Dynamic Fractional Resource Scheduling for HPC Workloads

Provided by: INRIA
Topic: Big Data
Format: PDF
The authors propose a novel job scheduling approach for homogeneous cluster computing platforms. Its key feature is the use of virtual machine technology for sharing resources in a precise and controlled manner. They justify their approach and propose several job scheduling algorithms. They present results obtained in simulations for synthetic and real-world High Performance Computing (HPC) workloads, in which they compare their proposed algorithms with standard batch scheduling algorithms. They find that their approach widely outperforms batch scheduling.

Find By Topic