[05:07] <dpb1> lazyPower: It's a general juju problem, afaict.  I filed a bug: https://bugs.launchpad.net/juju-core/+bug/1328269
[07:20] <bigjools> any takers? https://code.launchpad.net/~julian-edwards/maas/static-range-in-ui-api/+merge/222590
[07:27] <rvba> bigjools: on it
[07:30] <bigjools> ta
[09:55] <bigjools> rvba: solved my chained task queueing
[09:55] <bigjools> it's rather easy
[09:55] <rvba> bigjools: \o/
[10:05] <bigjools> rvba: create a subtask/signature with mytask.s(args, kwargs)
[10:05] <bigjools> then call .set(queue=foo) on it
[10:05] <bigjools> :)
[10:05] <bigjools> then you can throw them in a chained task, which itself doesn't require any queue applied, you just call its apply_async()
[10:05] <rvba> bigjools: don't you have a special keyword argument to pass that directly when creating the task?
[10:05] <bigjools> no
[10:06] <bigjools> because I am passing the task's args
[10:06] <bigjools> I am instantiating the task, just not executing it
[10:06] <bigjools> rvba: docs.celeryproject.org/en/latest/userguide/canvas.html#signatures
[10:07] <bigjools> there is a way of doing it if you use the string name for the task, but that's rather ugly IMO
[10:08] <rvba> Right, using '.set(queue=foo)' is fine anyway.
[12:53] <gmb> rvba: If you can re-review https://code.launchpad.net/~gmb/maas/fix-commissioning-page-distro-list-bug-1312844/+merge/222010 that’d be great. Thanks for your help this morning, too :)
[12:57] <rvba> gmb: I'll have a look now
[13:09] <gmb> rvba: Ah, damn, I’d forgotten the getZZZ() methods. Will fix that. Thanks.
[13:12] <rvba> np
[13:16] <jtv> "Get-some-sleep methods"?
[13:52] <rvba> jtv: allenap: any idea what might be wrong? paste.ubuntu.com/7623634/
[13:52] <rvba> RabbitMQ didn't got installed properly.
[13:52] <jtv> "Yes you're using the mouse buttons to copy/paste from chrome"
[13:52] <rvba> heh
[13:52] <jtv> Hmmm...  dns delay?
[13:53] <rvba> I think it's rabbitMQ not dealing well with the hostname changing or something…
[13:53] <jtv> It's been ages since I heard "DNS can take 15 minutes to propagate" so I have no idea what the status is today, but maybe we're relying on synchronicity when all we have is instantaneity?
[13:54]  * rvba tries to remove and re-install it.
[13:54] <allenap> rvba: It’s not a perfect fit, but http://stackoverflow.com/questions/14659335/rabbitmq-server-fails-to-start-after-hostname-has-changed-since-first-starting has a suggestion.
[13:54] <rvba> Didn't help (purging then restarting rabbitMQ)
[13:55] <jtv> Different error though, no?
[13:55] <rvba> No, same error.
[13:55] <allenap> rvba: Did purging actually remove /var/lib/rabbitmq?
[13:55] <rvba> I used `apt-get purge`.
[14:21] <jtv> rvba, there's another workaround for a similar problem on: http://serverfault.com/questions/225795/error-when-installing-rabbitmq-server-on-ubuntu-10-10
[14:21]  * jtv reboots
[14:40] <rkdemon> HI: I was trying to install maas on another server and I see this message "Import of boot images started on all cluster controllers. Importing the boot images can take a long time depending on the available bandwidth."
[14:41] <rkdemon> Is there an indication anything is happening coz I don't see the boot images ever getting to the cluster
[14:41] <rkdemon> Without this I cannot add a node to the cluster even though I plan to add on VMs to the cluster ?
[14:41] <rkdemon> Any clues anyone ?
[14:44] <rkdemon> When a VM is added as a node to a cluster: the "status" on the node says "commissioning", what does this really mean ?
[15:05] <jtv> rkdemon: to see what's going on with the downloads, have a look in /var/log/maas/celery.log... look for "import."
[15:06] <jtv> A status of "commissioning" means that the node has been told to boot into a temporary environment (not installed on disk) for basic self-check and network discovery.
[15:06] <jtv> Eventually it should contact the server to say that it's done commissioning (whether successfully or not) and its state should become Ready (or Failed, as the case may be).
[16:33] <rkdemon> jtv: thanks
[20:58] <d_`> I have a MAAS region with 2 cluster controllers in it, one is on the same node as the region controller, one is on its own node. all the machines that register against the shared role node work fine, all the nodes that are registering against the stand alone cluster controller are stuck in the declaration process booting the discovery image where it prints the route table out