Fault Tolerance in Cloud Computing

Cloud Computing is an important computational paradigm which provide on-demand services to users and in low cost. Fault tolerance and reliability are of great importance that provides correct result even in the presence of faulty components. Most of the systems are safety critical and highly reliable. So to achieve reliability in real time computing, the demand for fault tolerance is increased. In real time computing, the capabilities of intensive computing can be an advantage to execute real time tasks. In most of the applications of real time cloud, processing is done on remote cloud computing nodes. Therefore, due to loose control over the computing node, chances of errors increase. Fault tolerance techniques are used to predict these failures and take an appropriate action before failures actually occur. So to achieve reliability in real time computing, the requirement for fault tolerance increases. The reliability of virtual machines changes after every computing cycle i.e. it is adaptive in nature. In this paper, we introduce an unusual, system level aspect on the creation and management of fault tolerance in clouds. A complete high level approach is introduced to execute the fault tolerant techniques by