2 kinds of scenarios in terms of HA failover behavior on Nexenta

To clarify how rsf-cluster plugin works, I did some test on our office lab.

Below are detailed steps:

Example1:

Background

• 2 nodes( node1 and node2, node1 is primary node) 
• 1 Jbod
• 10G connect to switch respectively
  

1)  pull out 10G net cable of node1 which as heartbeat

      The behavior certainly tiggers automatical failover to node2 (standby node) which is still have connection to switch 

2)  pull out 10G net cable of node2 which as heartbeat

      no Faiover, but node2 is shown as down because of disconnection between node2 and switch.

Example2:

Background

• 2 nodes( node1 and node2, node1 is primary node)
• 1 Jbod
• 1G connect to each other with NIC not configured

3)  pull out 1G net cable of node1which as heartbeat

     no failover

4)  pull out 1G net cable of node2 which as heartbeat

      no failover

5) insert 1G cable back to 2 servers

     nothing change