Restart sequence for OPDK components one by one like a rolling restart

[#1]

Server restart is required for all components one by one to back up data. When all components are running, what is the best sequence to perform a rolling restart one by one?

--SF924745--

[#2]

@Janice Hunt Any Input from your side?

[remeeshnair]

As per OPDK document below is the sequence, Hope this helps.

Stop order: If you install Edge on multiple nodes, then you should stop Edge components on those nodes in the following order: 1. Apigee UI 2. Management Server 3. Router 4. Message Processor 5. Qpid Server 6. Postgres Server 7. OpenLDAP/ApacheDS 8. Qpidd 9. PostgreSQL 10. Cassandra 11. ZooKeeper

Start order: If you install Edge on multiple nodes, then you should start Edge components on those nodes in the following order: 1. ZooKeeper 2. Cassandra 3. OpenLDAP/ApacheDS 4. Qpidd 5. PostgresSQL 6. Management Server 7.Router 8. Message Processor 9. Qpid Server 10. Postgres Server 11. Apigee UI

[#4]

@Reemesh, Does the same order apply for a rolling restart, meaning stop and restart 1. Apigee UI, then stop and restart 2. Management Server and so on?

[remeeshnair]

From my experience each component is independent, we do the restart depends on our need.

[#6]

@Arjav Goswami - I was wondering how you ended up addressing this issue. We are seeing a similar need in some maintenance tasks we're working on. We wanted a script to cycle through the nodes and restart one by one.

[#7]

Hi Steven, In a clustered installation with more than 5 nodes, any one node in the cluster can be down at a time, and impact to runtime traffic can be avoided.