基于Docker容器故障恢复的属性权重快照选择策略

doi:10.3969/j.issn.1671-1122.2021.05.002

信息网络安全 ›› 2021, Vol. 21 ›› Issue (5): 12-18.doi: 10.3969/j.issn.1671-1122.2021.05.002

基于Docker容器故障恢复的属性权重快照选择策略

郑军¹, 聂榕², 王守信³(), 谭毓安²

1.北京理工大学网络空间安全学院,北京 100089
2.北京理工大学计算机学院,北京 100089
3.中国卫通集团股份有限公司,北京 100190

收稿日期:2021-02-22 出版日期:2021-05-10 发布日期:2021-06-22
通讯作者: 王守信 E-mail:shouxin_wang@126.com
作者简介:郑军（1969—）,女,辽宁,教授,博士,主要研究方向为网络与信息安全、人工智能和教育技术|聂榕（1996—）,女,山东,硕士研究生,主要研究方向为网络安全|王守信（1979—）,男,辽宁,高级工程师,博士,主要研究方向为软件工程、卫星互联网应用和人工智能机器视觉应用|谭毓安（1972—）,男,重庆,教授,博士,主要研究方向为人工智能安全和系统安全
基金资助:
国家自然科学基金(U1936218);国家自然科学基金(62072037);之江实验室开放课题(2020LE0AB02)

Attribute Weight Snapshot Selection Strategy Based on Docker Container Fault Recovery

ZHENG Jun¹, NIE Rong², WANG Shouxin³(), TAN Yu’an²

1. School of Cyberspace Security, Beijing Institute of Technology, Beijing 100089, China
2. School of Computer Science, Beijing Institute of Technology, Beijing 100089, China
3. China Satellite Communications Co., Ltd., Beijing 100190, China

Received:2021-02-22 Online:2021-05-10 Published:2021-06-22
Contact: WANG Shouxin E-mail:shouxin_wang@126.com

摘要/Abstract

摘要：

近年来,Docker容器技术作为一种新型的虚拟化技术,凭借其隔离性好、成本低、部署安全和方便等优点,得到广泛关注和使用。Docker容器的可靠性问题成为当前研究的热点。然而,Docker容器随时都有很多任务在运行过程中出现故障,且几乎无法避免。容器快照技术成为故障恢复的主流方法之一。当前主要快照选择方法有距离故障发生最小时间差快照选择策略和人工快照选择策略,但这些策略都缺乏对目标容器的综合评估和严谨决策的过程。文章提出一种基于Docker容器故障恢复的属性权重快照选择策略,在进行目标容器恢复操作时,根据不同的需求场景和该容器在使用时的功能特点,结合快照的属性指标权重配置方法与综合评分选择算法,得出适合目标容器进行故障恢复的最优快照。通过对几项容器资源指标进行评估,发现与最小时间差选择策略相比,文章策略能使容器在恢复后的性能状态更佳。

关键词: Docker容器技术, 故障恢复, 快照选择策略

Abstract:

In recent years, as a new virtualization technology, Docker container technology has been widely concerned and used because of its advantages of good isolation, low cost, safe and convenient deployment. The reliability of Docker container has become the focus of current research. However, there are many tasks in Docker container that break down at any time during operation, which is almost unavoidable. Container snapshot technology is one of the main methods for fault recovery. At present, the main snapshot selection methods are minimum time difference from fault snapshot selection strategy and manual snapshot selection strategy, but none of these strategies has a comprehensive evaluation and rigorous decision-making process for the target container. This paper proposes an attribute weight snapshot selection strategy based on Docker container fault recovery. When recovering the target container, according to different requirements scenarios and function characteristics of the container in use, the attribute index weight allocation method and comprehensive scoring selection algorithm of the snapshot are combined to get the optimal snapshot suitable for the target container to recover from a fault. Through the evaluation of several container resource indicators, it is found that compared with the minimum time difference selection strategy, the strategy proposed in this paper can make the container in a better performance state after recovery.

Key words: Docker container technology, fault recovery, snapshot selection strategy

中图分类号:

TP309

郑军, 聂榕, 王守信, 谭毓安. 基于Docker容器故障恢复的属性权重快照选择策略[J]. 信息网络安全, 2021, 21(5): 12-18.

ZHENG Jun, NIE Rong, WANG Shouxin, TAN Yu’an. Attribute Weight Snapshot Selection Strategy Based on Docker Container Fault Recovery[J]. Netinfo Security, 2021, 21(5): 12-18.

