Updated to Terraform Client 0.7 Updated docs

[mc-old]

I encountered some issues while trying to run the demo so I'm sharing what I did to get up and running. PR: Capgemini/tf_tls#46 should be dealt with first and references to the mattcoffey fork in main.tf should be reverted back to Capgemini.

N.B.: Only tested with Digital Ocean

[enxebre]

Why is this here?

[enxebre]

Why is this needed?

[enxebre]

hey @mattcoffey thanks a lot for this, just dropped a few comments

[enxebre]

Could you point the modules to github.com/tamsky/tf_tls/ for consistency which is already being used for aws and supports terraform >= 0.7.3 then we can move all back in one go to Cap

[mc-old]

Thanks for the comments @enxebre The odd stuff in inventory/inventory and comment about local ssh server was related to making the ansible part work without having specified the TF_VAR_STATE_ROOT properly previously. When I realised that was the problem it became obvious that it wasn't necessary to copy the IPs from the output of terraform apply or have a local ssh server running.

[enxebre]

Hey @mattcoffey as part of the upgrade for terraform 0.7 we'd need to update the the templates to be "Data Sources" analogous to https://github.com/Capgemini/kubeform/pull/172/files#diff-f7c82420f5bbe654c1ba49edd173dab3R14
and to be consistent with the list format as per https://github.com/Capgemini/kubeform/pull/172/files#diff-530c171466f13a11a8a7c715564dbc4bR85
I'll wait in case you wanna add support for that to this PR, otherwise I'll merge this anyway as It's an improvement over what we have now.
Thanks!

[mc-old]

I'll update the templates when I get back from travel on the 27th. Thanks for the pointers Alberto!