摘要：In modern times there is an increasing trend of applications for handling
Big data. However, negotiating with the concepts of the Big data is an extremely
difficult issue today. The MapReduce framework has been in focus recently for
serious consideration. The aim of this study is to get the task-scheduling over Big
data using Hadoop. Initially, we prioritize the tasks with the help of k-means
clustering algorithm. Then, the MapReduce framework is employed. The available
resource is optimally selected using optimization technique in map-phase. The
proposed method uses the FireFly Algorithm and BAT algorithms (FFABAT) for
choosing the optimal resource with minimum cost value. The bat-inspired algorithm
is a meta-heuristic optimization method developed by Xin-She Yang (2010). This bat
algorithm is established on the echo-location behaviour of micro-bats with variable
pulse rates of emission and loudness. Finally, the tasks are scheduled with the
optimal resource in reducer-phase and stored in the cloud. The performance of the
algorithm is analysed, based on the total cost, time and memory utilization.
