection failure. [16:30] 2012-04-19 10:43:45,297 ERROR Traceback (most recent call last): [16:30] Failure: juju.errors.ProviderInteractionError: Unexpected TimeoutError [16:30] interacting with provider: User timeout caused connection failure. Is there a place to see more details about this error? [16:35] matsubara: not that I've heard of... most of the hw crew are at the openstack summit atm (talking about maas) [16:37] matsubara: juju -v bootstrap might help a bit [16:38] and yes, we're all at openstack [16:38] keynotes going now [16:38] SpamapS, http://pastebin.ubuntu.com/937068/ [16:42] matsubara: seems rather clear, timed out trying to connect to ... something ;) [16:43] SpamapS, is there any other juj log or more verbose option I could look at to see what that something is? [16:46] matsubara: not really in this case no. [16:46] matsubara: whatever is raising that ProviderError is being very terse, and should have included a message === matsubara is now known as matsubara-lunch === almaisan-away is now known as al-maisan [17:16] sabdfl about to take stage at openstack conference :) [17:17] \o/ [17:17] current .... keynote .... boring. [17:17] losing .... conciousness ... [17:31] hah [17:41] http://openstack.org/conference/san-francisco-2012/ like the new wheels jcastro [17:41] thanks! === matsubara-lunch is now known as matsubara === carif_ is now known as carif === carif_ is now known as carif === rog is now known as Guest27265 === Guest27265 is now known as rogpeppe [18:00] AlanBell: I wanted to go to that event so bad === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [18:38] hi, i'm deploying openstack with juju, on ubuntu precise. 'juju deplay mysql' is ok, but 'juju deploy rabbitmq' give Error processing 'cs:precise/rabbitmq': entry not found [18:39] i'm using doc from here: https://help.ubuntu.com/community/UbuntuCloudInfrastructure [18:45] <_mup_> Bug #985812 was filed: Juju commands return SSL errors < https://launchpad.net/bugs/985812 > [18:52] and bzr branch lp:charms/rabbitmq give "bzr: ERROR: Not a branch: "http://bazaar.launchpad.net/~charmers/charms/precise/rabbitmq-server/precise/"." Strange, because it's a copy/paste from https://code.launchpad.net/~charmers/charms/precise/rabbitmq-server/trunk (Get this branch) [18:59] I have rebooted my laptop after an lxc juju setup. how do I restart it ? [19:20] and "Browse the code" (https://code.launchpad.net/~charmers/charms/precise/rabbitmq-server/trunk) on lp give 404 ! Is launchpad charms broken ? [20:28] flepied: juju destroy the env and re bootstrap it' [20:48] imbrandon, I don't want to loose what I have done. there is no other way ? [20:49] I manage to modify bootstrap to not stop and restart at least zookeeper. now I need to start the containers... [20:49] not with lxc no [20:50] flepied: its an open bug [20:50] whith ec2 how would it work ? [20:50] flepied: with EC2 you can reboot the nodes, though if the zookeeper node changes IP, things stop working. [20:51] SpamapS: ok [20:52] SpamapS, what is the bug number ? [20:54] SpamapS: you can change the EIP and things still work [20:54] juju ssh unit/# stops working, but juju ssh machine# still works. The data for serviecs falls out of sync, but machine data stays up to date [21:01] marcoceppi: you can't change the internal IP of the zookeeper node. Bad things happen. [21:01] SpamapS: Oh, for the zk node [21:01] I thought it was any instance [21:01] flepied: bug #955576 [21:01] <_mup_> Bug #955576: 'local:' services not started on reboot < https://launchpad.net/bugs/955576 > [21:02] I guess I need to test my newest charm against precise now [21:08] SpamapS: did you see sabdfl's keynote? [21:18] is there video > [21:18] ? [21:24] imbrandon: eventually [21:24] marcoceppi: yes, it was awesome [21:24] not AWSome.. but Awesome :) [21:24] lol [21:25] m_3: http://charmtests.markmims.com/ ... can I get some precise up in here? === matsubara is now known as matsubara-afk [21:51] Is