基于改进DRF算法的农业微服务负载均衡

Agricultural microservice load balancing based on improved dominant resource fairness(DRF)algorithm

微服务作为一种新型的互联网平台架构设计,在现代智慧农业中具有细粒度、低耦合度和高可靠性的优势。在负载均衡智慧农业微服务领域,目前针对复杂农业环境下的特定微服务负载均衡方案的研究较少,而高性能微服务负载均衡策略能够有效提升农业物联网的网络体验和性能指标。本研究提出一种改进的主导资源公平分配(Dominant resource fairness,DRF)算法,通过引入多角度的性能测评因素,着力于提升在智慧农业平台的微服务化下实现服务高效负载均衡的性能。经测试,本研究提出的算法相比于目前流行的微服务平台自带的负载均衡算法在响应时间、吞吐率和稳定性方面有明显提升,实现了各个农业生产单位物联网设备和数据终端以低耦合度的方式接入系统平台,同时实现高效地利用、治理和规范化来自各个数据源的农情数据,具有较高的实用价值。

Microservice is a new type of architecture design for internet platform,with the advantages of fine granularity,low coupling degree and high reliability in modern smart agriculture.In the field of load balancing of smart agricultural microservices,there are few researches on load balancing schemes of specific microservice under complex agricultural environments in the current,and load balancing strategies of efficient microservice can effectively improve the network experience and performance index of agricultural internet of things(IOT).This paper proposed an improved dominant resource fairness(DRF)algorithm and focused on improving the performance of service efficient load balancing under micro-servitization of smart agriculture platform by introducing multi-angle performance evaluation factors.After testing,the algorithm proposed in this paper showed significant improvement in response time,throughput and stability compared with the load balancing algorithms provided by current popular microservice platforms.IOT equipment and data terminal of each agricultural production unit can be connected to the system platform in a mode of low coupling degree,while agricultural data from various data sources can be effectively utilized,managed and standardized.The algorithm proposed in this paper shows good practical value.