[09:53] <SingleOne> Hello all
[09:54] <SingleOne> Do you know where I can find some documentation / tutorial on how to build a Windows image suitable for MAAS ?
[14:15] <thebozz> Hey guys, some nodes in our cluster are failing deployment with " grub-install: error: /usr/lib/grub/modinfo.sh doesn't exist. Please specify --target or --directory.". What could be causing this? I read that disabling UEFI would fix it, but it doesn't seem to be the case
[14:24] <thebozz> ...or it does work. The node I disabled UEFI on did boot correctly.
[14:34] <roadmr> thebozz: see this bug report: https://bugs.launchpad.net/curtin/+bug/1383727
[14:34] <roadmr> thebozz: looks like upgrading curtin may fix it, just ensure you get the version mentioned in the last comment (or something newer)
[14:48] <thebozz> roadmr: thanks, I'll talk about it with the MAAS admin when he arrives.
[14:49] <roadmr> thebozz: let us know if it works :) I no longer have access to a UEFI system so I haven't been able to confirm that
[14:50] <thebozz> At least disabling UEFI works :) which is enough for now
[18:39] <thebozz> Hey guys, we're trying to install Juju + Openstack on top of MAAS. The auto-install fails with this error log: http://pastebin.com/EWAwCEuN ... what could be wrong? We can SSH into our nodes.
[18:40] <blake_r> thebozz: are you able to ssh using the hostname?
[18:40] <blake_r> thebozz: before the timeout occurred?
[18:48] <thebozz> You were right, we couldn't SSH using the hostnames. We edited resolv.conf and now that works fine, we're going to try again.
[19:04] <thebozz> Nope, same error:  http://pastebin.com/EWAwCEuN
[19:26] <blake_r> thebozz: so the system now resolves the ip address?
[19:26] <blake_r> thebozz: is that the correct ip address in the log?
[19:26] <blake_r> thebozz: for the node
[19:28] <thebozz> blake_r: yeah, that was the correct IP.
[19:28] <blake_r> thebozz: are you able to ssh using the ip address and hostname?
[19:31] <thebozz> Yeah, it works with both. However, while doing the install we're not using the ubuntu default user on MAAS, and I have to connect using `ssh ubunthu@[server]`. Could that have something to do with it?
[19:40] <blake_r> thebozz: the hsot only has the user ubuntu
[19:40] <blake_r> thebozz: you need to use that username
[19:42] <thebozz> Wait... now that I think about it, that shouldn't affect the installation. I mean, the command is `sudo openstack-install`, so that should use the root user. Is there any way to know what user will the installer try to connect to the node as?
[19:43] <blake_r> thebozz: openstack-install uses juju and juju knows to use the ubuntu user
[19:45] <thebozz> :/ then I'm back to square one. I have no idea what could be causing the problem now, especially since we're able to SSH into the nodes.
[19:47] <blake_r> thebozz: can you login to the juju bootstrap node and see if it can resolve the other nodes hostnames?
[20:03] <thebozz> Dunno if we actually have juju already installed - I'm only helping with the debug with barely any knowledge about this. However, my boss says he's been following this tutorial: http://www.ubuntu.com/download/cloud/install-ubuntu-openstack he's currently on step 4.
[21:13] <thebozz> Hey guys, we're trying to install OpenStack on top of MAAS using the ubuntu tutorial (http://www.ubuntu.com/download/cloud/install-ubuntu-openstack).  We're currently on step 4, but it keeps failing with this error: http://pastebin.com/EWAwCEuN
[22:40] <blake_r> thebozz: does the node get marked deployed before the 15 minute timeout