Review on failure forecast in cloud for a fault tolerant system
J. Nandhini, T. Gnanasekaran · International Journal of Engineering & Technology · 2018
Cloud Computing is an increasingly popular computer paradigm constituting a large infrastructure involving storage, memory, servers and applications accessible via computer network. The cloud system design aims to provide on-demand services with scalability on diverse resources to ensure efficient resource utilization in addition to effectiveness. As cloud is a service-oriented infrastructure, it is critically imperative that the system is highly reliable to meet the Service Level Agreement (SLA). To achieve reliability, cloud requires a very efficient fault tolerance mechanism. Serviceability and reliability is impacted by any failure in the system. Prior prediction of faults in the system helps in overcoming failures. The Fault Tolerance in cloud involves ascertaining the resource fitness to execute scheduled task. The process involves prior screening of resources against various tasks as part of scheduling process. The scheduling process relies significantly on the virtualization of resources to maintain high efficiency.