ha_corosync_stability_check failed on checking resourse status

[Egor Kotko]

After reboot ('service corosync start && service pacemaker restart') on the first node
status pacemacker was not changed on online during 20 seconds

https://product-ci.infra.mirantis.net/job/8.0.system_test.ubuntu.ha_neutron_destructive/11/testReport/%28root%29/ha_corosync_stability_check/ha_corosync_stability_check/

Scenario:
            1. On the first controller kill corosync
            2. Verify for all controllers "pcs status nodes" reports
            3. Start corosync on the first controller
            4. Repeat steps 1-3 500 times

Syslog contains:
http://paste.openstack.org/show/475431/

VERSION:
  feature_groups:
    - mirantis
  production: "docker"
  release: "8.0"
  openstack_version: "2015.1.0-8.0"
  api: "1.0"
  build_number: "122"
  build_id: "122"
  fuel-nailgun_sha: "bebc5e74ba24d4c36c839f4c2a18e05b7bdf87e4"
  python-fuelclient_sha: "56fbd6bad7f60f0944b3845c2db14d0b8cabd4d3"
  fuel-agent_sha: "e881f0dabd09af4be4f3e22768b02fe76278e20e"
  fuel-nailgun-agent_sha: "d66f188a1832a9c23b04884a14ef00fc5605ec6d"
  astute_sha: "0f753467a3f16e4d46e7e9f1979905fb178e4d5b"
  fuel-library_sha: "37e1a0ee9f4a8fe026c9c903657d9541bfa77964"
  fuel-ostf_sha: "41aa5059243cbb25d7a80b97f8e1060a502b99dd"
  fuelmain_sha: "3b2b9dc84289227edc6d7ab945442c993935e016"

[Matthew Mosesohn (raytrac3r)]

Only 20s downtime? This is not a high priority bug if it doesn't break deployment.

[Bogdan Dobrelya (bogdando)]

According to logs, the cluster state became broken:
./remote/node-2.test.domain.local/pengine.log:
2015-10-05T23:53:31.401950+00:00 warning: warning: stage6: YOUR RESOURCES ARE NOW LIKELY COMPROMISED
2015-10-05T23:53:31.401950+00:00 err: error: stage6: ENABLE STONITH TO KEEP YOUR RESOURCES SAFE

this state can only be fixed by STONITH enabled. Cannot be fixed in Fuel.