摘要： 在分布式海量存储测试中，测试节点可能会出现资源耗尽甚至宕机的情况，严重影响测试结果的准确性和可信性，另外由于测试平台涉及大量的数据库操作，需要对MySQL数据库的性能有一个准确的评估。针对上述问题，设计并实现一个分布式监控管理系统，该系统可实时监控测试节点及MySQL的性能，并能提供报警机制。对其进行测试验证，结果表明，该系统能合理调整测试任务、分配系统资源，并提高系统整体性能。

关键词: 海量存储系统, 系统测试, 分布式监控, 监控管理, 报警机制

Abstract: In the distributed mass storage test, some test nodes may be the situation of depletion of resources even crash, which seriously affects the results of accuracy and credibility. Meanwhile, since there are a large number of database operations on the test platform, MySQL database’s performance nust be assessed exactly. To solve the above problems, this paper designs and implements a distributed monitoring management system, it can real-time monitor test nodes and MySQL's performance, provide alarm system as well. The distributed monitoring management system is tested, result shows that the system can reasonably adjust test tasks, allocate system resources and improve overall system’s performance.

Key words: mass storage system, system test, distributed monitoring, monitoring management, alarm mechanism

中图分类号: 

• TP391