[7.0][SWARM] deploy_heat_ha: Time limit exceeded while waiting for launching the new instance

[Vladimir Jigulin]

Steps to reproduce:
1. Create cluster
2. Add 3 node with controller role and mongo
3. Add 1 nodes with compute role
4. Set Ceilometer install option
5. Deploy the cluster
6. Verify Heat and Ceilometer services
7. Run OSTF platform tests

Expected results: OSTF pass, logs clear.

Actual result:
OSTF Check stack autoscaling (fuel_health.tests.tests_platform.test_heat.HeatSmokeTests.test_autoscaling) fail:
Test status: failure, message: Time limit exceeded while waiting for launching the new instance per autoscaling alarm to finish. Please refer to OpenStack logs for more details.'

Some error messages from deployment logs:
http://paste.openstack.org/show/I5mAcw0Q9vbN17OxV5T4/
also:
2017-02-26T03:30:35.415972+00:00 debug: (Puppet::Type::Cs_rsc_colocation::ProviderCrm) Command '/usr/sbin/crm configure load update /tmp/puppet_crm_update20170226-7404-eix286' failed, and return RC=1, output:
2017-02-26T03:30:35.416396+00:00 debug: (Puppet::Type::Cs_rsc_colocation::ProviderCrm) ERROR: constraint ntp-with-vrouter-ns references a resource clone_p_ntp which doesn't exist

Reproducibility:
deploy_heat_ha_one_controller_neutron and deploy_heat_ha tests
3 times in row. First fail - 22.02.2017. Last pass - 20.02.2017

[Vladimir Jigulin (vjigulin)]

Test still failing on swarm 05/31/17