Performance analysis of cloud computing for complex scientific workflows

Abstract

The scientific workflow management is always a keen issue in newlinevarious industries where there is a need to allocate available resources newlinetowards different jobs. So that identifying the order of execution of tiny task newlineover available resources has become a challenging issue. On the other side, newlinenot all organizations have the capability to afford the cost which required newlinepurchasing costlier resources. The growth of information technology has newlineopened the gate for such organizations to execute the jobs by inventing newlinedistributed, grid, parallel computing strategies. On the way with a growth of newlinegrid computing the modern cloud environment has been developed. The cloud newlineenvironment has the beauty of maintaining various organizations data in newlinedifferent data centers which are located in global locations. Also, the service newlineprovider enables the access of data through number of services which are newlineprovided by Cloud Service Provider (CSP). However, the growing size of newlineorganization and data increases the challenge of scheduling the resources and newlineexecuting the task given in form of request from the clients. Whatever the newlinestrategy being used for workflow management, the quality of service of the newlinecloud is depending on various constrains like the time complexity, power newlineconsumption, resource utilization. Towards the development of workflow management, different algorithms have been discussed earlier. Some of the algorithms consider the newlineexecution time or makes span time, but does not consider the other factors newlinelike throughput, energy utilization and so on. newline newline