OSTF try to ssh to the node that was destroyed

VERSION:
  feature_groups:
    - mirantis
  production: "docker"
  release: "6.1"
  openstack_version: "2014.2-6.1"
  api: "1.0"
  build_number: "310"
  build_id: "2015-04-13_22-54-31"
  nailgun_sha: "d22c074dec091e5ddd8ea3003c37665058303cd5"
  python-fuelclient_sha: "9208ff4a08dcb674ce2df132399a5aa3ddfac21c"
  astute_sha: "d96a80b63198a578b2c159edbd76048819039eb0"
  fuellib_sha: "8b80657e9ceed8d59c2dff1c11e1481c7e69380e"
  ostf_sha: "c2a76a60ec4ebbd78e508216c2e12787bf25e423"
  fuelmain_sha: "335d3ed09ed79bd37e1f7a90442c4831c8845582"

Deploy ceph with in HA mode

Scenario:
1. Create cluster with ubuntu
2. Add 3 nodes with controller and ceph OSD roles
3. Add 1 node with ceph OSD roles
4. Add 2 nodes with compute and ceph OSD roles
5. Deploy the cluster
6. Check ceph status
7. Cold retsart
8. Check ceph status

Next tests are failed
AssertionError: Failed tests, fails: 3 should fail: 1 failed tests name: [{u'Check that required services are running (failure)': u'Some nova services have not been started.. Please refer to OpenStack logs for more details.'}, {u'Check network connectivity from instance via floating IP (failure)': u'Instance is not reachable by IP. Please refer to OpenStack logs for more details.'}, {u'Launch instance with file injection (failure)': u'Instance is not reachable by IP. Please refer to OpenStack logs for more details.'}]

In ostf logs we can see that ssh to the node 5 failed(and acually we destroy it), so need to investigate why ostf goes to the offline node