基于Kubernetes的卫星遥感数据容器云平台

A Kubernetes based Container Cloud-Platform of Satellite Remote Sensing Data

摘要:

随着卫星遥感行业进入大数据时代，传统的数据处理平台日渐难以满足卫星遥感数据快速积累的发展需求，同时也对系统在动态调度资源、易维护、弹性扩展、复杂度可控和灵活部署等方面提出了更高的要求。针对上述问题，本文引入容器技术及Kubernetes容器集群管理系统，设计实现了基于Kubernetes的开源容器云平台，将单个业务系统拆分成多个独立运行在相互隔离容器中的服务，实现了应用服务容器的调度和管理、快速部署和迁移,并以GF7高分卫星数据为例进行验证，验证了容器集群能够有效处理卫星遥感数据，提高了卫星遥感数据处理平台的资源利用率和运维效率。文末结合卫星遥感行业的现状及特点，就容器化应用的前景及面临的困难进行了展望分析。

Abstract:

As satellite remote sensing industry enters big data era, traditional data processing platforms are gradually difficult to meet the rapid increase of satellite remote sensing data. It also requires better performance of the system on dynamic resource scheduling，maintenance，extendibility and deployment，etc．These features happen to be the advantages of the open source framework Kubernetes. Therefore, a Kubenetes based container cloud-platform of satellite remote sensing data was developed, spreading an individual business system into multiple independent running services into mutual isolation containers. The platform realizes the scheduling and management of application service container, rapid deployment and migration, and improves the resource utilization and operation and maintenance efficiency. Besides, this article analyzes the prospects and difficulties of container applications combined the feature of satellite remote sensing industry.