Remove a Cluster Node

Note:

Oracle recommends that you perform this task during a planned downtime, such as a maintenance period to avoid system disruption.
  1. Log into the OCSDM GUI on a node belonging to the running cluster.
  2. Select Tools, Health Monitor.
  3. In the Health Monitor Console pane, view the Heartbeat Summary table to know which node is cluster member node that is running the master databases.
  4. Log into the node to be removed from the cluster as the nncentral user.
  5. Change directory to the bin directory.
    For example:
    cd /home/nncentral/AcmePacket/NNC8x/bin
  6. Execute the shutdownnnc.sh script.
    ./shutdownnnc.sh
  7. Uninstall OCSDM.
    ./uninstall.sh
  8. Shut down all remaining nodes starting with the replicas, followed by the master cluster member node.
    ./shutdownnnc.sh
  9. Switch to the root user on the master cluster member node and enter the root user password when you are prompted.
    su root
  10. From the bin directory on the master cluster member node, execute the setup.sh script.
    ./setup.sh
  11. Enter 2 for the Custom option and press Enter.
  12. Enter Yes to continue.
  13. Enter 6 for the Cluster management option and press Enter.
  14. Enter Yes to continue.
  15. Option 1 Configure and manage members in a cluster is selected by default. Press Enter.
  16. Enter Yes to continue.
  17. Enter 2 for the Remove all remote members option that removes all remote cluster member nodes from the cluster configuration and press Enter.
  18. Enter Yes to continue.
  19. Enter 1 for the Proceed with removing all remote members option and press Enter.
  20. Enter Yes to continue.
  21. Enter 3 for the Apply new cluster configuration option and press Enter.
  22. Enter Yes to continue.
  23. Enter 3 for the Quit out of cluster configuration option and press Enter.
  24. Switch to the nncentral user on the master cluster member node.
    su nncentral
  25. Execute the startnnc.sh script on the master cluster member node.
    ./startnnc.sh
  26. Execute the startnnc.sh script on each remaining replica node in the cluster.