[00:00] m_3, seems like the timeout should be 5m instead of 2m [00:00] that's a remaininng bug I'll push tomorrow... I adjusted manually [00:01] m_3, yeah.. some of those look ok, when it goes to record the state its working already [00:01] but the watcher timed out [00:01] the current set of runs are with the new timeout [00:01] they seem to be passing [00:02] one strange hang still... hadoop-mapreduce [00:02] green is wrong there...an artifact of how I killed it [00:07] * m_3 out... back on later [00:10] SpamapS, you are awesome === Furao_ is now known as Furao [04:58] <_mup_> Bug #956697 was filed: Maas provider can't start second node for deployment (CloudInitError: Incomplete cloud-init: you need to call set_machine_id) < https://launchpad.net/bugs/956697 > === fenris is now known as Guest98977 === almaisan-away is now known as al-maisan === fenris is now known as Guest77154 === TheMue_ is now known as TheMue === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === carif__ is now known as carif [14:25] What does "I:" items mean in charm proof? I understand W:, etc but no I: [14:30] what's the best way of writing a charm with a generic db relation such that it can support either mysql or postgresql? [14:31] obviously, you could have two separate relationships defined, one for each type, but then you lose the meaning of a required db interface [14:32] it'd be nice if you could say db interface was required in metadata but then specify which implementers can satisfy that requirement [14:35] bbcmicrocomputer, hmm [14:35] that would be nice but is not a current feature [14:35] I did look at the existing charms but didn't find any that have attempted this.. yet [14:36] if you want to support diff database types you will have to specify different relations [14:36] I'd pick one to start with and be opinionated.... [14:36] jamespage: :) === al-maisan is now known as almaisan-away [15:08] marcoceppi: I is just informational. [15:08] Ah, thank you SpamapS [15:08] marcoceppi: short for FYI ;) [15:11] hum, not sure if this is a bug [15:11] I added a new environment to my environments.yaml file [15:11] The name of it is "720", which throws this error: "error: Environments configuration error: /home/marco/.juju/environments.yaml: environments: expected string, got 720" [15:12] nevermind, I wrapped it in quotes [15:13] Okay, back to the real problem, when I bootstrap in EC2 machine 0's state is always "not-started" [15:13] subsequent deploys don't seem to work [15:17] marcoceppi: also you may have problems with the first char not being a number [15:17] marcoceppi: that means the agent isn't starting [15:18] marcoceppi: check the console output [15:18] marcoceppi: what version of juju are you using? [15:20] SpamapS: this is a boostrap/deployment on an environment that's all chars and not numberic, not sure which juju version since juju --version doesn't work but dpkg says it's : 0.5+bzr470-1juju2~oneiric1 [15:21] marcoceppi: ok, and if you ssh to the machine directly, does it have the same juju version? [15:22] One sec, let me check [15:25] marcoceppi: also do check the console output of the instance. I'd bet it had problems starting the machine agent. [15:25] SpamapS: System Log from AWS doesn't show anything [15:27] Not sure how to connect to the machine at this point, AWS Console won't connect because it has no key-pair, juju ssh 0 says "Waiting for machine to come up" and just a straight SSH timesout [15:42] bcsaller: ping [15:42] SpamapS: Did an upgrade on my machine for juju, good to go [15:47] marcoceppi: you could have just ssh'd to the public hostname ;) [15:47] SpamapS: Tried, it didn't accept my connections [15:50] marcoceppi: did you try euca-get-console-output ? (or ec2-get-console-output) [15:50] Nope, I didn't know about that [15:51] marcoceppi: very important commands. :) [15:55] <_mup_> juju/refactor-machine-agent r462 committed by jim.baker@canonical.com [15:55] <_mup_> Fix failing tests due to refactorings in other merges [15:58] robbiew: hey [15:59] bcsaller: just a friendly reminder about objectives ;) [15:59] righto [16:03] jamespage, thanks for that review, it's excellent [16:03] jcastro, no problemo! [16:04] especially the tips to the sections of the documentation, that's clutch for new charmers [16:07] jcastro, I commented on your sun-jdk/hadoop bug - what do we want todo with