Evaluating Online Global Recovery with Fenix Using Application-Aware In-Memory Checkpointing Techniques

Keyphrases

• Checkpointing 100%
• Application-aware 100%
• In-memory 100%
• Global Recovery 100%
• Fenix 100%
• Exascale 66%
• Checkpoint-restart 66%
• Failure Time 33%
• Production System 33%
• Number of Components 33%
• Application-oriented 33%
• Recovery Framework 33%
• Stable Storage 33%
• Node Failure 33%
• Cray 33%
• Number of Cores 33%
• MPI Applications 33%
• Fault-tolerance Techniques 33%
• Exascale Systems 33%
• Process Failure 33%
• In-memory Checkpoint 33%
• Mini-application 33%
• User-level Failure Mitigation 33%

Computer Science

• Checkpointing 100%
• Checkpoint Restart 100%
• fault-tolerance 50%
• Tolerance Technique 50%
• Exascale System 50%
• Production System 50%
• Stable Storage 50%