MapReduce Based Implementation of Aggregate Functions on Cassandra
MapReduce is a simple and powerful processing model that allows parallel scalable programs to run on large volume of data on massive cluster of computers. Besides, Cassandra is a popular database of NoSQL solutions. According to scientific knowledge, still there are no general suitable procedures to perform arbitrary calculations in this database on MapReduce model. So, in this paper the authors propose some procedures based on MapReduce model that are needed generally to perform variety of aggregate operations on Cassandra. Their evaluation, compare with the most common methods shows significant improvement in performance on multi-core computers and a set of peer machines.