Description of problem:

OSPd sets throughput-performance as default tuned profile for all server roles

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

python-rdomanager-oscplugin-0.0.8-44.el7ost.noarch

How reproducible:


Steps to Reproduce:
1. Deploy undercloud
2. Deploy overcloud with control, compute, ceph server roles
3. run tuned-adm list on all servers

Actual results:
throughput-performance role for all servers

Expected results:

There is conflicting information in performance docs.

https://bugzilla.redhat.com/show_bug.cgi?id=974688 requests virtual-host as default tuned profile for compute nodes.

https://access.redhat.com/articles/1321163 -- recommends virtual-host profile for Ceph OSD nodes

https://access.redhat.com/articles/1507893 -- recommends throughput-performance for Ceph OSD nodes but virtual-host for Compute nodes

Additional info: I believe virtual-host should be default for compute nodes and OSP 7 + RHCS 1.3 should be regression tested to determine whether virtual-host or throughput-performance is appropriate for ceph nodes.