The performance evaluation of proactive fault tolerant scheme over cloud using CloudSim simulator

Abstract

The main issues in a cloud based environment are security, process fail rate and performance. Fault tolerance plays a key role in ensuring high serviceability and reliability in cloud. Nowadays, demands for high fault tolerance, high serviceability and high reliability are becoming unprecedentedly strong, building a high fault tolerance, high serviceability and high reliability cloud is a critical, challenging, and urgently required task. A lot of research is currently underway to analyze how clouds can provide fault tolerance for an application. When the number of processes are too many and the virtual machine is overloaded then the processes are failed causing lot of rework and annoyance for the users. The major cause of the failure of the processes at the virtual machine level are overloading of virtual machines, extra resource requirements of the existing processes etc. This paper introduces dynamic load balancing techniques for cloud environment in which RAM/Broker (resource awareness module) proactively decides whether the process can be applied on an existing virtual machine or it should be assigned to a different virtual machine created a fresh or any other existing virtual machine. So, in this way it can tackle the occurrence of the fault. This paper also proposed a mechanism which proactively decides the load on virtual machines and according to the requirement either creates a new virtual machine or uses an existing virtual machine for assigning the process. Once a process is complete, it will update the virtual machine status on the broker service so that other processes can be assigned to it.