Maintaining a load balancing cluster

If synchronized data fails to be committed to a node in the cluster, remove that node from the cluster for corrective action. Use the load balancing consolecli commands to analyze and update cluster nodes.

The following consolecli.sh|bat commands for maintaining the cluster are available from DASH_HOME/bin:

• Run the ListHAModules command to return a list of the component modules in the cluster:
  • consolecli.sh ListHAModules --username console_admin_user_ID --password console_admin_password [--nodename true|false]
  • consolecli.sh ListHAModules --username console_admin_user_ID --password console_admin_password [--nodename true|false]

  nodename is optional parameter to the ListHAModules command. When set to true, the local component modules are also listed, otherwise only modules from the database are listed.

• Run the ListHANodes command to list the current nodes in the cluster, determine whether they are active or not, view their synchronization status, and their version level of Dashboard Application Services Hub:
  • consolecli.sh ListHANodes --username console_admin_user_ID --password console_admin_password
  • consolecli.sh ListHANodes --username console_admin_user_ID --password console_admin_password
• To force a database update, run the ForceHAUpdate command. The ForceHAUpdate command pushes the local configuration to the database and updates the modules table to match the local node's module versions. Notifications are sent to other nodes to synchronize. Notified nodes with module versions that match those of the originating node are synchronized. Notified nodes with module versions that do not match, go into maintenance mode until an administrator updates their modules accordingly.
  • consolecli.sh ForceHAUpdate --username console_admin_user_ID --password console_admin_password
  • consolecli.bat ForceHAUpdate --username console_admin_user_ID --password console_admin_password
• Run the RemoveHANode command on a node to remove it from the cluster.

  This command is used to permanently remove a node from the cluster before you delete the WebSphere Application Server data source. If the data source was deleted beforehand, this command can be run from another node to remove a separate node by specifying the relevant node name. There is also an optional active parameter that is used for cleanup purposes. Its options are true|false|unreachable. If true is specified, then all the active nodes in the database with status of active that are not reachable are deleted. If false is specified, then all the inactive nodes in the database are deleted. If unreachable is specified, then all the nodes that are unreachable from that node are deleted.

  Note: If the active parameter is specified, then you do not need to use the nodename parameter. If you specify both a nodename parameter and the active parameter then the active parameter is ignored.
  • consolecli.sh RemoveHANode --username console_admin_user_ID --password console_admin_password [--nodename node_name]|[-- active true|false|unreachable]
  • consolecli.sh RemoveHANode --username console_admin_user_ID --password console_admin_password [--nodename node_name]|[--active true|false|unreachable]