the existing hadoop charms for oneiric? [16:07] jamespage, mims is at a conference, I was going to wait for his input [16:07] jcastro, sounds sensible [16:07] that charm was mostly his right? [16:07] yeah [16:08] I have a sneaking suspicion post-12.04 that no one will care for hadoop on 11.10. [16:08] so maybe it'll just start down the road to obscurity [16:10] jcastro, well the charm I wrote for 12.04 works with 11.10 as well - backported the packages I did for 11.10 in the hadoop-ubuntu PPA's [16:11] ah ok, so maybe just supercede it [16:12] maybe [16:14] jamespage: backporting charms to oneiric is a good idea. Though I think we'll do that quite a bit less after April :) [16:14] SpamapS, most likely [16:18] SpamapS, that was more of a between now and april type plan from my perspective [16:19] having one set is better I think. [16:19] but hey, I don't want to make claims on someone's baby when they're not around. [16:19] lol [16:20] I'm sure m_3 is listening... [16:20] "hey so welcome back, we threw away the charm you slaved on that was our premier use case for the entire project and stuff. Hope that's ok." [16:20] actually, knowing mims he's probably happy that hadoop is now a james page problem and not his, heh [16:22] "I don't always do hadoop. But when I do, I do it with Sun Java". -- James Page [16:27] * m_3 backscroll [16:28] jcastro: negronjl wrote the hadoop charms... I just got the natty packages building on oneiric [16:29] jcastro: I'm happy with the new charms/packages... and'd love to see the oneiric cdh pkg do an end-of-life [16:29] forget that - I do it with OpenJDK [16:30] ha! [16:30] (its the only option on armhf) [16:30] negronjl, mira, how do you feel about just using jamespage's hadoop charms in 11.10? [16:33] OMG... finally! [16:34] we've got consistently reproducible full-graph charm-tests.... charmtests.markmims.com [16:35] hazmat: whoohoo! [16:35] m_3, nice! [16:36] you guys are unbelievably amazing [16:36] m_3, thats incredibly awesome, nice work [16:36] http://www.youtube.com/watch?v=Vh78T--ZUxY <-- message from the entire juju community to hazmat and m_3 [16:37] * m_3 is in public atm [16:37] m_3: SFW [16:37] in fact, I order you to watch it [16:37] ORDER YOU [16:37] lol [16:37] SpamapS, got a moment to talk about the namespace-from-env branch? [16:37] nice one m_3 [16:38] hazmat: I saw your comments. Very frustrating parsing effort in general and I'm more than happy to shorten that code to a few RE's if that will work.. I just didn't give it much thought after I got the tests passing ;) [16:39] SpamapS, i'm split out the parsing to a separate function on your branch, it simplifies it a little, i'll put a diff link on the merge, outside of that if you can finish up the coverage, i'm fine with it [16:39] s/i'm/i've [16:43] hazmat: er, do you want to just do a new MP that supersedes imine? [16:44] hazmat: this has gotten a bit out of hand for me.. originally it was a 1 line change. [16:44] SpamapS, fair enough, i'll finish it [16:44] hazmat: so at this point its easier for me to move forward with a wrapper than wait for this to pass review [16:44] SpamapS, thanks for getting it this far [16:48] <_mup_> juju/refactor-machine-agent r463 committed by jim.baker@canonical.com [16:48] <_mup_> Expand comment on the reason for machine_id type assertion [16:54] <_mup_> juju/robust-test-removed-service-unit r461 committed by jim.baker@canonical.com [16:54] <_mup_> Merged upstream [17:25] SpamapS, charmstests.markmims.com is back for me now [17:27] jcastro: yeah, ^^ [17:34] * marcoceppi checks in on Minecraft charm [17:38] 2012-03-16 16:09:24,226 juju.service_watch:ERROR unit minecraft/3 startup failed install_error [17:38] SpamapS: Yeah, saw that [17:40] I really wish I could figure out the local provider, I still can't get it to bootstrap or deploy here [17:42] SpamapS: have you ever seen this when deploying? http://paste.ubuntu.com/886618/ [17:42] marcoceppi: yes [17:42] marcoceppi: https://bugs.launchpad.net/juju/+bug/952397 [17:42] <_mup_> Bug #952397: Juju is forcing *ALL* charms in a repo to be perfect to be able to deploy *any* < https://launchpad.net/bugs/952397 > [17:42] jamespage, would you consider the etherpad-lite charm complete enough to be demoable? [17:43] marcoceppi: you need to fix your charms [17:43] ah, thanks [17:45] marcoceppi: if you have the latest