Timothy Roscoe, Professor, Systems Group of the Computer Science Department ETH Zurich, the Swiss Federal Institute of Technology
Abstract:
This talk will argue for sophisticated automated reasoning capabilities as a first-class OS service. With such a service, one can delegate many OS policy decisions and calculations to a component which is highly flexible, expressive, and dynamic, providing considerable advantages of hard-coding such functionality in C or scripts. Modern operating systems face several engineering challenges: hardware is increasingly complex, increasingly diverse, and evolving rapidly. This, combined with parallel workloads having complex performance interactions with hardware make it hard to build a simple OS kernel which delivers good performance for a variety of platforms and workloads. We decided to tackle this head-on by building a reasoning engine as a first class service (the “System Knowledge Base”) in the Barrelfish OS, borrowing ideas such fields as knowledge representation, constraint satisfaction, logic programming, and optimization. Doing so was not without problems, but we found it highly convenient in a number of widely different application areas – for example, PCI programming, process coordination, spatial scheduling, and message routing. I’ll discuss several of these, and how the structure of the OS as a whole change when a facility like the SKB is availability.
Bio:
Timothy Roscoe is a Professor in the Systems Group of the Computer Science Department at ETH Zurich, the Swiss Federal Institute of Technology. He received a Ph.D. from the Computer Laboratory of the University of Cambridge, where he was a principal designer and builder of the Nemesis operating system, as well as working on the Wanda microkernel and Pandora multimedia system. After three years building web-based collaboration systems at a startup company in North Carolina, Mothy joined Sprint’s Advanced Technology Lab in Burlingame, California, working on application hosting platforms and networking monitoring. Mothy joined Intel Research at Berkeley in April 2002 as a principal architect of PlanetLab, an open, shared platform for developing and deploying planetary-scale services. In September 2006 he spent four months as a visiting researcher in the Embedded and Real-Time Operating Systems group at National ICT Australia in Sydney, before joining ETH Zurich in January 2007. His current research interests include operating systems for heterogeneous multicore systems and network architectures for ubquitous computing. The first generation of chief privacy officers (CPO) were typically attorneys, charged with the formulation and enforcement of privacy policies. Times have changed. Given the speed and complexity of technology, the privacy policy is necessary but hardly sufficient. Because we live much of our lives in public, both online and offline, the Internet is transforming the anonymity of our cities into the familiarity of small towns. Privacy is deeply ingrained in the technology that manages this personal data. The products and services driving this transformation must consider privacy from the earliest design sessions.