Shared state management for time-sensitive distributed applications

Distributed applications that share dynamically changing state are increasingly being deployed in wide-area environments. Such applications must access the state in a consistent manner, but the consistency requirements vary significantly from other systems. For example, shared memory models such as sequential consistency focus on the ordering of operations and same level of consistency is provided to each process. In interactive distributed applications, the timeliness of updates becoming effective could be an extremely important consistency requirement and it could be different across different users. mechanisms and policies for its management, it, is possible support the needs of such interactive distributed application in a wide area environment. For or attributes for protocols such as the time-sensitivity for their operations applications for achieving better performance under resource constraints, enabled in an object system that we have developed. The system provides both non-timed and time sensitive read and write operations for dynamic shared state. For example, a timed read can be used by a process to read a recently written value whereas a timed write can make a new value available to all readers within a certain amount of time. We have developed a new consistency model called the Timed Consistency (TS), that precisely defines the semantics of timed and non-timed read and write operations. Protocols providing the guarantees of the TS consistency as well as other consistency models applicable to interactive applications have been implemented in our object caching framework. The framework can be configured for a particular client through its QoS interface. The clients only need to specify their consistency provided by the system.