Considered the deficiencies of the existing performance management system, the paper proposes the performance management method based on Graph Database, using Graph search and sequence mining algorithms to improve the accuracy and efifciency of the bottleneck analysis, which is different from the traditional method based on rules and relational table queries. Finally, this method is applied in the actual performance management system to achieve secure management, convenient interactive and easy to deploy in the large-scale clusters.%文章针对现有性能管理应用的不足,提出了一种有别于传统基于规则和关系表查询的基于Graph Database的性能管理方法,利用Graph Search及序列挖掘算法对性能瓶颈进行分析,提高了性能分析的准确度和效率。最后,在实际性能管理系统中应用了此方案设计,实现了管理安全,交互便捷,方便集群大规模部署。
展开▼