Software overhead in messaging layers: where does the time go?
ASPLOS VI Proceedings of the sixth international conference on Architectural support for programming languages and operating systems
Fast parallel algorithms for short-range molecular dynamics
Journal of Computational Physics
Algorithm-Based Fault Tolerance for FFT Networks
IEEE Transactions on Computers
An Efficient Algorithm-Based Concurrent Error Detection for FFT Networks
IEEE Transactions on Computers
Experimental Study of Internet Stability and Backbone Failures
FTCS '99 Proceedings of the Twenty-Ninth Annual International Symposium on Fault-Tolerant Computing
Tolerating Network Failures in System Area Networks
ICPP '02 Proceedings of the 2002 International Conference on Parallel Processing
High performance MPI design using unreliable datagram for ultra-scale InfiniBand clusters
Proceedings of the 21st annual international conference on Supercomputing
Algorithm-Based Fault Tolerance for Matrix Operations
IEEE Transactions on Computers
Characterization of failures in an operational IP backbone network
IEEE/ACM Transactions on Networking (TON)
Zero-copy protocol for MPI using infiniband unreliable datagram
CLUSTER '07 Proceedings of the 2007 IEEE International Conference on Cluster Computing
Fault Tolerant Parallel FFT Using Parallel Failure Recovery
ICCSA '09 Proceedings of the 2009 International Conference on Computational Science and Its Applications
International Journal of High Performance Computing Applications
Dlib-ml: A Machine Learning Toolkit
The Journal of Machine Learning Research
California fault lines: understanding the causes and impact of network failures
Proceedings of the ACM SIGCOMM 2010 conference
IPDPS'06 Proceedings of the 20th international conference on Parallel and distributed processing
Understanding network failures in data centers: measurement, analysis, and implications
Proceedings of the ACM SIGCOMM 2011 conference
Algorithm-based fault tolerance for dense matrix factorizations
Proceedings of the 17th ACM SIGPLAN symposium on Principles and Practice of Parallel Programming
| Hi-index | 0.00 |
With the increasing number of components in HPC systems, transient faults will become commonplace. Today, network transient faults, such as lost or corrupted network packets, are addressed by middleware libraries at the cost of high memory usage and packet retransmissions. These costs, however, can be eliminated using application-level fault tolerance. In this paper, we propose recovery methods for transient network faults at the application level. These methods reconstruct missing or corrupted data via interpolation. We derive a realistic fault model using network fault rates from a production HPC cluster and use it to demonstrate the effectiveness of our reconstruction methods in an FFT kernel. We found that the normalized root-mean-square error for FFT computations can be as low as 0.1% and, thus, demonstrates that network faults can be handled at the application level with low perturbation in applications that have smoothness in their computed data.