charm-tools you can do 'charm update --fix path/to/a/dir' and it will pull all fixes in [17:46] SpamapS: I do, but all the repos are setup against the old charm/* lp branches [17:49] hey does any other charm other than phpmyadmin have a "get from upstream instead" option? [17:50] jcastro: jenkins does [17:50] ta [17:50] rails, node, django do too... [17:50] but they're not in the store except for node [17:50] node does? [17:50] jcastro, no - its broken [17:51] * m_3 looking [17:51] yeah I just need an example [17:51] I'm working on my companion blog post to the store landing [17:51] and I want to show an example of "don't want what's in the archive? Fine, here you go" [17:52] jcastro: yeah, node is using the pacakge... I thought I had npm updateing the node version, but it's only updating (node) package deps... not node itself [17:53] marcoceppi: --fix fixes that :) [17:53] yeah I am waiting for the node ppa guy to get back from SXSW to talk to him about it [17:53] hopefully he'll dig it [17:53] cool [17:53] node is the perfect example of something that we can rock by having an up to date charm [17:53] but do we have one that works now? [17:53] oh yeah [17:53] yeah [17:53] I used it in the mongodb demo [17:53] phpmaydmin works [17:53] node talks to a mongodb replset [17:53] but it's not a very fun charm [17:54] marcoceppi, yeah I am hoping for something sexy and cloudy [17:54] not to take away from your charm of course. :) [17:54] *nod* [17:54] naw, it's PMA, it's neither sexy or cloudy :) [17:54] jcastro: http://paste.ubuntu.com/886640/ [17:55] ok which part there does the upgraded node? [17:55] do you just have that local? [17:57] oops, sorry that was unclear... node just uses the pacakge only... I was wrong [17:57] aha! [17:58] hey so I see jamespage uses "lts" and "trunk" to differentiate his versions [17:58] but marco does it differently in phpmyadmin [17:58] the node guy should update it to get npm to update node itself to the version the app actually depends on [17:58] we should probably best practice one [17:58] yup [17:58] jcastro, those are quite specific to upstream naming [17:58] m_3, yea, that's what I was thinking [17:58] they have LTS and trunk releases [17:58] jamespage, ah ok, so we'll likely run into service-specific conventions then [17:59] but for a jenkins person that totally makes sense then [17:59] jcastro, take a look at the zookeeper charm [17:59] although it's really pretty obvious.... might not need to be so std [17:59] that uses distro OR some PPA options [17:59] m_3, indeed, I am overthinking [18:02] jamespage, brilliant, that's exactly what I need [18:02] actually for PPA's .. [18:03] we really need to be able to override that for system policy. I think a subordinate would be able to do that. [18:03] jamespage, ok so ... you can do dev, testing, or stable, what does the command look like [18:03] like, do I set a config or ... ? [18:04] set a config [18:04] juju deploy --config config.yaml --reposistory .... [18:04] juju set config source=dev [18:04] typically my charms don't support switching [18:04] i.e. that information is use in the install hook only [18:04] switching is a bit too sexy and scary for now I think [18:05] its one more thing to test.... [18:05] I think just an install switch is still way awesomer than what people have to do now [18:07] jcastro, I'm going to switch the etherpad-lite charm to use chris lea's PPA for node/npm [18:07] jamespage, I sent chris a mail [18:07] trying to make it work from upstream git for npm is to hard and unreliable [18:07] I am going to try to work with him on convincing him about the node charm [18:07] everyone uses his PPA for stuff [18:08] we might as well make it a nice option in the node charm [18:08] that would be slick. [18:11] jcastro, arggh [18:11] jcastro: I think the node one should be done like rails... the app specifies which version of node/rails it depends on thru package.json or Gemfile [18:11] etherpad-lite won't run on 0.6 node js [18:11] * jamespage put this down for the week [18:11] I need a beer [18:12] have a nice weekend folks [18:12] you've had a good week james, chill! [18:12] you've earned it! [18:12] jamespage: I can help clean that up next week [18:12] night! [18:12] m_3, are you at a conference still? I'd like to start running my blog post past you guys [18:40] ninjix: welcome :) [18:40] hey [18:40] I'm trying to do some juju with virtualized servers [18:40] ninjix: excellent. That should work. :) [18:41] running 11.10 orchestra and juju [18:41] getting a cannot connect error [18:42] when I run juju status [18:42] virtual machines were PXE installed with Orchestra [18:42] Uh, juju local is prompting for a password when I try to juju ssh [18:42] ninjix: backup.. did you run all the steps? [18:43] marcoceppi: ssh password or sudo password? [18:43] SSH [18:43] marcoceppi: actually it would have to be ssh password [18:43] ubuntu@localhost's password: [18:43] marcoceppi: most likely there was a problem during master-customize [18:43] marcoceppi: though 'localhost' sounds wrong [18:43] I haven't manually copied any ssh keys [18:43] I thought the juju profile was doing that [18:43] ninjix: don't get confused, I'm talking to marco about ssh [18:44] ;) [18:44] ninjix: for orchestra, the steps should be: 1) add VMs to cobbler (orchestra-provisioning-server) including adding them to the available mgmt class. Did you do that? [18:44] SpamapS: I followed the Dustin's blog howto [18:45] ninjix: Ok, so you did 'bootstrap' and 'deploy' already? [18:45] Found out the problem with the Minecraft charm, the "store" doesn't have the latest code. [18:45] Scratch that [18:45] talking out of my butt [18:45] marcoceppi: makes sense because what you just said stinks ;) [18:46] ninjix: I'd recommend getting juju from the juju PPA. The version in 11.10 is fairly buggy. [18:46] SpamapS: I added the mgmt classes but only after I ran into the "no available" error [18:46] SpamapS: did you actually merge the changes I had in to the main charm? [18:47] marcoceppi: not sure [18:47] nbd if you didn't I can merge it [18:47] but the code in charms/minecraft doesn't have my changes [18:47] marcoceppi: usually the proposer, not the approver, merges. [18:47] SpamapS: do the orchestra-juju-available need to be added before the cobbler install runs? [18:47] SpamapS: I'll keep that in mind, thanks [18:47] marcoceppi: the merge proposal would be 'Merged' if the code was merged. :) [18:47] <3 so foolish of me [18:48] ninjix: definitely [18:48] ok [18:48] SpamapS: do I need both *acquired and *available enabled? === almaisan-away is now known as al-maisan [18:58] ninjix: no [18:58] ninjix: only available [18:58] ninjix: the sequence is to set them to 'available', then 'bootstrap', then reboot the box (so it installs w/ juju) [18:59] ninjix: can you paste the HOWTO url you're using? I want to make sure thats what it says. :) [19:01] https://wiki.ubuntu.com/ServerTeam/OrchestraJuju [19:03] that wiki page is missing the default-series [19:04] ninjix: so right after the 'bootstrap' step, you should reboot the server that was acquired. [19:04] SpamapS: thanks, minecraft is updated now \o/ [19:04] ninjix: I just added some clarifying notes to that in the wiki page [19:04] marcoceppi: been a while since I lost myself in minecraft.. might have to go digging for lunch.. :) [19:05] marcoceppi: favorite thing to do: dig a ridiculously deep and long tunnel, then dig back up.. then find the ridiculously tall tower I built at my old base. ;) [19:05] I'm always a lot closer than I think I should be. :-P [19:06] yeah, that's what my first base was, a giant pillar with torches so I could find it from afar :) [19:08] I can't really get into playing on big long living multiplayer servers [19:08] Every time I do.. I come back and all my s*** is gone. [19:09] even when I hide it under sand. ;) [19:10] SpamapS: restarted the kvm and it was reinstalled by cobbler [19:12] ninjix: ok, it should now have zookeeper running, and 'juju status' should work [19:12] just saw the zookeeper get installed on the kvm console's echo [19:15] SpamapS: ahhh... just checked the firewall and a see some blocked traffic between the orchestra kvm's VLAN and the test machines [19:17] SpamapS: juju status working now [19:17] SpamapS: thank you [19:20] going to switch over to the ppa juju now [19:20] ninjix: you're not the first to get confused by that. Hopefully the wiki update will help. [19:20] SpamapS: it did [19:21] ninjix: I added default-series too [19:21] good move === dvestal|away is now known as dvestal [19:29] can juju do a master-master mysql relationship? [19:31] i see in some references to slave instances [19:32] ninjix: yes it can, but there's no charm for