Mach, AFS Papers Given SIGOPS Hall of Fame Awards
Two papers by Carnegie Mellon researchers were recognized as being among the most important contributions in operating systems research. The recognition, called the SIGOPS Hall of Fame Award, is announced annually at one of the two major symposia in OS research, either SOSP or OSDI, to the most influential peer-reviewed research papers that appeared at least ten years ago.
This year, seminal papers on the Mach operating system (Machine-independent virtual memory management for paged uniprocessor and multiprocessor architectures, by Rick Rashid, Avi Tevanian, Michael Young, David Golub, Bob Baron, David Black, William Bolosky, and Jonathan Chew) and AFS (Scale and performance in a distributed file system, John Howard, Michael Kazar, Sherri Kazar, David Nichols, M. Satyanarayanan, Robert Sidebotham, and Michael West) received the award. The seminal systems contributions of Multics, Grapevine, and DISCO were also recognized.
This is the third award for CMU work. In 2006, my own paper co-authored with George Necula — the first published paper on “proof-carrying code” — also received the SIGOPS Hall of Fame Award.
It’s really great to see CMU research work in systems recognized like this. It’s even better to be reminded of some of the really great work from a decade or more ago. Some of the early work on Unix, RPC, authentication logics, and more have influenced almost every computing system in use today.
It’s natural to wonder, of course, about the state of health of systems research today. Over the past decade there has been an undercurrent of concern about systems research becoming too incremental. Indeed, I remember my own stint on the SOSP program committee some years ago when there was considerable discussion about how to avoid selecting only incremental (albeit high quality) work and accommodate riskier, less-well worked-out ideas.
While I think that systems research did in fact (and perhaps still has) problems with too many “small result” papers, my sense today is that systems research is in a renaissance. A look through the past year or two of SOSP and OSDI proceedings reveals a plethora of new, possbly groundbreaking work, spanning cloud computing, secure operating systems, parallel programming mechanisms, storage, and much, much more. Furthermore, the huge challenges facing the field in parallel computing (i.e., the “multicore challenge”) and data-intensive computing (”big data”) are starting to generate more and more research contributions. While it is impossible to predict which of today’s papers will win SIGOPS Hall of Fame awards in 2018 and beyond, it seems likely to me that at least some of them will be every bit as influential as today’s award winners. Definitely something to look forward to.
In the mean time, congratulations to the old Mach and AFS groups!
Peter Lee @ December 16, 2008