Deduplication potential of HPC applications' checkpoints

Kaiser, Jurgen; Gad, Ramy; Suss, Tim; Padua, Federico; Nagel, Lars; Brinkmann, Andre

- No file added yet -

Deduplication potential of HPC applications' checkpoints

conference contribution

posted on 2018-02-06, 11:34 authored by Jurgen Kaiser, Ramy Gad, Tim Suss, Federico Padua, Lars NagelLars Nagel, Andre Brinkmann

© 2016 IEEE. HPC systems contain an increasing number of components, decreasing the mean time between failures. Checkpoint mechanisms help to overcome such failures for long-running applications. A viable solution to remove the resulting pressure from the I/O backends is to deduplicate the checkpoints. However, there is little knowledge about the potential to save I/Os for HPC applications by using deduplication within the checkpointing process. In this paper, we perform a broad study about the deduplication behavior of HPC application checkpointing and its impact on system design.

Funding

This work was partially supported by the German Ministry for Education and Research (BMBF) under project grant 01IH13004 (Project FAST) and the Intel Parallel Computing Centers (IPCC) project.

History

School

Science

Department

Computer Science

Published in

Proceedings - IEEE International Conference on Cluster Computing, ICCC

Pages

413 - 422

Citation

KAISER, J., ... et al., 2016. Deduplication potential of HPC applications' checkpoints. Presented at the 2016 IEEE International Conference on Cluster Computing (CLUSTER), Taipei, Taiwan, 12-16th Sept, pp. 413-422.

Publisher

Version

AM (Accepted Manuscript)

Publication date

2016

Notes

Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.