Devops slaves with nodegroup 2 are not discovered.

[Alexandr Kostrikov]

During creation of deployment devops slaves in nodegroup 2 are not discovered.
Need further investigation.

<<<<<##############################[ prepare_slaves_5 ]##############################>>>>>
Bootstrap 5 slave nodes

        Scenario:
            1. Revert snapshot "ready"
            2. Start 5 slave nodes

        Snapshot: ready_with_5_slaves

2015-09-03 09:25:11,558 - INFO environment.py:280 -- We have snapshot with such name: ready
2015-09-03 09:25:11,558 - INFO environment.py:282 -- Reverting the snapshot 'ready' ....
2015-09-03 09:25:22,325 - INFO environment.py:285 -- Resuming the snapshot 'ready' ....
2015-09-03 09:25:22,687 - INFO environment.py:323 -- Admin node login name: 'root' , password: 'r00tme'
2015-09-03 09:26:04,482 - INFO environment.py:108 -- Bootstrapping node: slave-01
2015-09-03 09:26:08,072 - INFO environment.py:108 -- Bootstrapping node: slave-02
2015-09-03 09:26:11,448 - INFO environment.py:108 -- Bootstrapping node: slave-03
2015-09-03 09:26:14,851 - INFO environment.py:108 -- Bootstrapping node: slave-04
2015-09-03 09:26:18,139 - INFO environment.py:108 -- Bootstrapping node: slave-05
2015-09-03 09:36:22,270 - INFO fuel_web_client.py:1177 -- Wait for task 600 seconds: {
 "status": "running",
 "name": "dump",
 "cluster": null,
 "result": {},
 "progress": 0,
 "message": null,
 "id": 1,
 "uuid": "4d242940-f103-40c5-b77b-e57ca54aa336"
}
2015-09-03 09:36:52,398 - INFO fuel_web_client.py:1194 -- Task finished. Took 30 seconds. {
 "status": "ready",
 "name": "dump",
 "cluster": null,
 "result": {},
 "progress": 100,
 "message": "/api/dump/fuel-snapshot-2015-09-03_09-36-30.tar.xz",
 "id": 1,
 "uuid": "4d242940-f103-40c5-b77b-e57ca54aa336"
}
2015-09-03 09:36:52,406 - INFO decorators.py:46 -- Saving logs to "/home/akostrikov/workspace/Stackforge/fuel-qa/logs/fail_error_prepare_slaves_5-fuel-snapshot-2015-09-03_09-36-30.tar.xz" file
2015-09-03 09:38:40,862 - INFO decorators.py:295 -- <<<<<****************************************************************************************************>>>>>
2015-09-03 09:38:40,863 - INFO decorators.py:296 -- Failed in method 'prepare_slaves_5'. Make snapshot: error_prepare_slaves_5
2015-09-03 09:38:40,863 - INFO decorators.py:306 -- You could revert and ssh to master node: [source /home/akostrikov/workspace/Stackforge/fuel-devops-venv/bin/activate; dos.py revert-resume flexible_templating_old_test23 --snapshot-name error_prepare_slaves_5 && ssh root@10.109.0.2]

[Alexandr Kostrikov (akostrikov-mirantis)]

Information from node in nodegroup 2
http://paste.openstack.org/show/444084/

[Alexandr Kostrikov (akostrikov-mirantis)]

There is a fix after which node is discovered
http://paste.openstack.org/show/444108/

[Alexandr Kostrikov (akostrikov-mirantis)]

Bug is going to be fixed with documentation/code in fuel-qa

[Alexandr Kostrikov (akostrikov-mirantis)]

We should add check to verify that sysctl contains values
net.bridge.bridge-nf-call-ip6tables = 0
net.bridge.bridge-nf-call-iptables = 0
net.bridge.bridge-nf-call-arptables = 0

Check should be performed with MULTIPLE_NETWORKS feature.
The point is where to set that check - in environment creation or before each test.
Seems that only after environment creation is not enough because after restart settings may not apply and user still have environment without
net.bridge.bridge-nf-call-ip6tables = 0

[Alexandr Kostrikov (akostrikov-mirantis)]

We have decided with Dennis Dmitriev that check should be performed before creation of virtual machines to not to overload host.

[Nastya Urlapova (aurlapova)]

Moved to 9.0, due to Medium status.