电信科学 ›› 2020, Vol. 36 ›› Issue (12): 105-112.doi: 10.11959/j.issn.1000-0801.2020320

• 专栏:云原生技术 • 上一篇    下一篇

面向云原生的智能运维架构和关键技术

薛龙,陆钢,周奇,张会炎,万亭君   

  1. 中国电信股份有限公司研究院,广东 广州 510630
  • 修回日期:2020-12-08 出版日期:2020-12-20 发布日期:2020-12-23
  • 作者简介:薛龙(1989- ),男,中国电信股份有限公司研究院工程师,主要从事行业数字化技术、云原生、智能运维及相关人工智能算法研究工作|陆钢(1978- ),男,中国电信股份有限公司研究院企业运营支撑中心副主任、教授级高级工程师,主要从事行业数字化、云计算、物联网、5G网络及应用研发工作|周奇(1992- ),男,中国电信股份有限公司研究院工程师,主要从事行业数字化技术、5G网络及应用相关研发工作|张会炎(1990- ),男,中国电信股份有限公司研究院工程师,主要从事行业数字化技术、5G和IT系统平台研发工作|万亭君(1996- ),女,中国电信股份有限公司研究院工程师,主要从事行业数字化技术、云原生、IT系统平台研发等工作
  • 基金资助:
    国家重点研发计划基金资助项目(2018YFB1802400)

Cloud native intelligent operation and maintenance technology

Long XUE,Gang LU,Qi ZHOU,Huiyan ZHANG,Tingjun WAN   

  1. Research Institute of China Telecom Co.,Ltd.,Guangzhou 510630,China
  • Revised:2020-12-08 Online:2020-12-20 Published:2020-12-23
  • Supported by:
    The National Key Research and Development Program of China(2018YFB1802400)

摘要:

凭借应用简单快捷、部署应用轻松自如、运行应用按需伸缩等优点,云原生技术得到迅速推广。但随着云原生的应用,系统架构愈加复杂,资源规模愈加庞大,迫切需要更加智能化的运维。利用智能运维,可以更好地助力云原生的应用开发,降低运维成本,提高服务质量。分析了智能运维的应用场景、关键点、经典实践以及存在的问题,提出了一种面向云原生的智能运维架构,详细介绍了核心服务算法平台中趋势预测、数据异常检测和故障定位诊断3个场景主要原理,最后展望智能运维未来。

关键词: 云原生, 智能运维, 微服务, 容器

Abstract:

Relying on the advantages of simple and fast applications,easy deployment of applications,and on-demand scaling of running applications,cloud native technology has been rapidly promoted.However,with the application of cloud native,the system architecture has become more complex and the scale of resources has become larger,and there is an urgent need for more intelligent operation and maintenance.Using intelligent operation and maintenance can better assist cloud-native application development,reduce operation and maintenance costs,and improve service quality.The application scenarios,key points,classic practices and existing problems of intelligent operation and maintenance were analyzed,and a cloud-native intelligent operation and maintenance architecture were proposed,and the main principles of the three scenarios:the trend prediction,data anomaly detection and fault location diagnosing in the core service algorithm platform were introduced in detail,and finally the future of intelligent operation and maintenance were looked forward.

Key words: cloud native, AIOps, micro service, container

中图分类号: 

No Suggested Reading articles found!