Highly available distributed services and fault-tolerant distributed garbage collection
PODC '86 Proceedings of the fifth annual ACM symposium on Principles of distributed computing
Timestamp-Based Orphan Elimination
IEEE Transactions on Software Engineering
Distributed, object-based programming systems
ACM Computing Surveys (CSUR)
On the correctness of orphan management algorithms
Journal of the ACM (JACM)
Atomic data abstractions in a distributed collaborative editing system
POPL '86 Proceedings of the 13th ACM SIGACT-SIGPLAN symposium on Principles of programming languages
| Hi-index | 0.00 |
In a distributed system, an activity running at one node can request another node to perform some service. This request results in an activity being created at the latter node to perform the requested service. The former node may then crash, destroying the activity that requested the service, but leaving behind the activity performing the service. Such surviving activities are known as orphans. Orphans are undesirable since they waste resources and can view inconsistent data. This thesis presents an algorithm that detects and exterminates orphans before they can view inconsistent data. The algorithm has the desirable property that no non-orphans are mistakenly identified as orphans and exterminated. An underlying premise of the algorithm is that orphan detection and extermination should delay normal computation as little as possible. The algorithm works by piggybacking information concerning orphans on various messages that flow about the system. The algorithm piggybacks an impractical amount of data on messages. The main contribution of this thesis is the development of a method called deadlining. This method works in conjunction with the algorithm to detect orphans before they view inconsistent data, while substantially reducing the amount of data piggybacked on messages. An analytic model is used to predict the actual performance of deadlining. This report is a minor revision of a thesis of the same title submitted to the Department of Electrical Engineering and Computer Science on May 25, 1984 in partial fulfillment of the requirements for the Degree of Master of Science.