Map Reduce (MR) system symphonizes the processing of high data intensive jobs in distributed cloud environments. Nowadays all the MR tasks have been deadline constraint based job for execution. At the same time ourcurrent time critical systems donot consider the following issues such as execution cost, cluster performance, nodes performance, task running time. Moreover, in heterogeneous environment, it is very difficult to schedule a job in its preferred resources in order to achieve high performance. Hence, to address the above issues, a new map reduce scheduler is modelled using Toilet queue based scheduling scheme (TQSS). The proposed scheduling model uses the population based candidate solutions rather than single path solution as in traditional scheduling mechanism which avoids trapping in local optimum. To evaluate the proposed model, thereal time experimentation has been carried out with proposed scheduler in virtual cluster with 16 Nodes.In essence, this paper also analyses the performance of cluster based on the running time and cluster efficiency. Finally, from the experimental results it is observed that proposed scheduler provides the above mentioned state of the art scheduling model with better efficiency in theminimumexecution time.
Volume 11 | 06-Special Issue
Pages: 647-655