海量存储系统的性能管理与监测方法研究

海量存储系统是一个大规模分布式的集群存储系统,拥有着PB级的存储容量,存在着数以千计的并发访问。为实现对海量存储系统的监控和管理,采用以ganglia为基础并且辅以各种改进和优化的监控模型,以满足海量存储系统的要求。从性能、功能、可靠性以及资源优化等多方面入手,使模型不断完善,最终达到一个比较好的性能,为集群的监控和管理提供一个可靠的保障。

MASS STORAGE SYSTEM PERFORMANCE MANAGEMENT AND MONITORING METHODS RESEARCH

The Mass Storage System is a large-scale distributed clustering storage system.It has a PB-level storage capacity where there are thousands of concurrent accesses.To realize mass storage system monitoring and management,the paper uses a ganglia-based and richly improved and optimized monitoring model to satisfy mass storage system requirements.The paper makes efforts on a number of aspects such as performance,function,reliability and resource optimization etc.to continuously perfect the model for achieving a better performance,so that it can provide a reliable assurance for cluster monitoring and management.