that yet [19:32] ninjix: I'd like to add it to the mysql charm at some point [19:32] nice. Let me wrap my mind around these charms and maybe I can help out [19:35] looks like I'm going need to write a Proxmox API extension for provisioning and controlling the kvm [19:36] is that what the "virtualization" settings are for in the Cobbler profile? [19:39] ninjix: dunno. I believe cobbler has support for libvirt and/or kvm [19:40] ninjix: you might want to consider trying the 'local' provider which does things in LXC containers rather than on VMs controlled by cobbler. [19:41] SpamapS: thanks but I am working on project to leverage our mucho macho server nodes [19:42] we've been running proxmox for several years now [19:42] ninjix: ah cool :) [19:43] just saw the first mysql charmed kvm get provisioned [19:43] ninjix: what exactly is proxmox? I only see a mail gateway at proxmox.com [19:43] http://pve.proxmox.com/wiki/Category:Proxmox_VE_2.0 [19:44] SpamapS: mysql install is still pending [19:45] ninjix: yeah, we've been looking at ways to allow you to install all of the nodes in parallel before you know what you want them to do. ;) [19:45] ninjix: since the install takes a few minutes. :-P [19:45] SpamapS: logged into the provisioned mysql charmed instance [19:46] SpamapS: can see mysql there but the juju status still shows state: not-started [19:48] SpamapS: kvm is idling now [19:48] ninjix: interesting. So is there a juju agent running? [19:48] ninjix: because it should have started up, and called back in to the zookeeper node. [19:49] SpamapS: I see it running [19:49] and it installed mysql [19:50] ninjix: ok, can you maybe pastebin your 'juju status' ? [19:50] ninjix: because that state should be 'running' or 'started' or something [19:51] http://pastebin.com/Tz5UCMx9 [19:55] SpamapS: how do you restart the juju agent [19:56] don't see juju in the /etc/init [19:57] SpamapS: I see an error in the machine-agent.log [19:58] SpamapS: but I am not sure if it occurred before I fixed the firewall rules [19:59] ninjix: if you're not on the PPA, you can't restart the agent [19:59] ninjix: the one that shipped in 11.10 just ran it as part of the first boot.. newer versions ship an upstart job for each agent [19:59] ninjix: dpkg -l juju , what does it show? [20:00] SpamapS: 0.5+bzr398-0ubuntu [20:00] that's from the mysql instance [20:01] the orchestra machine is now running 0.5+bzr481-1juju3~oneiric [20:02] SpamapS, hazmat, m_3, I have some stuff to run by you guys [20:02] if you're feeling for a Friday afternoon hangout [20:03] ninjix: yeah thats the old distro version, and you can't restart the agent (and it dies a lot if your servers get overloaded because the timeouts aren't handled right) [20:03] SpamapS, oh, kapil and ben are already hanging out [20:04] SpamapS: so... I should reboot the mysql instance? [20:05] SpamapS: or can I bang on its pid? [20:06] <_mup_> Bug #956000 was filed: 'juju' with no arguments gives confusing message < https://launchpad.net/bugs/956000 > [20:07] SpamapS: killed the juju pid and restarted it with the same command line args [20:07] now I see the error [20:07] SpamapS, invite sent! [20:07] getting "No zookeeper connection configured." along with python trace [20:11] jcastro: sorry, no can do atm [20:13] does that mean it's having trouble talking to the bootstrap? === al-maisan is now known as almaisan-away [20:16] ninjix: it takes stuff from the environment too [20:18] SpamapS: ok [20:19] going to terminate this run and try again [20:25] ninjix: definitely get on the PPA version, much better [20:26] jamespage: I just pushed a new version of subway charm... I have not yet had a chance to deploy myself since I need to setup LXC and I am writing this charm as a favor of sorts ;p [20:26] SpamapS: do I need to change a cobbler snippet to get the instances to use PPA juju? [20:28] jcastro: you did say --repository needs to point to my charm yes? [20:28] ninjix: no [20:28] ninjix: juju pushes the commands to deploy itself every time you bootstrap or deploy [20:34] hi all [20:36] jcastro: http://paste.ubuntu.com/886877/ :( [20:38] you want [20:38] juju deploy --repository= . subway local:subway [20:38] but you need to be in ~Work/Development [20:42] jcastro: juju: error: unrecognized arguments: local:subway [20:42] from ~Work/Development [20:44] one sec [20:44] on a call [20:47] SpamapS: got a little further with this run [20:47] now I see a "could not internally obtain zookeeper handle" error message [20:48] ninjix: hrm [20:55] ninjix: where are you seeing that? [20:55] hmm, why would local:subway not work? [20:55] clint? [20:55] jcastro: because --repository isn't pointing at something witht default-series/subway in it? [20:56] bkerensa: juju deploy --repository=~/Work/Development local:subway [20:56] actually [20:56] bkerensa: juju deploy --repository ~/Work/Development local:subway [20:56] no = [20:56] you need to let the shell expand ~ [20:57] * SpamapS goes to get lunch [20:58] SpamapS: found it. just had to read a little further up in the trace [20:59] SpamapS: DNS strikes again [21:01] SpamapS: same error [21:01] =/ [21:03] jcastro: no idea why its failwhaling [21:04] this is turning into a very productive afternoon. Thanks for your help, SpamapS. [21:05] indeed [21:05] nor me [21:05] hey, I'm going to finish off my charm and then just test yours bkerensa [21:06] jcastro: sure :) [21:06] you're not the first person who can't get it working locally [21:06] feel free to propose any fixes to its bad state [21:06] :D [21:06] but no worries, team effort. [21:14] SpamapS, znc incoming! [21:39] bkerensa: can you do ls -l ~/Work/Development | pastebinit ? [21:39] SpamapS: http://paste.ubuntu.com/886965/ [21:45] bkerensa: and ls -l ~/Work/Development/oneiric ? [22:04] SpamapS: http://paste.ubuntu.com/887000/ [22:07] bkerensa: they look identical.. that can't be right [22:07] bkerensa: did you forget to add /oneiric ? [22:08] SpamapS: add it where? [22:08] it exists [22:08] ~/Work/Development/oneiric/subway is where the charm resides [22:12] bkerensa: I was asking for ls -l ~/Work/Development/oneiric [22:12] bkerensa: also is the name set to 'subway' in metadata.yaml .. that can mess it up too [22:14] SpamapS: http://paste.ubuntu.com/887014/ ^ [22:15] bkerensa: ok cool, and 'cat ~/Work/Development/oneiric/subway/metadata.yaml' ? [22:15] bkerensa: you are almost as good as sshd at running my commands btw.. though I think the connection has a bit more latency than I'd like. ;) [22:15] SpamapS: http://paste.ubuntu.com/887015/ [22:16] bkerensa: and in environments.yaml, do you have 'default-series: oneiric' ? [22:18] SpamapS: where is this located? [22:23] bkerensa: ~/.juju/environments.yaml [22:23] bkerensa: you wouldn't have been able to bootstrap without that [22:25] SpamapS: http://paste.ubuntu.com/887026/ [22:29] bkerensa: interesting. You have two 'environments' sections. You should only have one. [22:29] bkerensa: if that fixes it, we have a bug to fix :) [22:32] SpamapS: Nope I removed the EC2 copy and tried again [22:32] http://paste.ubuntu.com/887037/ [22:34] bkerensa: AHA! [22:34] provides: subway: interface: http [22:34] well, irc ate that [22:34] bkerensa: interface needs to be indented [22:34] what a terrible error message. [22:35] 2012-03-16 15:31:02,246 WARNING Charm 'subway' has an error: MetaDataError('Bad data in charm info: /home/bkerensa/Work/Development/oneiric/subway/metadata.yaml: provides.subway: expected unicode or utf-8 string, got None',) Bad data in charm info: /home/bkerensa/Work/Development/oneiric/subway/metadata.yaml: provides.subway: expected unicode or utf-8 string, got None [22:35] bkerensa: I'm opening a bug to make charm validation more friendly. :) [22:36] actually bug 899433 is already open :) [22:36] <_mup_> Bug #899433: YAML errors in charms should be obvious to users < https://launchpad.net/bugs/899433 > [22:36] hm no [22:37] That wasn't even logging an error [22:37] bkerensa: so, yeah, indent interface: http one more level, and it should work fine [22:40] <_mup_> Bug #957498 was filed: Metadata errors should be user friendly < https://launchpad.net/bugs/957498 > [22:51] SpamapS: That fixed it [22:51] jcastro: My charm works fully now [22:51] I just deployed it succesfully so its working fine here and I just pushed the final changes [22:51] bkerensa: *woot* [22:51] bkerensa: whats your new-charm tagged bug #? I'll review it first since you've been such a good sport working this out. :-D [22:54] SpamapS: Bug #944246 [22:54] <_mup_> Bug #944246: Charm Needed: Subway IRC client/server < https://launchpad.net/bugs/944246 > [22:55] bkerensa: ok, I have one other thing blocking me.. then I'll start review in about 30min [22:55]