Performance Analysis for Dynamic Task Migration in Cloud Computing

摘要

The popularity of cloud computing as an attractive alternative to classic Information processing systems has been growing. Cloud Computing offers various remotely accessible services to users either free or on a pay-as-you-go (PAYG) basis. To provide high level of reliability and availability in such large-scale systems, it is critical to detect, diagnose, and fix these failures as they happen. The failures can be attributed to high CPU usage, memory consumption, and network delay. Therefore, resource monitoring plays an important role in avoiding failures. CPU usage is mostly used by researchers to discover virtual machines failure. Resource provisioning and virtual machines or tasks migration are used to maintain task performance. This study focuses on Infrastructure as a Service (IaaS). We use an effective method for avoiding failures by using proactive method through live monitoring CPU usage and dynamic task migration. The performance of model evaluated by measuring execution time, CPU utilization, and makespan for VMs different virtual. This study applied dynamic task migration as management resource for predicting failure event. The proposed model decreases the average execution time, CPU utilization while, model has low performance with makespan. Our model was developed and tested using cloud computing simulator called CloudSim.