rejoin stack isn't supported

[Shlomo]

The DF controller isn't load after stack rejoin.

Reproduction steps:
- stack
- reboot
- rejoin
- access the df-controller service

------------------------------------------------------------------------------------------------------

result:

2015-10-28 09:46:04.070 ERROR neutron.agent.linux.utils [-]
Command: ['ovs-vsctl', 'set-manager', 'ptcp:6640:10.100.100.45']
Exit code: 1
Stdin:
Stdout:
Stderr: ovs-vsctl: unix:/usr/local/var/run/openvswitch/db.sock: database connection failed (No such file or directory)

2015-10-28 09:46:04.070 CRITICAL neutron [-] RuntimeError:
Command: ['ovs-vsctl', 'set-manager', 'ptcp:6640:10.100.100.45']
Exit code: 1
Stdin:
Stdout:
Stderr: ovs-vsctl: unix:/usr/local/var/run/openvswitch/db.sock: database connection failed (No such file or directory)

2015-10-28 09:46:04.070 TRACE neutron Traceback (most recent call last):
2015-10-28 09:46:04.070 TRACE neutron File "/opt/stack/dragonflow/dragonflow/controller/df_local_controller.py", line 296, in <module>
2015-10-28 09:46:04.070 TRACE neutron main()
2015-10-28 09:46:04.070 TRACE neutron File "/opt/stack/dragonflow/dragonflow/controller/df_local_controller.py", line 293, in main
2015-10-28 09:46:04.070 TRACE neutron controller.run()
2015-10-28 09:46:04.070 TRACE neutron File "/opt/stack/dragonflow/dragonflow/controller/df_local_controller.py", line 68, in run
2015-10-28 09:46:04.070 TRACE neutron self.vswitch_api.initialize()
2015-10-28 09:46:04.070 TRACE neutron File "/opt/stack/dragonflow/dragonflow/db/drivers/ovsdb_vswitch_impl.py", line 42, in initialize
2015-10-28 09:46:04.070 TRACE neutron self.ovsdb.start()
2015-10-28 09:46:04.070 TRACE neutron File "/opt/stack/neutron/neutron/agent/ovsdb/native/connection.py", line 72, in start
2015-10-28 09:46:04.070 TRACE neutron helpers.enable_connection_uri(self.connection)
2015-10-28 09:46:04.070 TRACE neutron File "/opt/stack/neutron/neutron/agent/ovsdb/native/helpers.py", line 29, in enable_connection_uri
2015-10-28 09:46:04.070 TRACE neutron utils.execute(['ovs-vsctl', 'set-manager', manager_uri], run_as_root=True)
2015-10-28 09:46:04.070 TRACE neutron File "/opt/stack/neutron/neutron/agent/linux/utils.py", line 158, in execute
2015-10-28 09:46:04.070 TRACE neutron raise RuntimeError(m)
2015-10-28 09:46:04.070 TRACE neutron RuntimeError:
2015-10-28 09:46:04.070 TRACE neutron Command: ['ovs-vsctl', 'set-manager', 'ptcp:6640:10.100.100.45']
2015-10-28 09:46:04.070 TRACE neutron Exit code: 1
2015-10-28 09:46:04.070 TRACE neutron Stdin:
2015-10-28 09:46:04.070 TRACE neutron Stdout:
2015-10-28 09:46:04.070 TRACE neutron Stderr: ovs-vsctl: unix:/usr/local/var/run/openvswitch/db.sock: database connection failed (No such file or directory)
2015-10-28 09:46:04.070 TRACE neutron
2015-10-28 09:46:04.070 TRACE neutron
2015-10-28 09:46:04.086 INFO oslo_rootwrap.client [-] Stopping rootwrap daemon process with pid=3940

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/245802
Committed: https://git.openstack.org/cgit/openstack/dragonflow/commit/?id=dfe0ee649561449b937b96040ce3ec5a0de99055
Submitter: Jenkins
Branch: master

commit dfe0ee649561449b937b96040ce3ec5a0de99055
Author: Yuli <email address hidden>
Date: Mon Nov 16 16:49:11 2015 +0200

    Start etcd and ovs services when doing stack rejoin

    Closes-Bug: #1510812
    Change-Id: Ic76a70776fcb65db265a2b005ef0faa95d56e8ed