数据网格中的一个多级Cache模型

A Model Providing Multi-Level Cache in Data Grid

数据网格是一个典型的分布式系统,访问其中的广域分布的海量数据需要很大的时间开销。本文介绍一个数据网格系统中实现高速数据统一访问的Cache模型,该模型采用二级Cache机制,使用二个数据缓冲表来快速定位缓冲数据和控制缓冲数据访问,给出了各级Cache的数据替换算法,并提供了灵活的配置方法,可以将Cache与客户机、服务器独立分布,实现了Cache的可扩展性。

Accessing to widely distributed massive data in Data Grid which is a typical distributed system costs a lot of time. This paper introduces a model providing multi-level cache to implement fast data uniform access in Data Grid. It uses two-level cache, which adopts two Cache-Data Tables to locate Cache-Data and control data access, designs replacing data algorithm in its cache, and provides a set of flexible config methods which can separate cache from client or server and implement scalable cache in client or server.