HPC Applications deployment on distributed heterogeneous computing platforms via OMF, OML and P2PDC

International audienceA new tool and web portal are presented for deployment of High Performance Computing applications on distributed heterogeneous computing platforms. This tool relies on the decentralized environment P2PDC and the OMF and OML multithreaded control, instrumentation and measurement libraries. Deployment on PlanetLab of a numerical simulation application is studied. A first series of computational results is displayed and analyzed. I. INTRODUCTION The domains of parallel and distributed computing are currently undergoing profound mutations. New concepts like peer-to-peer, global computing and cloud computing have recently emerged (see [1]). More recently, heterogeneous technologies have started to be used with success for High Performance Computing (HPC) applications as well (see [2] and [3]). In particular, the combination of distributed and heterogeneous technologies seems to be very promising for HPC applications. These technologies have led to architectures with hundred thousands or millions of cores where application deployment, heterogeneity, power consumption and fault-tolerance are key issues. In this paper, we present the principle of an original solution related to a web portal for peer-to-peer HPC application deployment. The Portal is the combination of the decentralized environment for high performance Peer-to-Peer Distributed Computing, P2PDC, (see [4] and [5]) with tools like OML, OMF and OMF Portal that facilitate the deployment, management of P2PDC applications as well as the retrieval and analysis of results [6]–[9]. We introduce also in this paper a new measurement channel for P2PDC on OML. Section II deals with related work and background material. The motivations of the study are presented in Section III. We detail measurement channel for P2PDC and task deployment in Section IV. Experiments carried out on PlanetLab are presented in Section V. Finally, conclusions and future work are presented in Section VI