Add SSH keys to Vagrant environment

[mbarnes]

Allows Ansible to talk to the cluster nodes in the Vagrant environment, so now the host creation (a.k.a. @slow) tests that sit and wait for bootstrapping to finish actually work.

(Reviewer should verify that with behave -D use-vagrant)

This should also allow us to write tests to actually exercise cluster upgrade, reboot, etc. and catch failed stunts like my threadpool attempt awhile back.

[mbarnes]

(2) [multi-host-mgr] Install SSH keys in Vagrant cluster nodes

[ashcrow]

Reviewing now

[ashcrow]

@mbarnes The actual vagrant work is working great and I'm good with merging.

The change does point out an issue in one of the steps though. I see failures for we have a host at {host} in this block.

It also is showing a non fatal error though I'm not sure where this is being triggered:

Exception in thread Thread-1:
Traceback (most recent call last):
  File "/usr/lib64/python2.7/threading.py", line 804, in __bootstrap_inner
    self.run()
  File "/usr/lib64/python2.7/threading.py", line 757, in run
    self.__target(*self.__args, **self.__kwargs)
  File "[...]/commissaire/src/commissaire/cherrypy_plugins/investigator.py", line 69, in __response_thread
    host, exception = response
TypeError: 'object' object is not iterable

Since these issues are technical unrelated to this PR should make issues for these items?

[mbarnes]

That might be an uncaught exception in the investigator process.

Unrelated to this PR, but I'll investigate further.

[ashcrow]

@mbarnes ok I'll hold off from making either item github issues. Let me know if you decide to split it out or have a fix for it.

[mbarnes]

The change does point out an issue in one of the steps though. I see failures for we have a host at {host} in this block.

I think I overlooked this part, or thought it was the same issue. What does the failure look like for you? Is it causing the test to fail?

[ashcrow]

@mbarnes Correct. It causes it to fail. It gets a 404 instead of a 200.

[ashcrow]

@mbarnes looks like the test failure was on me. I didn't re-provision the node.

[ashcrow]

LGTM!