Design Of The Cloud Platform For Agricultural Information Service With Workload Balancing

In the process of the development of agricultural informatization in China, a large number of agricultural knowledge, data, resources, technology and market demand information not in rural areas are widely quickly spread, grassroots agricultural personnel unable to obtain valuable information, namely, the existence of the "last mile" phenomenon. In view of this problem, this paper presents a cloud platform based on OpenStack for agricultural information service. The cloud platform provides agricultural information dissemination, information inquiry and high performance computing, storage and other hardware and software resource sharing services for agricultural research institutions, agricultural enterprises and institutions and individual farmers. The main work of this paper is as follows:Firstly, the architecture of agricultural information service cloud platform based on Openstack is proposed. Based on the research status of agricultural informationization at home and abroad, and combining with the characteristics of agricultural informationization, the design framework of the agricultural information service cloud platform is put forward. The platform is based on the system hierarchy, which is divided into basic layer, application service layer, management layer and access layer. Between layer and layer using standard interface design, realize single sign on, unified management. On the basis of existing facilities deployed openstack, build a cloud platform, so as to realize the agriculture related resources integration, agricultural information publishing and searching, agricultural production management and interactive services.Then, the resource scheduling strategy for the agricultural information service cloud platform is improved. Aiming at the complexity, diversity and real-time feature of agricultural information service, the openstack original scheduling strategy was improved, put forward a migration of virtual machine based on dynamic resource scheduling strategy, the cloud service platform of agricultural information can better adapt to the change of load of agricultural information.Finally, the performance of the improved scheduling algorithm is verified by experiments. Using Cloudsim simulation software, the improved scheduling algorithm and the commonly used scheduling algorithm are compared. The experimental results show that the improved scheduling algorithm can ensure better load balancing, and good energy saving effect, in does not violate the Service-Level Agreement(SLA) under the premise, the overall energy consumption and response time of the data centers were significantly lower.