(*) If the server is part of a virtualisation platform, where the conditions below apply:

• All data are saved on a shared medium, such as e.g. an SAN (Storage Area Network)
• Each physical machine must have sufficient memory to be able to start several virtual servers of a failed node at any time.
• The platform is at least composed of 2 physical servers.

(**) If the server/service is part of a cluster or farm where its services are automatically taken over by an alternative server/service in the event of a failure. Related examples are database clusters, web farms or firewall clusters, where the redundancy mechanism is managed by Combell.

(***) In the case a server/service makes part of an environment where de services are automatically transferred to an alternative server/service, located in another data centre, during an electronic or mechanical failure.