Memetic Algorithm for Multi-objective Workflow Scheduling In Cloud

Objectives: Cloud computing is a service delivery over the internet where users pay based on the usage and the Quality of service (Qos). The cloud environment supports high performance computing based on protocols, which allow sharing of computation and storage. Scheduling in a cloud is the process of scheduling the virtual machines (VM) to meet the customer’s request. Methods/Statistical Analysis: The proposed evolutionary algorithm called Memetic Algorithm (MA) takes makespan and total cost as two objectives and gives an optimal workflow schedule of jobs. Findings: The algorithm is testing with different IaaS parameters from Amazon. Results show that MA gives significantly better solution than other algorithms like Genetic Algorithm (GA) and Iasi Cloud Partial Critical Path (IC-PCP). The schedule generated by MA gives more stability on most of the workflow instances. Application/Improvements: The proposed model applied to schedule the VMs in a cloud in an effective way.