=== Zetas_ is now known as Zetas [05:34] x58: yes [05:37] stub: Unfortunately adding a layer.yml doesn't fix the ignore issue :-( [05:42] x58: I don't think there is anything else I can do in the apt layer. You might need to make a local fork and remove README.md from it until it can be addressed in charm-tools. [05:43] Maybe this is why nobody else documents their layers ;) [05:46] stub: README.md is also coming from the layer:basic :P [05:46] If I only have a single layer I pull in layer:basic for example, ignore works just fine, if I pull in multiple layers, things go sideways. [05:47] rm build/mycharm/README.md :P [05:47] before doing a charm push . works :P [05:48] stub: Appreciate the work you do :-) === petevg_ is now known as petevg [06:57] Hello Juju World! === frankban|afk is now known as frankban [08:27] yesterday everyting went fine with juju, today im started new installation, same way as yesterday, but when i try to deploy application is got: panic: unknown channel "edge" [08:27] goroutine 1 [running]: [08:27] panic(0x1a24400, 0xc8200d7320) [08:33] Hello there, if i install neutron-gateway via juju (trusty-liberty) the neutron-gateway doesn't have l3 agent installed. [08:39] KpuCko: What version of Juju are you on? [08:51] 2.0 [08:53] zeestrat http://pastebin.com/MBusU0U5 [08:56] KpuCko: It looks like you might be on the beta15 release. There was an update to the charm store which breaks that release (https://blog.jujugui.org/2016/08/30/jujucharms-com-updated-with-new-channel-support/). I suggest you update your Juju to beta16. [08:56] how to update? [08:57] im using ubuntu 16.04 lts with the repository package [08:57] i do apt-get update && apt-get upgrade, and juju stay with the same version [08:58] same issue when i add latest stable release ot juju with ppa repository [09:04] KpuCko: Ah, I see. Juju is currently transitioning to version 2.0, but it is still in beta release. Unfortunately due to the transition the current default docs and packages in the stable repository are a bit of a mess. If you want to continue to use Juju 2.0 beta releases, then I suggest you add ppa:juju/devel repo and install beta16. Otherwise you can [09:04] install juju-1.25 package which is the current stable one. [09:05] magicaltrout: I have a trouble with JUJU now.... [09:05] http://imgur.com/BCOZSpC [09:05] any idea? [09:06] Hi. How can I set bugs-url and homepage for my newly developed charm? Please anyone respond to this. [09:06] I think the environments.yaml is wrong, but I have no idea what is wrong with it [09:20] magicaltrout: the rest of it: http://askubuntu.com/questions/819506/juju-bootstrap-fails-with-openstack-cloud-installer [09:34] PCdude: The screenshots you're posting don't show the error. Can you put the log file you're showing in nano on a pastebin (like paste.ubuntu.com) instead? [09:36] Hi! I am trying to add a Raspberry pi running Raspbian-linux as a new machine to to my juju model. juju add-machine ssh:pi@xxxx tells me that it cannot find provisioning script. How can I add a provisioning script? I cannot find a clear place in the source code where the provision scripts are generated? [09:37] I tried to make my Raspbian to idnetify itself as a 'xenial' instead of 'jessie', but still i get the same error. Maybe it is because of the ARM-architecture? [09:37] I am using juju 2.0beta15 [09:38] I would like to start using juju in our whole server infrastructure, but Raspberry Pi-based monitoring devices are part of that and I would like to control the lifecycle of the software therein using juju actions [09:39] babbageclunk: here is the pastebin: http://pastebin.com/42JXtMmD . this is the output of ~/.cloud-install/*.log [09:39] I would imagine if I add the Raspberry Pi as a new machine to my model, then deploy a charm to the machine, I should be able to manage them using juju? [09:40] babbageclunk: the error what I get is the above one in this question: http://askubuntu.com/questions/819506/juju-bootstrap-fails-with-openstack-cloud-installer [09:40] without actually instantiating any virtualization in the Raspbian itself === degville is now known as degville_ === degville_ is now known as degville [09:42] PCdude: The juju output (the line starting "Problem