test_list_servers_negative should dump server list when it times out

test_list_servers_negative requires that there is a clean database before it's tests run but if there are instances in an error state that aren't going to delete from the database correctly, the test always times out. The trace is this:

  File "/usr/lib/python2.6/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/usr/lib/python2.6/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/usr/lib/python2.6/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/usr/lib/python2.6/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/tmp/tempest/tempest/tempest/api/compute/servers/test_list_servers_negative.py", line 79, in setUpClass
    cls.os.tenant_name)
  File "/tmp/tempest/tempest/tempest/api/compute/servers/test_list_servers_negative.py", line 43, in _ensure_no_servers
    ignore_error=True)
  File "/tmp/tempest/tempest/tempest/services/compute/json/servers_client.py", line 172, in wait_for_server_termination
    raise exceptions.TimeoutException
Request timed out

The problem is in a CI environment, it's hard to easily tell which servers (and the tests that spawned them) are not getting automatically cleaned up. Most, if not all, instances should be named with the test class that's spawning them.

When test_list_servers_negative dies, it should dump the current nova instance list along with the instance's name, vm_state and task_state.