Efficient resource selection framework to enable cloud for HPC applications

Abstract

Cloud computing provides on demand services with pay-as-you-go manner. Cloud's Infrastructure as Service allows users to use hardware infrastructure (Compute, Network and Storage) and deploy their applications over a computing node as per demand. Users of High Performance Computing applications can take advantages of elasticity of cloud for their compute and data intensive HPC applications without deploying actual physical infrastructure. But virtualization technology degrades the performance of HPC applications. Most of the HPC applications distribute its workload among specified compute resources. To achieve better performance for an application, it is desirable to have homogeneity among the compute resources. Since in cloud environment the virtual machines are created dynamically as per customer needs, it is difficult to find homogeneous environment. In this work we are proposing a method to form cluster of heterogeneous compute resources for HPC base applications on cloud. Proposed method not only considers compute power also bandwidth among the resources. We observed homogeneity among the resources of the best cluster chosen by our method.