Description of problem:

When using the default harness for jobs running in a container (on Atomic for e.g.), the harness repo should not be checked and the job not aborted.

The check doesn't really make sense since we currently do the check for the harness repo corresponding to the base distro which is most likely different from the distro running in the container.

Version-Release number of selected component (if applicable):

0.18.4

How reproducible:

Always

Steps to Reproduce:
1. Provision a beaker job on an Atomic host with *no* harness ksmeta
2. 
3.

Actual results:

Job aborted

Expected results:


Additional info:


A simple workaround is to specify harness="beah" or harness="restraint" in the ksmeta.