Version 2.0: Breakthrough High Availability

Most failures on traditional server clusters consist of a single server failure or communication error, and ScaleOut StateServer, version 1.0, was designed to recover and maintain high availability after this typical class of failures. However, experience with Web and application server farms has shown that while multiple, simultaneous server failures are rare, these farms are regularly subjected to multiple system restarts, network switch outages, and other failure scenarios.

In response to customer feedback, ScaleOut StateServer's recovery capabilities have been significantly enhanced with a powerful, new recovery architecture specifically designed to maintain high availability after multiple, simultaneous failures. Now ScaleOut StateServer can handle almost any combination of failures that a server farm is ever likely to experience.

In addition, version 2.0 includes new global control functions that speed up and simplify management operations. ScaleOut StateServer’s distributed store can be quickly and easily tested to verify that it is active and ready to handle client requests.

Automatic Recovery after Multiple Failures

ScaleOut StateServer now maintains client access to the distributed store even after up to N-1 servers have failed in an N-server store. As with prior versions, ScaleOut StateServer continuously monitors the hosts in the server farm and automatically takes recovery actions in the case failures are detected. However, in version 2.0, ScaleOut StateServer heals the store and continues to handle client requests even after multiple, simultaneous failures. The service reports if it detects the possibility that data was lost due to a multi-server failure which removes all replicas for some objects. ScaleOut StateServer also now handles temporary network switch outages, such as power failures, and restores service when network connectivity is restored. In all of these situations, ScaleOut StateServer protects your data while maintaining scalable throughput using its patent-pending, quorum-based data replication and dynamic load-balancing.

Enhanced Global Management

Version 2.0 adds new control functions that speed up and simplify management operations. You can join all servers to the distributed store or have all servers leave the store with one command. In addition, you can simultaneously restart the StateServer service on all servers to form a new distributed store. The management console now reports when the distributed store is temporarily not ready while healing from server or network outages.

Also new in version 2.0 is a “Test Store” command that verifies client access from all active servers in the distributed store. This command creates, reads, updates, and removes several objects so that all servers are fully exercised. It reports the result of this test within several seconds so that system administrators can immediately verify that ScaleOut StateServer is ready to handle client requests.

Download ScaleOut StateServer evaluation software now!

Review the standard features and benefits of ScaleOut StateServer.