Process Migration for MPI Applications based on Coordinated Checkpoint

Authors:
Jiannong Cao;Yinghao Li;Minyi Guo
Affiliations:
Department of Computing The Hong Kong Polytechnic University Kowloon, Hung Hom, Hong Kong, China PR;Department of Computing The Hong Kong Polytechnic University Kowloon, Hung Hom, Hong Kong, China PR;Department of Computer Software The University of Aizu Aizu-Wakamatsu City
Venue:
ICPADS '05 Proceedings of the 11th International Conference on Parallel and Distributed Systems - Volume 01
Year:
2005

Citing 0
Cited 7

A serialization based approach for strong mobility of shared object

Proceedings of the 5th international symposium on Principles and practice of programming in Java
Proactive process-level live migration in HPC environments

Proceedings of the 2008 ACM/IEEE conference on Supercomputing
Dynamic load balancing for I/O-intensive applications on clusters

ACM Transactions on Storage (TOS)
Adapting grid applications to safety using fault-tolerant methods: Design, implementation and evaluations

Future Generation Computer Systems
A serialisation based approach for processes strong mobility

DAIS'07 Proceedings of the 7th IFIP WG 6.1 international conference on Distributed applications and interoperable systems
A scalable asynchronous replication-based strategy for fault tolerant MPI applications

HiPC'07 Proceedings of the 14th international conference on High performance computing
Proactive process-level live migration and back migration in HPC environments

Journal of Parallel and Distributed Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

A lot of research has been done on faulttolerance for MPI applications, some on checkpoint/restart, and some on network faulttolerance. Process migration, however, has not gained widespread use due to the additional complexity of the requirement that the knowledge about the new location of a migrated process has to be made known to every other process in the application. Here we present a simple yet effective method of process migration based on coordinated checkpointing of MPI applications. Migration is achieved by checkpointing the application, modifying the process location information in the checkpoint files, and restarting the application. Checkpoint/restart and migration are transparent to MPI applications. Performance evaluation results showed that the additional checkpoint/restart capability has little impact on application performance, and the migration method scales well on a large number of nodes.