多因素自适应心跳检测算法研究

发布时间：2018-04-04 18:05

本文选题：分布式系统　切入点：心跳检测　出处：《计算机工程与应用》2017年24期

【摘要】：分布式系统中心跳检测是节点故障检测机制的关键技术之一,心跳频率设定的合理性将影响到故障检测的准确性和完整性。针对大数据环境下,分布式系统产生故障受到网络、节点、作业多方面影响,为了提高心跳频率在多方面因素影响下的合理性设定,提出一种多因素心跳检测综合指标评价模型。在该模型下同时考虑网络负载情况和节点CPU工作状态及节点作业的大小对心跳检测过程的影响。在此基础上,提出了基于多因素评价模型的自适应心跳检测算法。该算法可以随网络环境、节点CPU占用率、作业任务大小自适应地改变心跳频率,综合各因素给出心跳频率设定的最优方案。最后通过实验验证了多因素对心跳频率自适应调整的影响。
[Abstract]:Distributed system center hop detection is one of the key technologies of node fault detection mechanism. The rationality of heartbeat frequency will affect the accuracy and integrity of fault detection.In the environment of big data, distributed system fault is affected by network, node and job. In order to improve the rationality of heartbeat frequency under the influence of many factors,A multi-factor comprehensive index evaluation model for heartbeat detection is proposed.In this model, the influence of network load, the working state of node CPU and the size of node job on the heartbeat detection process is considered simultaneously.On this basis, an adaptive heartbeat detection algorithm based on multi-factor evaluation model is proposed.The algorithm can adaptively change the heartbeat frequency according to the network environment, node CPU occupancy rate, job task size, and give the optimal scheme of heartbeat frequency setting by synthesizing various factors.Finally, the effects of multiple factors on the adaptive adjustment of heart rate are verified by experiments.
【作者单位】：辽宁大学创新创业学院;东北大学软件学院;
【基金】：国家青年自然科学基金(No.61502090) 教育部博士点基金新教师类(No.20130042120006);教育部基本科研业务费种子基金项目(No.151704003) 辽宁省档案局科技项目(No.L-2017-X-24) 辽宁省高校健康管理协同创新中心资助项目辽宁省教育科学“十三五”规划课题(No.JG17DB230)
【分类号】：TP301.6

【相似文献】