One issue is the need to search for patterns in event histories, either after execution has completed or on-line, as an event history is received. Since event histories are frequently large, e.g., millions of events, tools for finding unusual occurences are clearly important and on-line algorithms can be used in system-monitoring contexts to discover that a system is no longer behaving as desired.
A second issue is scaling our techniques to very large numbers of interacting processes (or threads or other concurrent entitities). Standard techniques for representing partial orders require space proportional to the product of the number of events and the number of concurrent processes. For thousands of processes and millions of events, novel techniques are required to achieve good efficiency without requiring absurd amounts of storage.
Finding time to produce a web version of my publication list does not seem to have achieved a high enough priority. My current best suggestion is to check DBLP, which is a good general solution for finding papers of people in computer science.
The research described above is taking
place in the Shoshin research group.
The group receives funding from a variety of sources.
The research on debugging of distributed applications has been funded
by the Natural Sciences and Engineering Research Council and by IBM.
Because of an inability to say ``no'' at strategic moments, I have served on numerous committees at various levels within the University, including both the University Senate and Board of Governors (and the Executive Committee of each). I have also, at various times, chaired the Advisory Committee on Appointments and the Curriculum Committee of the Department (now School) of Computer Science and the Mathematics Faculty Committee on Student Appeals. From July 1987 to June 1990 and again from July 1998 to June 1999, I served as Associate Chair, Undergraduate Studies, in the Computer Science Department. From January 2001 to December 2007, I served as Associate Dean (Undergraduate Studies) for the Faculty of Mathematics. As Associate Dean, I chaired several committees, including the Undergraduate Affairs Committee and the Standings and Promotions Committee of the Mathematics Faculty, and served on numerous other committees at the Faculty and University levels. As Director of the School, effective July 2010, I serve on numerous committees of the School and chair both the Executive Committee and the Tenure and Promotion Committee.
I have served on various program committees and have been both Program Chair and General Chair for the Symposium on Reliable Distributed Systems. I am currently a member of the editorial board of IEEE Transactions on Dependable and Secure Computing and previously served on the editorial board of IEEE Transactions on Knowledge and Data Engineering. I was Conference Coordinator for the International Conference on Dependable Systems and Networks in 2006 in Philadelphia and was persuaded to perform the same role for the 2010 conference in Chicago.
David R. Cheriton School of Computer Science University of Waterloo 200 University Avenue West Waterloo, Ontario, Canada N2L 3G1
Office: DC 2321
Phone: +1 (519) 888-4080
Fax: +1 (519) 885-1208
E-mail: (at top of page)
www: http://www.shoshin.uwaterloo.ca/~dtaylor