Attempt to provision a system which causes an error similar to that seen in the results below. You may be able to do this by running product on a system which has virt flags disabled.

• Error like the one below does not bubble up to UI.
  Rolling back due to a problem: [Settings up compute instance whynow.example.org 1 failed [#<Host::Managed id: nil, name: "whynow.example.org", ip: "192.168.100.13", environment: nil, last_compile: nil, last_freshcheck: nil, last_report: nil, updated_at: nil, source_file_id: nil, created_at: nil, mac: nil, root_pass: nil, serial: nil, puppet_status: 0, domain_id: 1, architecture_id: 1, operatingsystem_id: 1, environment_id: 9, subnet_id: 1, ptable_id: 1, medium_id: 5, build: true, comment: "", disk: "", installed_at: nil, model_id: nil, hostgroup_id: 1, owner_id: 1, owner_type: "User", enabled: true, puppet_ca_proxy_id: nil, managed: true, use_image: nil, image_file: nil, uuid: nil, compute_resource_id: 1, puppet_proxy_id: 1, certname: nil, image_id: nil, organization_id: 1, location_id: nil, type: "Host::Managed">, :setCompute]]

• Page is reloaded and user is blindly taken back to virt machine editing page.

• Virt machine is left half-baked, and while no new hosts show up in Hosts view, machines can no longer be created with the name(s) used in other futile attempts. For example, per the logs above, user can no longer try and create a system called "whynow" -- presumably because a basic storage device is listed in libvirt's storage directory -- but no such name appears in UI.

(https://bugzilla.redhat.com/show_bug.cgi?id=973814)