图/表 10

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

参考文献 18

[1]	ABENI L, BALSINI A, CUCINOTTA T. Container-based Real-time Scheduling in the Linux Kernel[J]. ACM SIGBED Review, 2019,16(3):33-38.
[2]	LIN Xin, LEI Lingguang, WANG Yuewu, et al. A Measurement Study on Linux Container Security: Attacks and Countermeasures[C]// ACM. The 34th Annual Computer Security Applications Conference, December 3-7, 2018, San Juan, PR, USA. New York: ACM, 2018: 418-429.
[3]	MA Hua, NIE Changhai, WU Huayao. History-based Fault Injection Testing for Cloud Platform[J]. Journal of Computer Science, 2019,42(10):2281-2296.
	马骅, 聂长海, 吴化尧. 基于历史的云平台故障注入测试[J]. 计算机学报, 2019,42(10):2281-2296.
[4]	LU Zhigang, XU Jiwei, HUANG Tao. Container Image Deduplication Method Based on Chunking Reuse of Multi-versions[J]. Journal of Software, 2020,31(6):1875-1888.
	陆志刚, 徐继伟, 黄涛. 基于分片复用的多版本容器镜像加载方法[J]. 软件学报, 2020,31(6):1875-1888.
[5]	YANG Na, LIU Jing. Optimized Fault Tolerance as Services Provisioning for Cloud Applications[J]. Journal of Software, 2019,30(4):1191-1202.
	杨娜, 刘靖. 面向云应用系统的容错即服务优化提供方法[J]. 软件学报, 2019,30(4):1191-1202.
[6]	ROMAN E. A Survey of Checkpoint/Restart Implementations[EB/OL]. https://xueshu.baidu.com/usercenter/paper/show?paperid=5b1401189e635ec43330021f49b72406&site=xueshu_se, 2020-12-20.
[7]	SCHULZ M, BRONEVETSKY G, FERNANDES R, et al. Implementation and Evaluation of a Scalable Application-level Checkpoint-recovery Scheme for MPI Programs[C]// IEEE. The 2004 ACM/IEEE Conference on Supercomputing In Supercomputing, November 6-12, 2004, Pittsburgh, PA, USA. NJ: IEEE, 2004: 38.
[8]	WANG Long, PATTABIRAMAN K, KALBARCZYK Z, et al. Modeling Coordinated Checkpointing for Large-scale Supercomputers[C]// IEEE. 2005 International Conference on Dependable Systems and Networks, June 28-July 1, 2005, Yokohama, Japan. NJ: IEEE, 2005: 812-821.
[9]	GUERMOUCHE A, ROPARS T, BRUNET E, et al. Uncoordinated Checkpointing without Domino Effect for Send-deterministic MPI Applications[C]// IEEE. 25th International Parallel & Distributed Processing Symposium, May 16-20, 2011, Anchorage, AK, USA. NJ: IEEE, 2011: 989-1000.
[10]	SAMIR A, PAHL C. Self-adaptive Healing for Containerized Cluster Architectures with Hidden Markov Models[C]// IEEE. The IEEE 4th International Conference on Fog and Mobile Edge Computing, Jun 10-13, 2019, Rome, Italy. NJ: IEEE, 2019: 68-73.
[11]	BELLA M R M, DATA M, YAHYA W. Web Server Load Balancing Based on Memory Utilization Using Docker Swarm[C]// IEEE. 2018 International Conference on Sustainable Information Engineering and Technology (SIET), November 10-12, 2018, Malang, Indonesia. NJ: IEEE, 2018: 220-223.
[12]	LEONG H V, CHAN A, NGAI G. Approximate Web Database Snapshots[EB/OL]. https://www.zhangqiaokeyan.com/open-access_resources_thesis/0100055856610.html, 2020-12-20.
[13]	ZHANG Hao, SUN Yuzhong, XIAO Li, et al. RainbowD: A Heterogeneous Cloud-oriented Efficient Docker Image Distribution System[J]. Journal of Computer Science, 2020,43(11):2067-2083.
	张浩, 孙毓忠, 肖立, 等. RainbowD: 一种异构云环境下高效的Docker镜像分发系统[J]. 计算机学报, 2020,43(11):2067-2083.
[14]	POTDAR A M, NARAYAN D G, KENGOND S, et al. Performance Evaluation of Docker Container and Virtual Machine[J]. Procedia Computer Science, 2020,171(4):1419-1428. doi: 10.1016/j.procs.2020.04.152 URL
[15]	AMIRANTE A, ROMANO S P. Container NATs and Session-oriented Standards: Friends or Foe?[J]. IEEE Internet Computing, 2019,23(6):28-37. doi: 10.1109/MIC.4236 URL
[16]	WANG Peijian. Redis High Availability Implementation Based on Docker Container[J]. Information & Communications, 2019,33(8):67-70.
	王佩剑. 基于Docker容器的Redis高可用实现方案[J]. 信息通信, 2019,33(8):67-70.
[17]	XU Jiwei, ZHANG Wenbo, WANG Tao, et al. A Genetic Algorithm-based Adaptive Strategy for Image Backup of Virtual Machines[J]. Journal of Computer Science, 2016,39(2):351-363.
	徐继伟, 张文博, 王焘, 等. 一种基于遗传算法的虚拟机镜像自适应备份策略[J]. 计算机学报, 2016,39(2):351-363.
[18]	WANG Wei, ZHANG Xiangliang, Gombault S, et al. Attribute Normalization in Network Intrusion Detection[EB/OL]. https://ieeexplore.ieee.org/document/5381578/keywords#keywords, 2020-12-20.

基于Docker容器故障恢复的属性权重快照选择策略

Attribute Weight Snapshot Selection Strategy Based on Docker Container Fault Recovery

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 18

相关文章 1

编辑推荐

Metrics

本文评价