Second Workshop on Evaluating and Architecting System dependabilitY (EASY)

http://www.crhc.uiuc.edu/EASY/

Sunday, 6 October 2002, San Jose, California, U.S.A.

Immediately precedes the Tenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS-X)

Workshop Links

Final program
Tutorial on Designing and Evaluating Dependable Systems (Saturday, October 5, 2002)
Call for papers

Workshop Description

Although end customers of systems care about performance, there is mounting evidence to suggest that they care even more about system availability, manageability and scalability. The increasing complexity of systems implies that no one research community is well-suited to design and evaluate these systems in isolation. Although the computer architecture, fault-tolerant systems, operating systems and database communities each have over twenty years of research experience in their respective domains, there is very little overlap between the communities.

The first EASY workshop was held in 2001, before the co-located ISCA and DSN conferences in Göteborg, Sweden. It served as a unique opportunity to start a dialog between the computer architecture and dependability communities on these topics. Several high-level lessons emerged from the discussions at the workshop:

Researchers are frustrated by the lack of publicly available failure data upon which to calibrate dependability models, and to use as inputs to fault injection techniques.
As all components of a system affect its dependability, discussions of end-to-end dependability benefit from input from the system and application software communities as well as the architecture and dependability communities.
The role of the human (as end user or administrator) is an important (perhaps the most important), but not well-explored, contribution to the failures of systems.
Many systems researchers are unaware of techniques developed by the dependability community.

The goals of the EASY-02 workshop are to expand the dialog to include researchers in the operating systems and programming language communities. The workshop is intended to foster interactive discussion between the ASPLOS communities on the evaluation of computer system dependability and design techniques to increase computer system dependability. We envision a highly interactive and interdisciplinary program, including a combination of refereed paper presentations, invited speakers, and experience reports on cutting-edge industrial techniques for increasing system dependability. In addition, a one-day tutorial on Designing and Evaluating Dependable Systems will be held in conjunction with the workshop.

Workshop Organizers

Kimberly Keeton, Hewlett-Packard Laboratories, kkeeton@hpl.hp.com
Steven Lumetta, University of Illinois at Urbana-Champaign, lumetta@uiuc.edu

Program Committee

George Candea, Stanford University		Thu Nguyen, Rutgers University
Zbigniew Kalbarczyk, University of Illinois		David Oppenheimer, UC Berkeley
Phil Koopman, Carnegie-Mellon University		Dave Patterson, UC Berkeley
Rich Martin, Rutgers University		Bill Sanders, University of Illinois
Subhasish Mitra, Intel and Stanford University		Lisa Spainhower, IBM
Shubu Mukherjee, Intel		Aad van Moorsel, HP Labs
Brendan Murphy, Microsoft