English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

As we have entered Exascale computing, the faults in high-performance systemsare expected to increase considerably. To compensate for a higher failure rate,the standard checkpoint/restart technique would need to create checkpoints at amuch higher frequency resulting in an excessive amount of overhead which wouldnot be sustainable for many scientific applications. Replication allows forfast recovery from failures by simply dropping the failed processes and usingtheir replicas to continue the regular operation of the application.  In this paper, we have implemented PartRePer-MPI, a novel fault-tolerant MPIlibrary that adopts partial replication of some of the launched MPI processesin order to provide resilience from failures. The novelty of our work is thatit combines both fault tolerance, due to the use of the User Level FailureMitigation (ULFM) framework in the Open MPI library, and high performance, dueto the use of communication protocols in the native MPI library that isgenerally fine-tuned for specific HPC platforms. We have implemented efficientand parallel communication strategies with computational and replica processes,and our library can seamlessly provide fault tolerance support to an existingMPI application. Our experiments using seven NAS Parallel Benchmarks and twoscientific applications show that the failure-free overheads in PartRePer-MPIwhen compared to the baseline MVAPICH2, are only up to 6.4% for the NASparallel benchmarks and up to 9.7% for the scientific applications.

PartRePer-MPI: Combining Fault Tolerance and Performance for MPI Applications