逻辑集群与数据缓存在网络数据库中的应用

Application of logical client clustering and caching technology in network databases

在网络计算环境下，当大量客户同时存取数据库资源，数据库服务器常成为性能瓶颈。将集群技术应用于数据库系统结构，通过将具有相似数据存取模式的客户形成一个逻辑集群，由一中间层服务器负责集群范围内的目录服务和缓存管理，形成客户用／服务器／服务器三层结构。为获得更好的性能，客户机，特别是中间层服务器规定有合适的缓存功能。该结构可有效降低服务器负载，提高数据库事务响应性能，有效扩大系统支持的客户规模，具有一定的应用价值。

In network computing environments, as large number of concurrent user access shared database, the database servers often become performance bottlenecks. This paper applies clustering technique to database system structure and proposes a client/server/server architecture. The group of clients that demonstrate similarities in their data access behavior are clustered into a logical cluster, and each such cluster is handled by an intermediate server that acts as a cluster-wide directory service and cache manager. To improve system performance, client data caching especially, intermediate server caching is planned. This scheme can reduce sever load significantly, improve system capability and transaction response times.