Checkpointing Speculative Distributed Shared Memory

Abstract. This paper describes a checkpointing mechanism destined for Distributed Shared Memory (DSM) systems with speculative prefetching. Speculation is a general technique involving prediction of the future of a computation, namely accesses to shared objects unavailable on the accessing node (read faults). Thanks to such predictions objects can be fetched before the actual access operation is performed, resulting, at least potentially, in considerable performance improvement. The proposed mechanism is based on independent incremental checkpointing integrated with a coherence protocol introducing little overhead. It ensures the consistency of checkpoints, allowing fast recovery from failures.

show abstract

“…Therefore, we investigate other possibilities, discussed in depth in [3]. Here we present one which solves all the problems described above.…”

Section: P2(x)mentioning

confidence: 99%

Speculation Meets Checkpointing

Danilecki

Szychowiak

2006

Computational Science – ICCS 2006

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the previous work on reliable speculative DSM systems ( [4], [3]), it has been shown that a naive implementation of the speculation and, more specifically, prefetching may reduce DSM efficiency by introducing false dependencies, which in turn increase the number of unnecessary checkpoints. Since prefetches are seen by the object owner as read accesses, they create dependencies.…”

Section: Previous Workmentioning

confidence: 99%

Using Speculative Push for Unnecessary Checkpoint Creation Avoidance

Danilecki

Szychowiak

2006

Distributed Applications and Interoperable Systems

Self Cite

View full text Add to dashboard Cite

This paper discusses a way of incorporating speculation techniques into Distributed Shared Memory (DSM) systems with checkpointing mechanism without creating unnecessary checkpoints. Speculation is a general technique involving prediction of the future of a computation, namely accesses to shared objects unavailable on the accessing node (read faults). Thanks to such predictions objects can be pushed to requesting nodes before the actual access operation is performed, resulting, at least potentially, in a considerable performance improvement. This mechanism is a foundation for the proposed SpecCkpt protocol based on independent checkpointing integrated with a coherence protocol for a given consistency model introducing little overhead. It ensures the consistency of checkpoints, at the same time allowing a fast recovery from failures.

show abstract

Checkpointing Speculative Distributed Shared Memory

Cited by 2 publications

References 6 publications

Speculation Meets Checkpointing

Speculation Meets Checkpointing

Using Speculative Push for Unnecessary Checkpoint Creation Avoidance

Contact Info

Product

Resources

About