An energy saving heuristic algorithm based on consolidation of virtual machines

Energy saving based on the consolidation of virtual machines is an effective approach to green computing and can reduce the operation cost of IT systems. However, how to achieve the optimal energy saving based on virtual machine consolidation is essentially a Variable Size Bin Packing Problem (VSBPP), which is a NP problem. A heuristic algorithm named Virtual Machine Consolidation Heuristic (VMCH) based on the typical data center configuration can convert the VSBPP on virtual machine consolidation into a 0-1 Knapsack problem and effectively simplify the algorithm decision process. VMCH consolidates virtual machines into physical servers as few as possible without obviously sacrificing the system performance. The efficiency and effectiveness of VMCH have been verified in a small prototype system.