[00:38] <Azendale> bigjools: so, I was wrong about it being fixed. But it is something weird. Running a packet capture on the host on the bridge interface shows the MaaS controller sending 'who has' arp packets that never get answered. Weirder is rebooting the maas in LXC fixes the problem just like in KVM. I tried removing and reattaching all interfaces from the bridge, bringing the bridge up and down, bringing up and down the tap device for the node maas was trying to reach, et
[00:55] <bigjools> Azendale: curious
[10:50] <rbasak> If I'm running my own DHCP, what are my options for getting enough DNS for juju to work?
[10:50] <rbasak> I don't have functional (enough) DNS on this network already. I do have a static mapping from MAC to IP.
[10:51] <rbasak> I can run a DNS server on the MAAS node, for example, and point all nodes to use that for DNS using DHCP.
[10:51] <rbasak> Is there anything that exists so that I don't have to do this manually?
[10:51] <rbasak> (especially as the hostnames are issued randomly)
[13:14] <Maassive> hello All
[13:14] <Maassive> can someone help me with geeting MAAS to work
[13:14] <Maassive> Ive followed all Documentation
[13:14] <Maassive> but struggling to get the node to be up
[13:15] <Maassive> last error was this: tftp timeout
[13:15] <Maassive> any idea pls
[13:16] <Maassive> I have followed the exact doc
[13:16] <Maassive> now I did whats in here http://fafadiatech.blogspot.com/2012/04/metal-as-service-maas-for-cloud.html
[13:22] <Maassive> also can anyone guide me which maas version is the less buggy one 12.04.3 lts or 12.10 or 13?
[13:22] <Maassive> Im looking for a stable one maas
[13:31] <Maassive> Hello?????
[13:31] <Maassive> is this room active?
[23:00] <wintrmute> anyone know how this message from the ubuntu maas web gui applies to ubuntu distros that are installed on the nodes?  i have no idea what they're packaged with:  If 'local' is chosen, nodes must be using mDNS. Leave empty to use hostnames without a domain for newly enlisted nodes.
[23:00] <bigjools> wintrmute: mDNS is Avahi
[23:01] <bigjools> it should work out the box provided everything is on the same lan segment
[23:01] <wintrmute> alright, ill leave everything default
[23:01] <wintrmute> ty
[23:01] <bigjools> it just saves running DNS
[23:03] <wintrmute> bigjools, so does that mean i leave the cluster interfaces "unmanaged"?  I was planning to set that to "manage DHCP and DNS" with the associated DHCP fields
[23:04] <wintrmute> s/that/eth0/
[23:04] <bigjools> no it needs to be managed
[23:04] <bigjools> you can get away with DHCP only but it won't make any difference if you do DNS as well
[23:04] <wintrmute> gotcha.  and is DNS zone name "master" (default) fine?
[23:04] <bigjools> yep
[23:05] <wintrmute> this is on a local LAN with bridged VMs and an existing DHCP/DNS server, if that matters
[23:06] <bigjools> you can't have an existing DHCP server on the same lan
[23:06] <bigjools> it will conflict with the maas dhcp server
[23:06] <wintrmute> i have it set up to ignore DHCP requests from PXE boots