摘要
针对系统资源紧张和带宽条件受限时无法区分系统服务重要性等级等问题,设计了一种基于服务状态动态感知的服务重要性分级方法。首先,利用基于应用性能管理(APM)链路追踪的原理实现服务状态监控;然后,通过服务状态监控数据生成服务调用关系拓扑图;最后,利用基于服务拓扑图的重要节点挖掘技术对服务进行动态划分等级,并将低等级服务降级处理,为服务资源动态重新分配提供可靠依据。
Due to the system resource constraints and the limited bandwidth,it is impossible to distin⁃guish the importance level of system services and realize the reallocation of system resources according to the service level.Aiming at this problem,a service importance classification method based on ser⁃vice state perception is designed.Firstly,service status monitoring is realized by using the principle of application performance measurement(APM)link tracking.Then,the service invocation relationship topology is generated from service state monitoring data.Finally,by using the important node mining technology based on the directed topology,the service is used to dynamically classify services and downgrade the low level services,and provides reliable basis for dynamic reallocation of service re⁃sources.
作者
毛宇
贲婷婷
汪敏
李荣宽
MAO Yu;BEN Tingting;WANG Min;LI Rongkuan(The 28th Research Institute of China Electronics Technology Group Corporation,Nanjing 210007,China)
出处
《指挥信息系统与技术》
2020年第3期82-86,共5页
Command Information System and Technology
基金
装备发展部“十三五”预研课题资助项目。
关键词
重要性分级
服务状态监控
拓扑图
重要节点挖掘技术
service importance classification
service status monitoring
topological graph
important node mining technology
