OnTimeDetect: Offline and Online Network Anomaly Notification Tool for perfSONAR Deployments

"OnTimeDetect" enables consumers of perfSONAR measurements within the data-intensive scientific communities in overcoming their existing limitations of network anomaly detection and notification


Background:


  • Effort to enhance the NLANR/SLAC implementations of a network performance "plateau-detector" algorithm
  • Evaluated anomaly detection performance for both synthetic and actual perfSONAR measurement traces
  • Developed "OnTimeDetect" v0.1 tool prototype from evaluation experiences

Significance:


  • perfSONAR web-service users need automated techniques and intuitive tools to analyze anomalies in real-time and offline manner
  • Network anomaly detector tool should produce minimum false alarms and detect bottleneck events quickly

Sponsor:


DOE