Time Services

Motivation

This seminar brought together a significant fraction of the world’s researchers and engineers of network and distributed time services. Until now, there had been little communication between those who implemented time services for computer networks and those exploring new algorithms and analysis techniqes for distributed time services. Our Dagstuhl seminar housed – for the first time – representatives from each of the major approaches to clock synchronization and each of the major providers of real-time synchronization services.

The formal presentations covered new clock synchronization algorithms, methods and actual analysis of these algorithms, and the details of fielded implementations. Participants learned not only the details of the current state of the Internet time protocol (NTP), but also about the history and origins of the algorithms in that implementation.

Some participants could see the practical impacts their fundamental research was having. Others learned about what problems are real and what problems could be safely put aside as academic curiosities. And, implementors learned of new algorithms ripe for trial.

The informal discussions and interactions were at least as important. The meeting forged a community of scientists and engineers with like interests where none had been. Workers in time services now realize the importance of their efforts and have a set of peers with whom to discuss their work. Clock synchronization and time services are a critical part of the infrastructure for any distributed network. By creating a community of workers in this area, the Dagstuhl seminar has itself created an infrastructure to facilitate further advances in the subject.

Schloss Dagstuhl provided an excellent atmosphere for this meeting. We like to thank the staff and the Dagstuhl foundation who made this exchange of research possible.