Evaluating Cassandra scalability with YCSB

Abstract

NoSQL data stores appeared to fill a gap in the database market: that of highly scalable data storage that can be used for simple storage and retrieval of key-indexed data while allowing easy data distribution over a possibly large number of servers. Cassandra has been pinpointed as one of the most efficient and scalable among currently existing NoSQL engines. Scalability of these engines means that, by adding nodes, we could have more served requests with the same performance and more nodes could result in reduced execution time of requests. However, we will see that adding nodes not always results in performance increase and we investigate how the workload, database size and the level of concurrency are related to the achieved scaling level. We will overview Cassandra data store engine, and then we evaluate experimentally how it behaves concerning scaling and request time speedup. We use the YCSB - Yahoo! Cloud Serving Benchmark for these experiments. © 2014 Springer International Publishing Switzerland.