Read Me
SLAML — Managing Large-Scale Systems via the Analysis of System Logs and the Application of Machine Learning Techniques
Program chairs: Peter Bodik, Marc Casas, Greg Bronevetsky
08:50-09:00
Introduction
09:00-10:00
Session 1
Practical Experiences with Chronics Discovery in Large Telecommunications Systems
Soila P. Kavulya (Carnegie Mellon University), Kaustubh Joshi, Matti Hiltunen, Scott Daniels (AT&T Labs), Rajeev Gandhi, Priya Narasimhan (Carnegie Mellon University)
BLR-D: Applying Bilinear Logistic Regression to Factored Diagnosis Problems
Sumit Basu (Microsoft Research), John Dunagan (Microsoft), Kevin Duh (NTT Labs), Kiran-Kumar Muniswamy-Reddy (Harvard University)
10:30-12:00
Session 2
Mining Temporal Invariants from Partially Ordered Logs
Ivan Beschastnikh, Yuriy Brun, Michael D. Ernst, Arvind Krishnamurthy, Thomas E. Anderson (University of Washington)
Adaptive Event Prediction Strategy with Dynamic Time Window for Large-Scale HPC Systems
Ana Gainaru (UIUC & UPB Bucharest), Franck Cappello (INRIA & UIUC), Joshi Fullop (UIUC), Stefan Trausan-Matu (UPB Bucharest), Bill Kramer (UIUC)
Mining large distributed log data in near real time
Stefan Weigert (TU Dresden), Matti Hiltunen (AT&T Labs), Christof Fetzer (TU Dresden)
13:30-15:00
Session 3
Web Analytics and the Art of Data Summarization
Archana Ganapathi, Steve Zhang (Splunk Inc)
Panel: Assessing and improving the quality of program logs
Panelists: Ari Rabkin (UC Berkeley), Ding Yuan (UIUC/UC San Diego), Wei Xu (Google), Steve Zhang (Splunk)
15:30-17:00
Session 4
PAL: Propagation-aware Anomaly Localization for Cloud Hosted Distributed Applications
Hiep Nguyen, Yongmin Tan, Xiaohui Gu (North Carolina State University)
Discussion