Load-Aware Fragment Allocation Strategy for NoSQL Database
Zhikun Chen, Shuqiang Yang, Li He, Shuang Tan, Lumin Zhang, Huiyu Yang, Ge Zhang · 2014
NoSQL databases are famed for the characteristics of high scalability, high availability and high fault-tolerance. So NoSQL databases are used in a lot of applications, especially in the Internet Applications. The computing model of the NoSQL database is transferred from "data close to computing" to "computing close to data". So the data partition strategy and fragment allocation strategy directly affect the load balance of the system. We can not play the advantages of the NoSQL database when the load of the system is unbalance. This paper will analyze the problem of fragment allocation strategy in NoSQL databases, and analyze the strategy how affecting the load balance of system. Then we will propose the load-aware fragment allocation strategy (LAFAS) to improve the load balance of the system. At last we will use some experiments to verify the effectiveness of LAFAS which is proposed in this paper.