Section: New Results

Optimal Cooperative Checkpointing for Shared High-Performance Computing Platforms

Participants : Dorian Arnold [Emory University, Atlanta, GA, USA] , George Bosilca [ICL, University of Tennessee Knoxville, USA] , Aurélien Bouteiller [ICL, University of Tennessee Knoxville, USA] , Jack Dongarra [ICL, University of Tennessee Knoxville, USA] , Kurt Ferreira [Center for Computing Research, Sandia National Laboratory, USA] , Thomas Hérault [ICL, University of Tennessee Knoxville, USA] , Yves Robert.

In high-performance computing environments, input/output (I/O) from various sources often contend for scarce available bandwidth. Adding to the I/O operations inherent to the failure-free execution of an application, I/O from checkpoint/restart (CR) operations (used to ensure progress in the presence of failures) places an additional burden as it increases I/O contention, leading to degraded performance. In this work, we consider a cooperative scheduling policy that optimizes the overall performance of concurrently executing CR-based applications which share valuable I/O resources. First, we provide a theoretical model and then derive a set of necessary constraints needed to minimize the global waste on the platform. Our results demonstrate that the optimal checkpoint interval as defined by Young/Daly, while providing a sensible metric for a single application, is not sufficient to optimally address resource contention at the platform scale. We therefore show that combining optimal checkpointing periods with I/O scheduling strategies can provide a significant improvement on the overall application performance, thereby maximizing platform throughput. Overall, these results provide critical analysis and direct guidance on checkpointing large-scale workloads in the presence of competing I/O while minimizing the impact on application performance.

This work is available as a research report and has been submitted to a conference.