Experiences with the Amoeba distributed operating system
Communications of the ACM
ACM Transactions on Computer Systems (TOCS)
On power-law relationships of the Internet topology
Proceedings of the conference on Applications, technologies, architectures, and protocols for computer communication
OceanStore: an architecture for global-scale persistent storage
ACM SIGPLAN Notices
Time, clocks, and the ordering of events in a distributed system
Communications of the ACM
Bayeux: an architecture for scalable and fault-tolerant wide-area data dissemination
NOSSDAV '01 Proceedings of the 11th international workshop on Network and operating systems support for digital audio and video
On scalable and efficient distributed failure detectors
Proceedings of the twentieth annual ACM symposium on Principles of distributed computing
The Arrow Distributed Directory Protocol
DISC '98 Proceedings of the 12th International Symposium on Distributed Computing
Failure Detection and Consensus in the Crash-Recovery Model
DISC '98 Proceedings of the 12th International Symposium on Distributed Computing
Scalable Fault-Tolerant Aggregation in Large Process Groups
DSN '01 Proceedings of the 2001 International Conference on Dependable Systems and Networks (formerly: FTCS)
Building Peer-to-Peer Systems with Chord, a Distributed Lookup Service
HOTOS '01 Proceedings of the Eighth Workshop on Hot Topics in Operating Systems
Hi-index | 0.00 |
Resource location is a fundamental problem for large-scale distributed applications. This paper discusses the problem from a probabilistic perspective. Contrary to deterministic approaches, which strive to produce a precise outcome, probabilistic approaches may sometimes expose users with incorrect results. The paper formalizes the probabilistic resource-location problem with the notion of probabilistic queries. A probabilistic query has a predicate as parameter and returns a set of sites where the predicate is believed to hold. The query is probabilistic because there are some chances that the predicate does not hold in all, or even in any, of the sites returned. To implement probabilistic queries, we introduce psearch, an epidemic-like algorithm that uses basic concepts of Bayesian statistical inference. Among its properties, PSEARCH is able to adapt itself to new system conditions caused, for example, by failures.