it just me, or should make be a part of build-essential? [21:51] grr [21:53] charm-upgrade didn't quite do what I expected, nvm [21:53] marcoceppi: agreed, I'm shocked that make isn't part of build-essential [21:54] * SpamapS tried to write a charm using make the other day. If there had been a puppy around I'd have kicked it [21:54] haha [21:54] I just kicked the jr. dev over here instead, and by kicked I mean shot with nerf guns [21:55] In other news, I think the shelr.tv charm is ready to be reviewed, and it uses mongodb! [21:55] just waiting for this last test to finish [21:56] nice [21:56] SpamapS: should I make a bug for these: 2012-04-19 21:55:11,412:14468(0x7f66884a5700):ZOO_WARN@zookeeper_interest@1461: Exceeded deadline by 86ms they seem exceisve [21:57] why can no one build a fskin sane php ppa ? fskin grumbles and goes off to do it himself possibly [21:57] marcoceppi, there is one extant [21:57] hazmat: cool, thanks [21:57] marcoceppi, its unfortunate they chose to do log that as a warning, cause its really not meaningful [22:01] yeah, I always scratch my head at it like "oh, okay?" [22:02] SpamapS: you know after a couple of days of looking re: ssh key server, their really isnt a universal one, or one peroid etc except LP, and me being the way I am I think I'm just gonna write one ... in Go ... public , secure , and indep not tied to github sf.net lp etc ( but will likely make use of them for imports or syncing ) [22:03] kinda like a gpg keyserv [22:04] bad idea ? i'm kinda on the fence about it [22:06] imbrandon: lol [22:06] imbrandon: glad you did the due dilligence [22:10] i figured if nothing else it will give me a "project" to learn Go with [22:10] SpamapS: i tend to retain more that way , and maybe later i can get my hands dirty with juju go easier [22:11] SpamapS: yup... working on getting charmtests up to precise... charmrunner's gotta take environment params first [22:11] SpamapS: congrats... I hear the talk went well today! [22:18] Does juju run with the locale env? [22:19] marcoceppi, i think not in the local provider [22:20] I'm wondering if that's why I'm getting this error, the command fails from the install hook, but not from the terminal [22:20] the fix [22:22] hum, doesn't appear to have LANG set, I just tried to echo it from the install hook === Furao_ is now known as Furao [22:22] I guess the easy way would be to just set it? [22:27] So two things, A) That fixed it B) what, if any ENV variables are included during execution? [22:28] marcoceppi: jsut evho env [22:28] echo [22:28] from the install hook and and it should show [22:28] err run env [22:28] not echo [22:28] /usr/bin/env [22:28] eh, I will if I have to push an update [22:48] before I submit a patch for this, can someone confirm that this is a bug? https://bugs.launchpad.net/charms/+source/mongodb/+bug/985939 [22:48] <_mup_> Bug #985939: Mongodb charm does not expose the port on relation-joined < https://launchpad.net/bugs/985939 > [23:05] jcastro / m_3 : Juju on OSX updated , run "brew update && brew upgrade juju" if you have it installed to get juju-0.5+bzr531 :) [23:06] imbrandon: cool [23:13] marcoceppi: when you get a moment can you pass me the two curent stanza's we're using ? [23:19] shelr.tv charm \o/ http://ec2-50-19-40-77.compute-1.amazonaws.com/records/4f909d76c2183956b1000002 [23:19] imbrandon: there's nothing in the new stanzas [23:19] everyting's back on the old setup [23:20] ahh ok, is thre a current staging ? or not right now ? [23:20] there's no staging [23:20] atm [23:20] kk [23:22] marcoceppi: shelr's cool [23:22] yeah, I love the work they've been doing [23:23] and it's another charm we can add to the list that uses mongo [23:24] need to clean up a few things and add a copyright, then it'll be up for review [23:24] yup [23:29] yea i did your charm juju install of omg playback via the shell on osx [23:29] that is soooooooo cool [23:29] marcoceppi: ^^ [23:29] alot smoother that way too