Recommendations for backend servers

We recommend that you install at least two backend servers to ensure service continuity in case of failure.

Install each server on different machines, virtual or physical.

Enable the Microsoft Network Load Balancing (NLB cluster) mechanism on each machine hosting a backend server, to ensure redundancy and load balancing.

For more information on the NLB mechanism, refer to the section Performing a standard installation.

With redundancy

If one backend server fails, we recommend installing another backend server to maintain redundancy.

To install a new backend server, use the installation center on the desired machine.

Make sure to first add the new machine to the NLB cluster and remove the old one.

No redundancy

If you only had one backend server, you need to restore the machine from a backup (virtual machine snapshot or physical machine disk image).

We recommend that you check the status of the cluster after restoring the machine. The restored machine must be visible in the cluster.

When the machine restarts, the backend server connects to the administration and log databases. Agent handlers and administration consoles normally connect to the backend server.