PROACTIVE LOAD BALANCING FAULT TOLERANCE IN REAL CLOUD COMPUTING (PLBFT)

Abstract

Reliability is one of cloud computing research issues; because of different reasons such as network instability, unprofessional use … etc. Some of these reasons cause high CPU utilization, memory crashes, full storage. The aim of our thesis is to increase the reliability of cloud systems. This is achieved by developing a proactive fault tolerance model with a load balancing system, this model deals with CPU faults in a virtual machine environment. The prediction of faults before occurrence is the main idea of our research. Finally, we tested our model with a real cloud system contains four nodes (controller node, network node, and two compute nodes)