Cross-area Node Selection Algorithm for Data Replica Storing in Cloud Data Centers

Abstract

With the arrival of cloud computing and big data, enterprise application systems are supported by massive data storing in cloud data centers (CDCs), in which the popular open-source NoSQL databases (Cassandra, HBase, MongoDB) are utilized to cope with high concurrency, high availability, and high scalability. There are still two challenging issues: (1) When the consistency level and replica number is specified, selecting the appropriate nodes for data replicas to reduce the communication latency is necessary. (2) A tradeoff between reliability and synchronization time should be discussed in CDCs. In this paper, a cross-area node selection algorithm (CANSA) is proposed to minimize the communication latency, then a new method to adjust reliability in the data center is designed. Numerical results in the Cassandra cluster of CDCs demonstrate that the CANSA provides excellent consensus ratio, synchronization, and latency.