Workload manager is most relevant to improve the performance of High Performance Computing (HPC) Cluster. Aiming at the recent issues of workload management in the HPC cluster within data center is so complex that it causes the whole data center load imbalanced. This paper introduces an open source solution for middleware layer which is having high efficiency and abilities to gain high resource utilization. There are various scheduling approaches available, some approaches are open source and some of them are commercial. Schedulers are useful for achieve 99% of utilization of available resources and also increase the usability, performance and acceptance of HPC Cluster. In this paper we describe the use of openlava scheduler to manage workload of HPC Cluster and compose the conclusions respecting performance and throughput of the systems by showing the scheme prospective for the second establishment for HPC Cluster with MPICH2 integration, improved up to 57%.
[1]
Deva Bodas,et al.
Simple Power-Aware Scheduler to Limit Power Consumption by HPC System within a Budget
,
2014,
2014 Energy Efficient Supercomputing Workshop.
[2]
Wolfgang Gentzsch,et al.
Sun Grid Engine: towards creating a compute power grid
,
2001,
Proceedings First IEEE/ACM International Symposium on Cluster Computing and the Grid.
[3]
Krzysztof Rzadca,et al.
A Scheduler-Level Incentive Mechanism for Energy Efficiency in HPC
,
2015,
2015 15th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing.
[4]
Seren Soner,et al.
Integer Programming Based Heterogeneous CPU-GPU Cluster Scheduler for SLURM Resource Manager
,
2012,
2012 IEEE 14th International Conference on High Performance Computing and Communication & 2012 IEEE 9th International Conference on Embedded Software and Systems.