摘要
日趋庞大、复杂、高速的网络使得传统网络测量技术已经不能满足当下网络智慧管控的需求。作为一种新型网络测量技术,网络遥测可以提供细粒度、精确的会话级或报文级遥测信息。目前已有的网络遥测方案在部署网络遥测路径时不考虑网络状态,大多以静态的方式部署网络遥测路径。这种方法无法适应网络的动态性及不可靠性,如果网络遥测包所经路径出现带宽饱和或者遭遇网络攻击,则会造成网络遥测包丢失,使得网络遥测可靠性变差。除此之外,现有网络遥测方案通常采用全链路覆盖方式实现,遥测冗余较大,探针数据包的有效载荷较低。为了解决上述问题,文中提出了基于协同迁移进化的自适应网络遥测路径编排方法(AdaPtive network telemetry Path OrchestratIoN method based on CooperaTivE MigRation Evolution, APPOINTER)。APPOINTER根据网络状态信息,计算能够覆盖全部网络设备的最优网络遥测路径,以转发遥测报文。实验结果表明,APPOINTER增强了网络遥测的可靠性,有效避免了遥测冗余,提高了遥测效率。
The increasingly large,complex and high-speed network makes the traditional network measurement technology cannot meet the current demand of intelligent network control.As a new network measurement technology,network telemetry can provide fine-grained and accurate session-level or message-level telemetry information.The existing network telemetry solutions do not consider the network state when deploying the network telemetry path,and deploy the network telemetry path in a static manner.These approaches cannot adapt to the dynamic and unreliable nature of the network.If the routing path that transfers the network telemetry packets is facing with network attacks or saturated,the network telemetry packets will be lost,and the reliabi-lity of network telemetry will decreased.In addition,the existing network telemetry methods are usually implemented by traversing all links,causing large telemetry redundancy and relatively low probe packet payload.To solve the above problems,this paper proposes APPOINTER,an adaptive network telemetry path scheduling method based on cooperative migration evolution.APPOINTER calculates the optimal network telemetry path that can traverses all network devices to forward telemetry messages based on network state information.Experimental results show that APPOINTER enhances the reliability of network telemetry,effectively avoids telemetry redundancy,and improves telemetry efficiency.
作者
郝炳炜
崔允贺
钱清
申国伟
郭春
HAO Bingwei;CUI Yunhe;QIAN Qing;SHEN Guowei;GUO Chun(School of Computer Science and Technology,Guizhou University,Guiyang 550025,China;State Key Laboratory of Public Big Data(Guizhou University),Guiyang 550025,China;School of Information,Guizhou University of Finance and Economics,Guiyang 550025,China)
出处
《计算机科学》
CSCD
北大核心
2023年第7期270-277,共8页
Computer Science
基金
国家自然科学基金(62102111)
贵州省科技计划项目([2020]1Y267)
贵州大学引进人才项目((2019)52)。
关键词
网络遥测
路径编排
协同迁移进化
Network telemetry
Path orchestration
Cooperative migration evolution