Research On Virtual Data Center Multi-domain Embedding Algorithm

Data center has been the focus of attention as a cost-effective infrastructure used to store a large amount of data and host service applications. Large computations link Amazon, Google, Facebook, and Yahoo! routinely use data centers for storage, Web search, and large-scale computation. The emergence of the data center virtualization technology solved the limitation problem of traditional data center, such as low server utilization, high operation cost and a lack of isolation. However, the data center virtualization technology also brought a variety of new challenges at the same time. Virtual data center embedding is one of the key challenges. In cloud computing environment, infrastructure providers have physical resources and provide these resources for service providers. A request for resources to a data center can be abstracted as a virtual data center(VDC) request. A VDC consist of the virtual machines(VMs) connected through virtual swiches, routers and links with guaranteed bandwidth. In this paper, we study the problem of VDC embedding across multiple data centers.First, we study the problem of reliable VDC embedding across multiple data centers, such that the total bandwidth consumption in the inter-data center backbone network is minimized, while satisfying the reliability requirement of each VDC request. We propose a reliable virtual data center multi-domain embedding algorithm. Simulation results show that the proposed algorithm performs better in terms of lowering physical resource consumption and VDC request blocking ratio compared with existing solution.Then, we also study the problem of dynamic VDC embedding across multiple data centers, such thar the total operation cost(domain cost and inter-domain cost) is minimized. We use the VM migration to reconfigure dynamic VDC, thus reducing the total operation cost. And through the VM migration, we consolidate the VDCs locally, thus reducing the number of active servers and energy consumpution. We proposed a dynamic VDC embedding across multiple data centers heuristic algorithm. The simulation results show that this algorithm performs better in terms of lowering total operation cost and energy consumpution compared with existing solution.