Designing an Adaptive Application-Level Checkpoint Management System for Malleable MPI Applications

TL;DR

This paper introduces iCheck, an adaptive checkpoint management system designed for malleable MPI applications, enhancing fault tolerance and data redistribution by leveraging dynamic resource management in high-performance computing.

Contribution

The paper presents a novel adaptive checkpointing system that integrates with malleable resource management to improve fault tolerance and data redistribution in HPC applications.

Findings

iCheck improves checkpointing efficiency in malleable MPI applications.

The system enhances fault tolerance and data redistribution during resource changes.

Experimental results demonstrate better resource utilization and application performance.

Abstract

Dynamic resource management opens up numerous opportunities in High Performance Computing. It improves the system-level services as well as application performance. Checkpointing can also be deemed as a system-level service and can reap the benefits offered by dynamism. A checkpointing system can have better resource availability by integrating with a malleable resource management system. In addition to fault tolerance, the checkpointing system can cater to the data redistribution demand of malleable applications during resource change. Therefore, we propose iCheck, an adaptive application-level checkpoint management system that can efficiently utilize the system and application level dynamism to provide better checkpointing and data redistribution services to applications.