Fault injection experiments on Cielo characterize fault-error-failure sequences in the Gemini network and nodes to identify unrecoverable failure scenarios for HPC resilience.
Analysis of gemini interconnect recovery mechanisms: Methods and observations,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Understanding Fault Scenarios and Impacts through Fault Injection Experiments in Cielo
Fault injection experiments on Cielo characterize fault-error-failure sequences in the Gemini network and nodes to identify unrecoverable failure scenarios for HPC resilience.