Genetic Clustering with Workload Multi-task Scheduler in Cloud Environment

Abstract

One of the challenging problems in the cloud environment is many-task computing paradigm as they contain large volumes of datasets and loosely coupled tasks. The data possession scheme using the cooperative provable model (CPDP) was based on homomorphism that provided reliability by automatically maintaining the multiple photocopy of information. In specific, the CPDP scheme for huge files still required to address the cluster network model for dynamically updating the CPDP parameters. Another scheduling scheme based on the multi-objective (MOS) scheme was specifically designed and applied using the ordinal optimization (OO) method for clouds. In addition, MOS also used different memory and disk requirements, increasing the workload while performing multi-tasking. To increase the performance during multi-tasking, Genetic Clustering with Workload Multi-task (GCWM) scheduler scheme is introduced. GCWM scheduler is based on clustering of similar workload using the genetic concepts which minimizes the computational cost and complexity involved during computation. GCWM scheduler scheme is applied to cluster ‘n’ tasks with initial population (i.e.,) tasks, selection, crossover and mutation operators for workload management. The fitness function in GCWM scheduler scheme cluster similar task in cloud zone and communicate with each other effectively. Genetic Clustering with Workload Multi-task scheduling scheme uses distributed computing resources. GCWM Scheduler ensures the multi tasking operation with efficient users' communication. Genetic Clustering Based Workload Multi-task scheduling scheme is experimented on the factors such as throughput, workload management efficacy, relative cost, and multi-task cluster effect.