An Efficient Workflow Scheduling Approach in Cloud Computing

Abstract

Cloud Computing is a latest trend in today’s world. It provides on demand services like hardware, software, platform, infrastructure and storage etc. dynamically to the user according to the “pay per use” model by using virtualized resources over the internet. Cloud computing is able to host various applications such as business, social networks and scientific applications. While Cloud computing provides various services like IaaS, PaaS and SaaS etc. to end users but due to novelty of cloud computing, it also suffers from many types of research issues such as security, performance, database management, virtual machine migration, server consolidation, fault tolerance and workflow scheduling etc. Among these workflow scheduling is major issue for scientific applications. Existing workflow scheduling algorithms in the grid and cloud environment focused on several QoS parameters such as cost, CPU time, makespan and reliability etc. have been surveyed out and some of the time based scheduling algorithms such as First Come First Serve (FCFS), Min-min, Max-min, and Minimum Completion Time (MCT) has also been discussed in this thesis. The time based scheduling algorithms have been used to minimize the execution time only but no awareness has given to utilize resources for reducing execution time. So, a new scheduler named MaxChild has been proposed to increase the resource utilization and to reduce the overall completion time. MaxChild scheduler focuses on the parent-child relationship in the tasks of workflow and schedules the task having the maximum number of Childs. The proposed scheme has been validated by using simulation based analysis though WorkflowSim.