during bootstrap") is still truncated - it looks like you cut and paste from within nano? [09:43] PCdude: If you install pastebinit you can paste a file directly without needing to cut and paste it. [09:43] babbageclunk: sorry that is all there is.. maybe I can higher the verbose level? [09:44] PCdude: no, I'm sure there's more in the file, it's just that the line is long so nano isn't displaying it. [09:45] babbageclunk: yup u are right, I thought I copied all of it [09:46] babbageclunk: let me make another pastebin [09:46] one second [09:46] PCdude: use pastebinit - it's a command-line util that will let you put the whole file into the pastebin. [09:47] babbageclunk: here is the pastebin: http://pastebin.com/ujM62KZH [09:47] babbageclunk: I will install that program next, did not know that exsisted [09:56] babbageclunk: is that enough information? [10:01] PCdude: sorry, on a call [10:04] babbageclunk: np, I am just curious what the problem is so I can fix it [10:11] zeestrat thanks a lot [10:11] babbageclunk: I am afk for20 minutes, but I am back after that [10:13] PCdude: I reformatted the last bit of the Juju output http://paste.ubuntu.com/23119674/ [10:14] You can see that downloading the agent failed because it couldn't get to https://streams.canonical.com/juju/tools/agent/1.25.6/juju-1.25.6-trusty-amd64.tgz [10:14] does that mean the node does not have internet? [10:14] I can download that, so possibly there's a connectivity problem on your end? [10:14] PCdude: I think so [10:16] babbageclunk: uhm strange, if u look a little earlier in the dump file, u can see it is reaching for the ubuntu archives and it gets there too [10:16] when I click the link of JUJU it works on my own PC, so no firewall crap or something like that [10:17] Hmm, not sure then [10:17] I will aquire a node in MAAS and try to download that package manually, just to check that it can get there [10:18] PCdude: I was just about to say the same thing [10:19] I think maas does some proxying for apt packages, so that might be working while general connectivity from a node to the internet isn't. [10:20] babbageclunk: uhm good point, that is true. [10:20] MAAS is deploying a node as we speak [10:36] babbageclunk: ok it is deployed how do I login again with the SSH key? (been some time haha) [10:36] PCdude: I think you had to specify a public key in maas? [10:37] yeah did that, only forgot the last letter of the key.... I am in [10:38] and bingo, it can't get to the internet [10:40] babbageclunk: the server with MAAS installed has internet, but the nodes should route through that node [10:42] PCdude: can it resolve DNS? Something I always forget is to set the upstream DNS - under Settings in the MAAS UI. [10:43] babbageclunk: haha, just what I was checking and yes it can resolve DNS it can also ping the gateway, but just not to the outside [10:43] let me check the settings again [10:43] maybe some stupid setttings [10:44] PCdude: Ok, sounds like you're on it [10:45] babbageclunk: at least thanks for the help, I will try some stuff and will keep u updated [10:45] PCdude: :) === ant_ is now known as Guest4912 [11:06] babbageclunk: how can I manage routing in MAAS? [11:07] PCdude: I'm not sure, sorry [11:09] babbageclunk: every apt command can indeed reach the internet. pings to the gateway succeed and dnslookups succeed too. ping an ip address outside the LAN fails and also to a hostname [11:12] PCdude: is the gateway for the subnet the nodes are on the IP of the controller, or the switch? [11:13] PCdude: My knowledge gets a bit spotty here, you might need someone else's help [11:13] its the IP of the controller [11:13] babbageclunk: ah ok, yeah the MAAS channel is pretty silent most of the time, maybe u know another channel? [11:26] PCdude: You could try setting it to the IP of the switch instead (not sure that's right though). [11:27] PCdude: You could ask on the maas mailing list? [11:28] http://askubuntu.com/questions/717803/openstack-install-problem-with-juju-bootstrap/718820#718820 [11:28] PCdude: this is the point where I'd often ask dimitern ;) [11:28] babbageclunk: that sounds promising [11:28] uhm ok haha, I am gonna remember that name [11:28] :) [11:28] PCdude: I'm having a look now [11:28] good