Troubleshoot Cassandra for OpenContrail 3.2¶

In case of issues with the OpenContrail 3.2 Cassandra database connection, the system response of the contrail-status command can be as follows:

== Contrail Analytics ==
supervisor-analytics:         active
contrail-alarm-gen            active
contrail-analytics-api        active
contrail-analytics-nodemgr    active
contrail-collector            initializing (Database:ctl02:Global connection down)
contrail-query-engine         timeout
contrail-snmp-collector       active
contrail-topology             active

Workaround:

1. Restart the supervisor-analytics service:

   service supervisor-analytics restart
   

2. Verify the status of the Neutron API:

   neutron net-list
   +------------------------+-------------------------+---------+
   | id                     | name                    | subnets |
   +------------------------+-------------------------+---------+
   | d6638b91-4e1d-4214-... | ip-fabric               |         |
   | ce66ee12-71b4-44ea-... | __link_local__          |         |
   | d452af3a-3b9f-442e-... | default-virtual-network |         |
   +------------------------+-------------------------+---------+
   

3. Restart nova-api to reflect installed OpenContrail dependencies:

   salt 'ctl*' service.restart nova-api
   ctl02.workshop.cloudlab.cz:
       True
   ctl03.workshop.cloudlab.cz:
       True
   ctl01.workshop.cloudlab.cz:
       True
   

4. Verify the system response of the nova list command:

   nova list
   

Now, the OpenStack and OpenContrail controller nodes are properly deployed.