Strategy-proof resource pricing in federated systems

There is growing interest in federated systems where globally distributed and commoditized resources are traded, such as in peer-to-peer computing, grids, and cloud computing. In this context, an important issue in the allocation of shared resources is that users are rational and attempt to maximize their self-interest. In this thesis, we investigate the use of resource pricing with financial incentives to allocate shared resources when users are rational. Using mechanism design, we propose a strategy-proof resource pricing scheme, with provable economic properties and computational efficiency. To improve scalability, we propose a distributed auction scheme that leverages on a peer-to-peer overlay network for resource lookups and distributes pricing by resource type. In a federated cloud composed of four Amazon EC2 regions, we show using traces that rational users with dynamic pricing increased welfare over EC2 spot pricing. As an application, we have prototyped SkyBoxz, a federated cloud platform for managing virtual machines across multiple public and private clouds.

Author dc:creator

• MARIAN MIHAILESCU

• federated clouds, resource allocation, dynamic resource pricing, strategy-proof, mechanism design, rationality