Moldable Job Scheduling for HPC as a Service

As cloud computing emerges and gains acceptance, more and more software applications of various domains are transforming into the SaaS model. Recently, the concept of HPC as a Service (HPCaaS) was proposed to bring the traditional high performance computing field into the era of cloud computing. One of its goals aims to allow users to get easier access to HPC facilities and applications. This paper deals with related job submission and scheduling issues to achieve such goal. Traditional HPC users in supercomputing centers are required to specify the amount of processors to use upon job submission. However, we think this requirement might not be necessary for HPCaaS users since most modern parallel jobs are moldable and they usually could not know how to choose an appropriate amount of processors to allow their jobs to finish earlier. Therefore, we propose a moldable job scheduling approach which relieves HPC users’ burden of selecting an appropriate number of processors and can achieve even better system performance than existing job scheduling methods. The experimental results indicate that our approach can achieve up to 75% performance improvement than the traditional rigid processor allocation method and 3% improvement than previous moldable job scheduling methods.