[16:00] <freakynl> Hi, would cloud-init be a good choice to use for configuring an ova/ovf template after deployment? It won't be running on public clouds. The default config seems to want to pull information from the default gw which seems very amazon EC2 like
[16:05] <smoser> freakynl, well,. it will look to the ec2 metadata service currently if it didnt find any ovf data.
[16:06] <smoser> but if you have an ovf iso transport cdrom attached, it should use that.
[16:15] <freakynl> smoser: Thanks for the info. Was more hoping of a yes this is the right tool or no you should look elsewhere kind of answer before diving in the docs :)
[16:16] <smoser> freakynl, yes!
[16:16] <smoser> that better ?
[16:32] <freakynl> smoser: hell yea :) thanks
[19:53] <larsks> harlowja: around?
[19:54] <larsks> smoser: cloudinit/net/sysconfig.py seems to be broken, particularly right here: https://github.com/cloud-init/cloud-init/blob/master/cloudinit/net/sysconfig.py#L286
[19:54] <larsks> At that point, "iface" is just a dict, and has no "children" attribute, so it all blows up.
[19:57] <smoser> :(
[19:57] <larsks> https://bugs.launchpad.net/cloud-init/+bug/1665441
[19:57] <smoser> do you have network config that showed this ?
[19:58] <larsks> I do, although I am pretty sure that any network config would show this...I don't see how you would get anything other than a dict at this point.
[19:58] <larsks> But I will update the bug with my network_data.json.
[19:59] <smoser> must of regressed. i am surprised harlowja woudlnt have cried.
[20:05] <rharper> looks like a nice candidate for integration tests
[20:08] <larsks> rharper: looks like a nice candidate for *unit* tests, even, I think.
[20:08] <rharper> oh, that too
[20:08] <rharper> there are sysconfig unittests
[20:08] <larsks> rharper: I am pretty sure this stems from someone writing "iface" when they meant "iface_cfg".
[20:08] <rharper> so that might be an easy way to expose it
[20:08] <rharper> y
[20:09] <larsks> With a couple small fixes I have it not failing and generating a semi-valid config file, although it appears to be missing an ip address.
[20:13] <larsks> Oh hey, actually I think that fixed it.  It just generates more alias configurations than I expected.
[21:04] <harlowja> who what