• Summary
• Related Documents
  • Active Bibliography
  • Co-citation
• Version History

Abstract:

Along with the fast progress the interconnection of computer systems makes the need for enterprise-wide distributed solutions grows. These systems have to support hundreds or even thousands of sites located all over the world. The distances between the sites will impose high communication costs on distributed activities thus significantly increasing their response times. This problem can be alleviated through migrating objects to the sites where they are needed. However, migration in such systems cannot be managed globally. Therefore, we present a migration protocol for (autonomous) objects which enables them autonomously to decide whether or not and which site to migrate to, thus making the system self-tuning. We also present two migration strategies objects can use to derive their migration decisions. The first one finds the optimal placement for an object while the second strategy in some cases returns a sub-optimal location but induces lower computation costs. The two strategies are evaluated through benchmarks in a distributed system of autonomous objects. The experiments show a significant decrease in communication costs when migration is employed. 1

Citations