Multi-resource Workload Consolidation in Cloud Data Centers

Consolidation of Virtual Machines (VMs) on the minimum number of physical servers has been recognized as a very efficient approach to increase the efficiency of virtualized data centers and save energy, as consolidation allows unloaded servers to be switched off or used to accommodate more load. The problem is so complex that centralized and deterministic solutions are useless in large data centers with hundreds or thousands of servers. This paper presents a self-organizing approach for the consolidation of VMs on two resources, CPU and RAM. Decisions on the assignment and migration of VMs are driven by probabilistic processes and are based on local information, which makes the solution simple to implement and scalable. Experiments on a real data center show that the approach rapidly consolidates the workload, and CPU-bound and RAM-bound VMs are balanced, so that both resources are exploited efficiently.