数据与计算发展前沿 ›› 2024, Vol. 6 ›› Issue (1): 94-101.

CSTR: 32002.14.jfdc.CN10-1649/TP.2024.01.009

doi: 10.11871/jfdc.issn.2096-742X.2024.01.009

• 技术与应用 • 上一篇    下一篇

SDGs科研工作台架构设计与实现

陈灿*(),朴英超   

  1. 中国科学院计算机网络信息中心,北京 100083
  • 收稿日期:2022-11-18 出版日期:2024-02-20 发布日期:2024-02-21
  • 通讯作者: * 陈灿(E-mail: chencan@cnic.cn
  • 作者简介:陈灿,中国科学院计算机网络信息中心,工学博士,高级工程师,主要研究方向为云原生技术、大数据分析技术。近年来主要参与了中国科学院A类先导专项、中国科学院十四五信息化工程、全球开放科学云计划等项目。
    在本文中主要承担工作为完成了对SDGs科研工作台的架构设计和系统研发,论文初撰。
    CHEN Can, Ph.D., Senior Engineer, Computer Network Information Center, CAS, China. His research interests include cloud native and big-data analysis. He worked in the Big Earth Data Science Engineering Program (CASEarth) and Global Open Science Cloud (GOSC) in recent years.
    In this paper, he is mainly responsible for designing the architecture and developing the system of the SDGs workbench.
    E-mail: chencan@cnic.cn
  • 基金资助:
    中国科学院A类战略先导项目“地球大数据科学工程”(XDA19020104)

Design and Implementation of SDGs Workbench Architecture

CHEN Can*(),PIAO Yingchao   

  1. Computer Network Information Center, Chinese Academy of Sciences, Beijing 100083, China
  • Received:2022-11-18 Online:2024-02-20 Published:2024-02-21

摘要:

【目的】地球大数据平台为可持续发展目标研究提供了计算资源和数据资源的支持,但各学科研究人员对于资源的占用不均衡,且对资源的使用率也不高。本文致力于解决地球大数据平台中资源使用的问题,更好地服务科研人员开展SDGs研究。【方法】本文采用云原生技术构建了一个面向可持续发展研究的一站式科研工作环境,为研究人员提供在线获取和处理数据、训练和使用模型、构建和部署应用软件等服务。【结果】极大地减轻了科研人员搭建科研软件栈的工作负担,提高了对占用资源的使用率,同时依托云原生的自动扩缩容能力,实现了资源的均衡使用。【结论】通过云原生架构构建的科研工作台,实现了云服务从研发、测试、部署、版本更新到使用的一体化应用生态,有效地支持了面向数据驱动的SDGs研究新范式。

关键词: 可持续发展目标, 云原生, 一站式科研环境, Kubernetes, 地球大数据

Abstract:

[Objective] The CASEarth Big Earth Data system provides support of computing resources and data resources for the study of sustainable development goals. The occupation of resources by researchers in various disciplines is uneven, and the utilization of resources is not high. The purpose of this paper is to solve the problem of resource usage in the current system and to better serve the researchers. [Methods] This paper has designed and constructed a one-stop scientific research environment for SDGs research based on cloud-native technology. Researchers can obtain and process data online, train and use models, and build and deploy applications online. [Results] The workbench can reduce the workload of researchers in building scientific research software stacks, and improve the utilization of occupied resources. At the same time, relying on the automatic scaling capabilities of cloud-native, the workbench can balance the use of resources. [Conclusions] The SDGs workbench can form an application ecology of cloud services for R&D, testing, deployment, and version updating, which enables a new data-driven SDG scientific research paradigm.

Key words: Sustainable Development Goals, cloud-native, one-stop scientific research environment, Kubernetes, CASEarth big earth data