Handling Persistent States in Process Checkpoint/Restart Mechanisms for HPC Systems

Pierre Riteau 1, * Adrien Lebre 1 Christine Morin 1
* Corresponding author
1 PARIS - Programming distributed parallel systems for large scale numerical simulation
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, ENS Cachan - École normale supérieure - Cachan, Inria Rennes – Bretagne Atlantique
Abstract : Computer clusters are today the reference architecture for high-performance computing. The large number of nodes in these systems induces a high failure rate. This makes fault tolerance mechanisms, e.g. process checkpoint/restart, a required technology to effectively exploit clusters. Most of the process checkpoint/restart implementations only handle volatile states and do not take into account persistent states of applications, which can lead to incoherent application restarts. In this paper, we introduce an efficient persistent state checkpoint/restoration approach that can be interconnected with a large number of file systems. To avoid the performance issues of a stable support relying on synchronous replication mechanisms, we present a failure resilience scheme optimized for such persistent state checkpointing techniques in a distributed environment. First evaluations of our implementation in the kDFS distributed file system show the negligible performance impact of our proposal.
Complete list of metadatas

Cited literature [19 references]  Display  Hide  Download

https://hal.inria.fr/inria-00346745
Contributor : Pierre Riteau <>
Submitted on : Friday, December 12, 2008 - 11:42:44 AM
Last modification on : Tuesday, December 4, 2018 - 9:36:02 AM
Long-term archiving on : Thursday, October 11, 2012 - 1:35:26 PM

File

RR-6765.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00346745, version 1

Citation

Pierre Riteau, Adrien Lebre, Christine Morin. Handling Persistent States in Process Checkpoint/Restart Mechanisms for HPC Systems. [Research Report] RR-6765, INRIA. 2008, pp.16. ⟨inria-00346745⟩

Share

Metrics

Record views

597

Files downloads

229