Integer programming based heterogeneous CPU–GPU cluster schedulers for SLURM resource manager

作者：

Highlights：

• Two integer programming based heterogeneous CPU–GPU cluster scheduling algorithms are developed for supercomputers.

• SLURM scheduler plug-ins are developed and realistic emulation tests are carried out on large clusters.

• Tests show that our algorithms perform better in terms of utilization, spread and packing over SLURM's backfill plug-in.

摘要

•Two integer programming based heterogeneous CPU–GPU cluster scheduling algorithms are developed for supercomputers.•SLURM scheduler plug-ins are developed and realistic emulation tests are carried out on large clusters.•Tests show that our algorithms perform better in terms of utilization, spread and packing over SLURM's backfill plug-in.

论文关键词：Job scheduling,SLURM,Integer programming

论文评审过程：Received 31 January 2013, Revised 4 October 2013, Accepted 23 January 2014, Available online 25 June 2014.

论文官网地址：https://doi.org/10.1016/j.jcss.2014.06.011