Evaluating Malleable Job Scheduling in HPC Clusters using Real-World Workloads

TL;DR

This paper evaluates the impact of resource-malleable job scheduling in HPC clusters using real workloads, demonstrating significant efficiency improvements even with partial adoption of malleability strategies.

Contribution

It introduces a comprehensive simulation study of malleable job scheduling strategies in HPC, highlighting their benefits and optimal configurations based on real-world workload traces.

Findings

Job turnaround times decrease by up to 67%.

Job wait times reduce by up to 99%.

Node utilization improves significantly, up to 52%.

Abstract

Optimizing resource utilization in high-performance computing (HPC) clusters is essential for maximizing both system efficiency and user satisfaction. However, traditional rigid job scheduling often results in underutilized resources and increased job waiting times. This work evaluates the benefits of resource elasticity, where the job scheduler dynamically adjusts the resource allocation of malleable jobs at runtime. Using real workload traces from the Cori, Eagle, and Theta supercomputers, we simulate varying proportions (0-100%) of malleable jobs with the ElastiSim software. We evaluate five job scheduling strategies, including a novel one that maintains malleable jobs at their preferred resource allocation when possible. Results show that, compared to fully rigid workloads, malleable jobs yield significant improvements across all key metrics. Considering the best-performing scheduling strategy for each supercomputer, job turnaround times decrease by 37-67%, job makespan by 16-65%, job wait times by 73-99%, and node utilization improves by 5-52%. Although improvements vary, gains remain substantial even at 20% malleable jobs. This work highlights important correlations between workload characteristics (e.g., job runtimes and node requirements), malleability proportions, and scheduling strategies. These findings confirm the potential of malleability to address inefficiencies in current HPC practices and demonstrate that even limited adoption can provide substantial advantages, encouraging its integration into HPC resource management.