Improving Cost-Efficiency through Failure-Aware Server Management and Scheduling in Cloud

We examine the problem of managing a server farm in a cost-efficient way that reduces the cost caused by server failures, according to an Infrastructure-as-a-Service model in cloud. Specifically, failures in cloud systems are so frequent that severely affect the normal operation of job requests and incurring high penalty cost. It is possible to increase the net revenue through reducing the energy cost and penalty by leveraging failure predictiors. First, we incorporate the malfunction and recovery states into the server management process, and improve the cost-efficiency of each server using failure predictor-based proactive recovery. Second, we present a revenue-driven cloud scheduling algorithm, which further increases net revenue in collaboration with server management algorithm. The formal and experimental analysis manifests our expected net revenue improvement.