Science & Engineering Research Support soCiety (SERSC)
Clouds focus on flexibility and on-demand provisioning of virtualized resources, which allow users to scale up and down the usage of resources according to their needs. These characteristics give rise to more challenges for resource allocation in cloud system. The optimization of Virtual Machines (VMs) allocation and the minimization of time and infrastructure cost are still needed to be improved. Meanwhile, failure-aware VMs rescheduling and allocation is another key point to be considered. This paper focuses on the resource failures and presents efficient VMs scheduling mechanisms and policies which try to keep the tradeoff between system performance and QoS.