A Shared Protection Algorithm for Data Center Network Nodes

doi:10.3969/j.issn.1671-1122.2020.10.004

Abstract

Abstract:

Aiming at the problem of the data center network survivability, this paper proposes a shared protection algorithm for important network nodes in data center. The algorithm firstly searches for nodes that are prone to failure to be protected. Then, protection routes are established according to the shortest path principle and the shared protection principle, and the nodes to be protected that are included in the current protection route are detected to avoid repeated protection and improve the utilization of network resources. Finally, according to the principle of capacity equivalence, the data transmission capacity relationship between the current protection route and the working route is judged. By adding links to ensure that the protection route capacity is not less than the working route capacity, ensure that data can be transmitted normally and effectively when the network fails, and improve the real-time response capability of the data center network. The simulation results show that the number and length of the protection routes established by the algorithm are less. On the basis of ensuring that the data center is not affected by faulty nodes and data is transmitted normally and effectively, the resource consumption of the protection route is reduced, and the survivability of the data center network is improved.

Key words: data center, survivability, protection route, shared protection

CLC Number:

TP309

LIU Yi, LI Jianhua, CHEN Yu, QI Zisen. A Shared Protection Algorithm for Data Center Network Nodes[J]. Netinfo Security, 2020, 20(10): 27-33.

Figures/Tables 11

References 15

[1]	WEI Xianglin, CHEN Ming, FAN Jianghua, et al. Data Center Network Architecture[J]. Software Journal, 2013,24(2):295-316.
	魏祥麟, 陈鸣, 范建华, 等. 数据中心网络的体系结构[J]. 软件学报, 2013,24(2):295-316.
[2]	RAMASWAMI R, SIVARAJAN K N, SASAKI G H. Optical Networks[M]. Amsterdam: Elsevier, 2010.
[3]	ABHISHEK R, TIPPER D, MEDHI D. Network Virtualization and Survivability of 5G Networks: Framework, Optimization Model, and Performance [C]//IEEE. 2018 IEEE Globecom Workshops (GC Wkshps), December 9-13, 2018, Abu Dhabi, United Arab Emirates. NJ: IEEE, 2019: 1-6.
[4]	XIAO Jie, WEN Hong, WU Bin, et al. Joint Design on DCN Placement and Survivable Cloud Service Provision over All-Optical Mesh Networks[J]. IEEE Transactions on Communications, 2014,62(1):235-245.
[5]	FRANGOUDIS P A, YALA L, KSENTINI A. CDN-as-a-Service Provision over a Telecom Operator's Cloud[J]. IEEE Transactions on Network & Service Management, 2017,14(3):702-716.
[6]	GU Feng, ALAZEMI H, RAYES A, et al. Survivable Cloud Networking Services [C]//IEEE. 2013 International Conference on Computing, Networking and Communications (ICNC), January 28-31, 2013, San Diego, CA, USA. NJ: IEEE, 2013: 1016-1020.
[7]	PENG Yuhuai, WANG Xiaojie, SHEN Dawei, et al. Design and Modeling of Survivable Network Planning for Software-defined Data Center Networks in Smart City[J]. International Journal of Communication Systems, 2018,31(16):1-14.
[8]	TAKITA Y, HASHIGUCHI T, TAJIMA K, et al. Towards Seamless Service Migration in Network Re-optimization for Optically Interconnected Datacenters[J]. Optical Switching and Networking, 2017,23(3):241-249. doi: 10.1016/j.osn.2016.11.002 URL
[9]	CHEN Tianping, MENG Xiangru, CUI Wenyan, et al. Active Service Drift Model Based on Situational Awareness of Network Survivability[J]. Journal of Air Force Engineering University: Natural Science Edition, 2015,16(6):64-68.
	陈天平, 孟相如, 崔文岩, 等. 基于网络可生存性态势感知的主动服务漂移模型[J]. 空军工程大学学报:自然科学版, 2015,16(6):64-68.
[10]	QIN Yang, YE Yibing, YANG Weihong, et al. Multicast Routing and Recovery-based in Data Center Networks [C]//IEEE. 2015 10th International Conference on Communications & Networking in China, August 15-17, 2015, Shanghai, China. NJ: IEEE, 2016: 864-870.
[11]	MEIXNER C C, DIKBIYIK F, TORNATORE M, et al. Disaster-resilient Virtual-network Mapping and Adaptation in Optical Networks [C]//IEEE. 2013 17th International Conference on Optical Networking Design and Modeling (ONDM), April 16-19, 2013, Brest, France. NJ: IEEE, 2013: 107-112.
[12]	YANG Yuan, YIN Shan, GUO Bingli, et al. Data Center Networks Recovery for Large-scale Disasters in Optical Cloud Networks [C]//IEEE. 2016 15th International Conference on Optical Communications & Networks, September 24-27, 2016, Hangzhou, China. NJ: IEEE, 2016: 1-3.
[13]	MOHANDESPOUR M, KAMAL A E. Multicast 1+1 Protection: The Case for Simple Network Coding [C]//IEEE. 2015 International Conference on Computing, February 16-19, 2015, Garden Grove, CA, USA. NJ: IEEE, 2015: 853-857.
[14]	YANG Yuan. Research on Protection and Restoration Methods Based on Data Center Optical Network[D]. Beijing: Beijing University of Posts and Telecommunications, 2017.
	杨源. 基于数据中心光网络的保护恢复方法研究[D]. 北京:北京邮电大学, 2017.
[15]	HU Yuxiang, LI Ziyong, HU Zongkui, et al. Software-defined Network Control Resource Optimization Mechanism Based on Traffic Engineering[J]. Journal of Electronics and Information, 2020,42(3):661-668.
	胡宇翔, 李子勇, 胡宗魁, 等. 基于流量工程的软件定义网络控制资源优化机制[J]. 电子与信息学报, 2020,42(3):661-668.