nice! [11:30] PCdude: AFAIK MAAS 2.1, currently in recently release alpha1 has some support for managing routes via the API [11:31] ah ok, well to try openstack and landscape I will need to stick to the 1.x version of MAAS [11:31] JUJU is still beta and does not work for me yet [11:31] dimitern: but the link I provided might be a solution [11:31] PCdude: but otherwise, earlier versions do the bare minimum, like allow you to specify different gateways for different subnets, which might not work that well [11:31] dimitern: indeed, very very basic stuff only [11:32] PCdude: if you only need to allow nodes to access the internet via NAT, that's easy - do basically what's described in that askubuntu answer [11:33] yeah, right now, that is sufficient, but I know myself pretty good with those things and I will be fucking around with those files within 1 week [11:33] dimitern: are there maybe extra programs I can add to manage all that? [11:34] PCdude: I'm sure there are more user-friendly ways to manage a firewall on linux than iptables CLI :), if that's what you're asking about [11:35] but usually I never bother with these [11:41] dimitern: added the firewall rules, lets try again :) [11:53] dimitern: and babbageclunk many thanks I have internet now on my nodes! [11:53] PCdude: \o/ awesome! :) [11:54] PCdude: woot! === natefinch-afk is now known as natefinch === infinityplusb_ is now known as infinityplusb [14:15] ah the chaos of my life [14:15] now to spin up demo #2 [14:28] is there a force flag for application removal? I have an app that has no machines attached to it that i can't remove [15:13] not that i'm aware of cholcombe [15:13] anything in the logs associated with the application? [15:15] lazyPower, lemme check [15:15] lazyPower, nadda [15:16] single service left in teh model? [15:16] is it related to anything? [16:04] <__szilard_cserey> hi [16:04] <__szilard_cserey> hi [16:06] cory_fu kwmonroe https://github.com/juju-solutions/charmbox/pull/55 [16:06] need an express on this if we can land it before seman comes in we'll have CWR fixed in time to intercept todays runs with a working build [16:07] lazyPower: Am I reading that right that installing snaps depends on systemd? [16:08] yep [16:08] ... [16:08] well at least system logging [16:08] Why? [16:08] *sigh* [16:08] there's no init, so theres no logging socket [16:08] same issue we ran into with conjure-up that stokachu fixed [16:08] i'll get a bug filed and xreference [16:08] and we can follow up on teh snappy list [16:12] I need to NAT juju gui from LXD to an external port [16:12] not websocket protocol [16:12] what have i forgotten? [16:13] actually ssh portforward may suffice [16:14] nope [16:14] gaa annoyance [16:15] oh it does [16:15] if i pass the full url [16:15] nice [16:17] https://ibin.co/2taneCgvEp1T.png [16:17] tonights demo [16:17] all in LXD [16:18] lazyPower: Reviewed and tested. +1 [16:18] nice magicaltrout [16:18] cory_fu ta, waiting for mbruzek to give final +1 and we'll cut a release [16:32] perfect, merged and cut [16:32] thanks for the help gents *hat tips* [16:47] any idea on this one lazyPower [16:47] The following errors occurred while retrieving bundle changes: cannot read bundle YAML: cannot unmarshal bundle data: yaml: unmarshal errors: line 1: cannot unmarshal !!str `xenial` into charm.legacyBundleData [16:47] exported my stuff [16:47] err [16:47] and tried to import it [16:47] pastebin the bundle? [16:47] on a new controller [16:47] http://paste.ubuntu.com/23120903/ [16:48] imported via the gui i presume? [16:48] same juju installation , just exported my lxd model and wanted to import it into a aws one [16:48] yeah [16:49] ok hang on [16:49] i'm trying from cli first [16:49] then i'll try import into a gui === frankban is now known as frankban|afk [16:54] command line works lazyPower don't worry about it [16:54] gui bug i guess [16:54] well it appears to be a gui bug [16:54] yeah [16:54] we should get that filed [16:54] can you do it, i'm well short on time before i need to head down [16:55] (pretty please) [16:55] you betchya [16:55] thanks [16:55] magicaltrout - which log did you get that output from? [16:55] browser or unit log on the controller? [16:55] the error message? [16:56] its