EECS Publication
An Evaluation of User-Level Failure Mitigation Support in MPI
Wesley Bland, Aurelien Bouteiller, Thomas Herault, Joshua Hursey, George Bosilca, and Jack J. Dongarra
As the scale of computing platforms becomes increasingly extreme, the requirements for application fault tolerance are becoming so as well. Techniques to address this problem by improving the resilience of algorithms have been developed; but they currently receive no support from the programming model, and without such support, they are bound to fail. This paper discusses the failure-free overhead and recovery impact aspects of the User-Level Failure Mitigation proposal presented in the MPI Forum. Experiments demonstrate that fault-aware MPI has little or no impact on performance for a range of applications, and produces satisfactory recovery times when there are failures.
Published 2012-05-31 04:00:00 as ut-cs-12-695 (ID:7)