An Efficient Fault Tolerance Mechanism Based on Moving Averages Algorithm

Fault Tolerance is concerned with all the techniques necessary to enable a system to tolerate faults remaining in the system after its development and often we encounter these failures using a well defined Fault Tolerance Mechanism (FTM).This technique is enabled using a Virtual Machine (VM) migration policy. This paper proposed a new method of identifying the need to migrate VM Machines, whenever there is an overload of work in the VM(s). The proposed algorithm provided with a way of computing the moving (dynamic) current thresholds as per current load of the data center calculated on the basis of effectiveness factor, which further provides us with implementation of a policy which remains refresh, since the payload of the datacenter varies during the day. Keywords— Cloud Computing, Fault Tolerance, VM Migration, Moving Averages, Effectiveness Factor