just spat out by juju gui to a warning modal [17:01] ok [17:14] magicaltrout https://github.com/juju/juju-gui/issues/1966 and https://bugs.launchpad.net/juju/+bug/1619389 [17:14] Bug #1619389: juju-gui fails to parse exported bundle in beta-16 [17:15] thanks lazyPower === bcsaller_ is now known as bcsaller [17:16] magicaltrout - its apparently fixed, will come with the next release of hte gui [17:17] k [17:19] Anita [17:20] that you are! === Anita is now known as Guest46979 [17:20] I am getting following error [17:20] root@ptcvm3:~# juju bootstrap local.lxd-test localhost Creating Juju controller "local.lxd-test" on localhost/localhost Bootstrapping model "controller" Starting new instance for initial controller Launching instance ERROR failed to bootstrap model: cannot start bootstrap instance: unable to get LXD image for ubuntu-xenial: The requested image couldn't be found. root@ptcvm3:~# [17:20] during juju boorstrap [17:20] any idea, why I am getting [17:21] someone else had that yesterday [17:21] i think the solution was to make sure you're on beta16 [17:21] beta16 has already an issue [17:21] Guest46979 - update juju-2.0 by adding the devel ppa, and then running sudo apt update && sudo apt upgrade juju-2.0 [17:21] thats why I downgraded [17:21] to beta15 [17:21] thats beta's for you [17:22] Yes, Its beta15 [17:22] indeed [17:22] which doesn't work with lxd [17:22] beta15 does not work with lxd? [17:22] nope [17:22] but beta 16 does [17:22] but in one of the other m/c i have beta15 [17:23] which works fine [17:23] Guest46979 - beta16 added the image fetching routine and made the entire lxd bootstrap experience more robust [17:23] lazypower_: do you suggest me to apply beta16? [17:24] Guest46979 - yep, here's the release notes of what changed https://jujucharms.com/docs/stable/temp-release-notes === lazyPower changed the topic of #juju to: Welcome to Juju! || Docs: http://jujucharms.com/docs || FAQ: http://goo.gl/MsNu4I || Review Queue: http://review.juju.solutions || Unanswered Questions: http://goo.gl/dNj8CP || Youtube: https://www.youtube.com/c/jujucharms || Charmer Summit Sept 12-14th http://summit.juju.solutions || Juju 2.0 beta-16 release notes: https://jujucharms.com/docs/devel/temp-release-notes [17:26] cory_fu sorry i missed one https://github.com/juju-solutions/charmbox/pull/56 [17:26] when i tested i didn't give it a full bundletester test, and it failed miserably. http://ci.containers.juju.solutions:8080/job/Kube-Core-Bundle-Test/12/console [17:26] but that little bump to six seems ot resolve it [17:27] lazyPower_:ok let me try beta16 [17:44] is there a way in actions.yaml to say i want the user to specify one parameter or the other but not both? [17:44] i suppose i could make one the default [17:51] cholcombe: check jsonschema, not sure but probably [17:54] marcoceppi, yeah good point. thanks [17:54] bdx I deployed your vault, consul, and consul-agent charms but the vault charm didn't seem to work with my charm [17:57] lazyPower_:lxc config set core.https_address [::] , with this work around, finished bootstrapping. [17:58] Guest46979 - ah i didnt setup ipv6 networking on my lxd bridge [17:58] makes sense though [17:59] lazyPower_:it was mentioned in one of the mail. I didnt setup ipv6... [17:59] lazyPower_:Thanks [19:27] icey: how so? [19:27] bdx it almost seems to not send keys in response to the token.requested state [19:27] bdx please tell me I'm just using it wrong :-P [19:28] it's my unlock_ceph charm that isn't working [19:28] whaa ... let me stand up a deploy real quick [19:28] want to see? [19:29] bdx ^ [20:12] icey: ya [20:12] icey: deploying now ... did you deploy -> https://gist.github.com/jamesbeedy/102d1bbcfe5dbf6227f014348026d0a3 [20:12] ? [20:12] basically, PM'd you a hangouts link [22:54] Is the juju 2.0 getting started guide up to date? [22:55] Seem to get 'ERROR failed to bootstrap model: cannot start bootstrap instance: unable to get LXD image for ubuntu-xenial: The requested image couldn't be found.' when bootstraping using lxd [23:37] Bro I got that same error [23:37] You can work around it by logging into the container host and copying in that image under that alias [23:38] lxc image copy ubuntu:16.04/amd64 local: --alias ubuntu-xenial