[00:50] hi [00:50] i'm trying to follow https://maas.ubuntu.com/docs/quantal/juju-quick-start.html [00:50] and having a few problems [02:41] mwhudson: what are the issues you're having? [02:41] marcoceppi: ah, hi [02:41] marcoceppi: i overcame my initial issues but now i get [02:41] root@calxeda02-11-00:~# juju bootstrap [02:41] 2013-04-19 21:41:45,622 INFO Bootstrapping environment 'maas' (origin: distro type: maas)... [02:41] 2013-04-19 21:41:47,124 ERROR No matching node is available. [02:42] mwhudson: Just so you know, I glanced over those docs and they're pretty outdated [02:42] marcoceppi: i sort of suspected as much [02:43] i'm wondering if this is some kind of series (precise vs quantal) thing [02:43] but it's all pretty opaque to me :) [02:43] Did you install Juju from the ppa or from the distro? [02:44] ppa [02:45] excellent, do you have nodes in maas that are 'ready' (or whatever the state is that maas displays)? [02:45] yeah [02:46] and clicking 'start node' on one of them worked [02:46] (eventually :-) [02:47] mwhudson: cool, let me poor over my notes from when I tried this about a year ago [02:47] i've got as far as seeing "MAASAPINotFound: File not found" and "NodesNotAvailable: No matching node is available." in the maas.log file when i try to bootstrap [02:47] but i don't know what arguments are being passed... [02:47] was poking around the juju source [02:48] marcoceppi: fwiw, this is an armhf setup [02:49] heh, so all bets are off ;) [02:49] i couldn't possibly comment! [02:49] When I tested this, I used virtual box to simulate maas nodes, so what I got working probably isn't a standard setup [02:50] mwhudson: A couple of things, are the nodes assigned to the same user whos oauth key you're using? [02:50] marcoceppi: the nodes are not assigned at all yet [02:50] there is only one user set up in maas yet [02:51] mwhudson: I think that user needs to "own" them to use them, which might be why you're getting the no matching node available [02:51] oh ok [02:51] I found this: http://askubuntu.com/questions/172011/no-matching-node-is-available-error-when-trying-to-bootstrap-juju [02:51] maas-cli default node acquire node-.* sort of thing? [02:52] well [02:52] commissioning is complete [02:52] mwhudson: I wouldn't be able to say [02:52] hm [02:52] the api method that bootstrapping is calling (and failing on) is acquire [02:53] ah, so it acquires them for the user [02:53] yeah [02:53] All the questions on Ask Ubuntu about that error say that means there are no machines "ready" but as you've said that's not the case [02:53] right [02:53] i'm about >< this far from editing maas source to add some logging methods :) [03:04] ah hahahaha [03:04] POST:, [03:04] i think i can see the problem here [03:05] OH, right it defaults to amd64 [03:06] For all cloud providers [03:06] You should be able to change that with --constraints "arch=armhf" in the bootstrap and subsequent commands [03:07] mwhudson: https://juju.ubuntu.com/docs/constraints.html [03:07] i can't put that into environments.yaml? [03:09] hm [03:09] progress [03:10] LookupError: SSH authorized/public key not found. [03:10] oh [03:10] right [03:10] the key isn't on this node... [03:10] mwhudson: unfortuntately no, you'll have to specify it for each bootstrap, and then either keep specifying it for each deploy or use set-constraints to change the default [03:15] ok [03:19] well, i presume *something* is happening :) [03:20] mwhudson: if bootstrap exited cleanly, then there's a good chance something _is_ happening after a few mins you'll find a node somewhere has turned on and is ready to roll [03:22] yeah [03:23] ah yes, life on the serial console of the allocated node [03:23] huzzah [03:23] i guess i'll be wanting to set up a local archive mirror if i'm going to do this a lot... [03:23] mwhudson: probably a good idea [03:24] maas esp can be pretty package hungry [03:30] although configuring packages seems to be taking the vast bulk of the time [03:30] i wonder if anything came of that "dd an image onto the disk" idea... [03:46] frigging ssh host key checking [03:49] marcoceppi: seems that the arch constraint was automatically propagated into the environment [03:50] mwhudson: they may have changed the default behavior of bootstrapping with constraints. It does make sense to make that the default [03:51] right, two more nodes pending [03:51] time to go home [03:51] marcoceppi: thanks for your help! [03:51] mwhudson: o/ no problem! === defunctzombie is now known as defunctzombie_zz === balloons_ is now known as balloons === balloons is now known as Guest92449 === kandinsk1 is now known as kandinski === popey_ is now known as popey === defunctzombie_zz is now known as defunctzombie === Camel` is now known as Guest71897 === Guest71897 is now known as Camel` === defunctzombie is now known as defunctzombie_zz === defunctzombie_zz is now known as defunctzombie === defunctzombie is now known as defunctzombie_zz === huats_ is now known as huats