符号	含义
${{l}_{i}}$	网络拓扑中链路${{e}_{i}}$的长度,${{e}_{i}}\in E$
${{c}_{i}}$	网络拓扑中链路${{e}_{i}}$的容量,$C=\left\{ {{c}_{1}},{{c}_{2}},\cdots ,{{c}_{N}} \right\}$,$N$表示网络链路总数
$s$	一次数据传输请求的源节点,$s\in V$
D	一次数据传输请求的目的节点集合,$D\in V-s$
J	一次数据传输的链路请求总数,目的节点总数决定了链路请求总数
d_j	一次数据传输请求里第j个链路请求的目的节点,${{d}_{j}}\in D$,j=1,2,…,J
$H$	一次数据传输请求的总数据量
${{h}_{j}}$	一次数据传输请求里第j个链路请求的数据量
${{R}_{s,D,H}}$	从s到D的一次数据传输请求,其数据量为$H$
${{r}_{s,{{d}_{j}},{{h}_{j}}}}$	从s到d_j的第j个链路请求,其数据量为${{h}_{j}}$
$W$	R的工作路由集合,总数为?
${{w}_{\tau }}$	R的一条工作路由,${{w}_{\tau }}\in W$,τ =1,2,…,?
$M$	$M=\{d_{\gamma }^{g},{{d}_{j}},l_{\gamma }^{g},{{h}_{\gamma }}\}$,表示数据传输请求中易受攻击而发生故障的重要网络节点d_j的参数集合。其中,$d_{\gamma }^{g}$表示由于d_j故障而待保护的网络节点,$l_{\gamma }^{g}$表示$d_{\gamma }^{g}$与d_j之间的路由距离,${{h}_{\gamma }}$表示经过d_j的链路的数据量
${{V}_{g}}$	${{V}_{g}}=\{d_{\gamma }^{g},l_{\gamma }^{g}\}$,表示待保护节点$d_{\gamma }^{g}$和相应保护路由距离$l_{\gamma }^{g}$的集合
$P$	待保护节点的保护路由集合,$P=\left\{ {{p}_{1}},{{p}_{2}},\cdots ,{{p}_{O}} \right\}$,O表示保护路由总数
$h_{\sigma }^{p}$	保护路由${{p}_{\sigma }}$的数据容量,$\sigma \text{=}1,2,\cdots ,O$

工作路由	数据容量
${{w}_{5-2}}$,${{w}_{5-4}}$,${{w}_{\text{5}-\text{12}}}$,${{w}_{\text{5}-\text{13}}}$	1
${{w}_{\text{5}-\text{1}}}$,${{w}_{\text{5}-\text{6}}}$,${{w}_{\text{5}-\text{10}}}$	2
${{w}_{\text{5}-\text{14}}}$	3

工作路由	经过的节点
${{w}_{\text{5}-1}}$	${{\varphi }_{5}}\to {{\varphi }_{4}}\to {{\varphi }_{1}}$
${{w}_{\text{5}-\text{2}}}$	${{\varphi }_{5}}\to {{\varphi }_{4}}\to {{\varphi }_{2}}$
${{w}_{5-\text{4}}}$	${{\varphi }_{5}}\to {{\varphi }_{4}}$
${{w}_{\text{5}-\text{6}}}$	${{\varphi }_{5}}\to {{\varphi }_{6}}$
${{w}_{\text{5}-\text{10}}}$	${{\varphi }_{5}}\to {{\varphi }_{12}}\to {{\varphi }_{10}}$
${{w}_{\text{5}-\text{12}}}$	${{\varphi }_{5}}\to {{\varphi }_{12}}$
${{w}_{\text{5}-\text{13}}}$	${{\varphi }_{5}}\to {{\varphi }_{6}}\to {{\varphi }_{7}}\to {{\varphi }_{8}}\to {{\varphi }_{13}}$
${{w}_{\text{5}-\text{14}}}$	${{\varphi }_{5}}\to {{\varphi }_{6}}\to {{\varphi }_{14}}$

网络链路	数据容量
${{e}_{2-4}}$,${{e}_{7-\text{8}}}$,${{e}_{11-\text{14}}}$	1
${{e}_{1-2}}$,${{e}_{\text{2}-\text{3}}}$,${{e}_{\text{4}-\text{5}}}$,${{e}_{\text{5}-\text{9}}}$,${{e}_{\text{5}-\text{12}}}$,${{e}_{10-13}}$	2
${{e}_{1-\text{4}}}$,${{e}_{1-\text{8}}}$,${{e}_{\text{3}-\text{6}}}$,${{e}_{\text{9}-\text{13}}}$,${{e}_{10-\text{12}}}$	3
${{e}_{5-6}}$,${{e}_{8-13}}$,${{e}_{10-\text{14}}}$	4
${{e}_{\text{6}-\text{7}}}$,${{e}_{\text{6}-\text{14}}}$,${{e}_{11-\text{13}}}$,${{e}_{11-12}}$	5

保护路由	经过的节点	数据容量
${{p}_{1}}$	${{\varphi }_{5}}\to {{\varphi }_{12}}\to {{\varphi }_{10}}\to {{\varphi }_{14}}$	2
${{p}_{2}}$	${{\varphi }_{5}}\to {{\varphi }_{9}}\to {{\varphi }_{13}}\to {{\varphi }_{11}}\to {{\varphi }_{14}}$	1
${{p}_{3}}$	${{\varphi }_{5}}\to {{\varphi }_{6}}\to {{\varphi }_{3}}\to {{\varphi }_{2}}\to {{\varphi }_{1}}$	2