Server interface

In Windows, Linux, and UNIX, clustering software is implemented to make Cloverleaf highly available. Two nodes (Cloverleaf machines) are configured: One is designated as the primary; the second is designated as the secondary.

Both nodes have access to a shared disk. The active node has the lock. If the primary fails, then the clustering software takes this action:

  • The active node begins shutting down with these steps:
    • Stops the application by calling the AS stop script.

      The stop action does not apply on a program failure.

    • Unmounts the shared disk at the Cloverleaf installed location.
    • Releases the Virtual IP (VIP) address. This is the Cloverleaf IP to external systems.
    Note: This does not apply to a program failure stop action.
  • The standby node (failing over):
    • Mounts the shared disk at the Cloverleaf installed location.
    • Assigns the VIP.
    • Starts the application by calling the AS start script.

VMWare Clustering on Linux and Windows

VMWare clustering, or virtual environment, monitors the hardware health. If the active server fails, then it relocates the VM’s running on that node to its standby server. On the standby node, AS scripts are implemented as a service that starts Cloverleaf at boot time.

During implementation, a CloverAS service is created and configured as automatic at boot time. This service calls the AS start script.