Empirical Analysis

This folder contains a representative sample of HPC (Cray) data used for reliability study of node failures.

If you are using these logs, please cite the following paper in which it was first released:

"Systemic Assessment of Node Failures in HPC Production Platforms" in IEEE International Parallel and Distributed Processing Symposium (IPDPS), 2021.

@inproceedings{DBLP:conf/ipps/Das0R21,
  author    = {Anwesha Das and
               Frank Mueller and
               Barry Rountree},
  title     = {Systemic Assessment of Node Failures in {HPC} Production Platforms},
  booktitle = {35th {IEEE} International Parallel and Distributed Processing Symposium,
               {IPDPS}},
  pages     = {267--276},
  publisher = {{IEEE}},
  year      = {2021} 
}