关键词: 服务恢复, 上下文, 分类, 持久化

Abstract: Service context is the key point for achieving transparent fault tolerance to the user, and reducing the cost during service recovery. However, the current research on service recovery lacks of a classification and representation method about context. Aiming to this problem, this paper presents a method of context classification, representation and persistence in service-oriented recovery. It utilizes the specific application scenario to describe the use of context data. The method provides a reliable basis to build context model for service recovery and the selection of equivalent services.

Key words: service recovery, context, classification, persistence