Windows systems support high availability by means of:
the use of cluster configurations
special disk peripherals such as RAID systems
software that identifies faults and enables you to switch quickly from defective hardware to backup systems
These measures, together with openUTM, can be utilized to their full potential because openUTM offers global transaction management and comprehensive restart capabilities right up to the client. This makes openUTM especially suited for use in clusters because processing can be continued on a different system after a restart without losing any data or allowing inconsistencies to arise.
Windows systems using PRIMERGY hardware can be used as follows as a cluster:
Symmetric cluster with the Microsoft Cluster Server:A Microsoft Cluster Server configuration currently consists of two independent Windows-based servers that are connected via a high-speed connection and that presents itself to the user as a single server system. The clients in the network can access all resources of the cluster transparently. If a server crashes, then the other servers automatically take over its function. Both servers can be run in mission-critical mode, i.e. they work productively on their own applications and also monitor each other.
Asymmetric cluster with the PRIMERGY ServerShield and SCSI switch:The cluster consists of primary and secondary servers. The primary server processes the business critical tasks by default while the secondary server assumes less critical tasks. The secondary server monitors the primary server. If it detects an error in the primary server, then it properly terminates its own applications and assumes the data of the primary server. The new start of the secondary server with the data and applications of the primary server quickly restores the availability of the overall system.