Verifying local HA operation

If you have a four-node MetroCluster configuration, you should verify the operation of the local HA pairs in the MetroCluster configuration. This is not required for two-node configurations.

About this task

Two-node MetroCluster configurations do not consist of local HA pairs and this task does not apply.

The examples in this task use standard naming conventions:

Procedure

  1. On cluster_A, perform a failover and giveback in both directions.
    1. Confirm that storage failover is enabled: storage failover show
      The output should indicate that takeover is possible for both nodes:
      cluster_A::> storage failover show
                                    Takeover
      Node           Partner        Possible State Description
      -------------- -------------- -------- ---------------------------
      controller_A_1 controller_A_2 true     Connected to controller_A_2
      
      controller_A_2 controller_A_1 true     Connected to controller_A_1
      2 entries were displayed.
      
      
    2. Take over controller_A_2 from controller_A_1: storage failover takeover controller_A_2
      You can use the storage failover show-takeover command to monitor the progress of the takeover operation.
    3. Confirm that the takeover is complete: storage failover show
      The output should indicate that controller_A_1 is in takeover state, meaning that it has taken over its HA partner:
      cluster_A::> storage failover show
                                    Takeover
      Node           Partner        Possible State Description
      -------------- -------------- -------- -----------------
      controller_A_1 controller_A_2 false    In takeover
      
      controller_A_2 controller_A_1 -        Unknown
      2 entries were displayed.
      
      
    4. Give back controller_A_2: storage failover giveback controller_A_2
      You can use the storage failover show-giveback command to monitor the progress of the giveback operation.
    5. Confirm that storage failover has returned to a normal state: storage failover show
      The output should indicate that takeover is possible for both nodes:
      cluster_A::> storage failover show
                                    Takeover
      Node           Partner        Possible State Description
      -------------- -------------- -------- ---------------------------
      controller_A_1 controller_A_2 true     Connected to controller_A_2
      
      controller_A_2 controller_A_1 true     Connected to controller_A_1
      2 entries were displayed.
      
      
    6. Repeat the previous substeps, this time taking over controller_A_1 from controller_A_2.
  2. Repeat the preceding steps on cluster_B.