Auto-scalable and fault-tolerant load balancing mechanism for cloud computing based on the proof-of-work election

作     者：Xiaoqin FENG Jianfeng MA Shaobin LIU Yinbin MIAO Ximeng LIU Xiaoqin FENG;Jianfeng MA;Shaobin LIU;Yinbin MIAO;Ximeng LIU

作者机构：School of Cyber Engineering Xidian University State Key Laboratory of Cryptology Shaanxi Key Laboratory of Network and System Security Xidian University College of Mathematics and Computer Science Fuzhou University 

出 版 物：《Science China(Information Sciences)》 (中国科学:信息科学(英文版))

年 卷 期：2022年第65卷第1期

页      面：131-146页

核心收录：

学科分类：0711[理学-系统科学] 08[工学] 0804[工学-仪器科学与技术] 080402[工学-测试计量技术及仪器] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported in part by Key Program of National Natural Science Foundation of China (NSFC)(Grant No. U1405255) in part by Shaanxi Science&Technology Coordination&Innovation Project (Grant No. 2016KTZDGY05-06) in part by Fundamental Research Funds for the Central Universities (Grant No. SA-ZD161504) in part by National Natural Science Foundation of China (Grant No. 61702404) in part by Fundamental Research Funds for the Central Universities (Grant No. JB171504) in part by Project Funded by China Postdoctoral Science Foundation (Grant No. 2017M613080) in part by Major Nature Science Foundation of China (Grant Nos. 61370078, 61309016) in part by Key Research and Development Plan of Jiangxi Province (Grant No. 0181ACE5002) 

主　　题：cloud computing load balancing auto-scalable proof-of-work fault-tolerant 

摘      要：Load balancing mechanism in technologies such as cloud computing has provided a huge opportunity for the development of large-scale projects. Although the conventional view is to build mechanisms that adopt a dynamic load balancing strategy, existing strategies cannot automatically scale platform(network) servers(nodes) to adapt for dynamic requests, but only guarantee load balancing for the pre-deployed nodes, thereby increasing resource consumption and decreasing networks’ efficiency. We contend that existing load balancing mechanisms are inadequate for deploying dynamic applications. In this regard, we first adopt both load balancing and cloud computing virtualization technologies to modularly design a load balancing mechanism that provides a dynamically auto-scalable solution for large-scale and dynamic computing scenarios. Furthermore, we adopt the proof-of-work consensus, for a novel use during the lifecycle of master nodes in case of system failure caused by a failed master node, to demonstrate a fault-tolerant load balancing mechanism. We theoretically evaluate the security requirement of our mechanism and analyze its performance. Experimental results show that the mechanism supports auto-scalability and has a better performance compared to existing mechanisms such as the ordinary cluster.