The notification based approach to implementing failure detectors in distributed systems
InfoScale '06 Proceedings of the 1st international conference on Scalable information systems
Evaluation of the QoS of crash-recovery failure detection
Proceedings of the 2007 ACM symposium on Applied computing
Efficient algorithms for fault tolerant mobile agent execution
International Journal of High Performance Computing and Networking
IEEE Journal on Selected Areas in Communications - Special issue on wireless and pervasive communications for healthcare
Implementation and performance evaluation of an adaptable failure detector in iSCSI
APPT'07 Proceedings of the 7th international conference on Advanced parallel processing technologies
Parallel algorithms for fault-tolerant mobile agent execution
ICA3PP'05 Proceedings of the 6th international conference on Algorithms and Architectures for Parallel Processing
QoS self-configuring failure detectors for distributed systems
DAIS'10 Proceedings of the 10th IFIP WG 6.1 international conference on Distributed Applications and Interoperable Systems
Adaptare: Supporting automatic and dependable adaptation in dynamic environments
ACM Transactions on Autonomous and Adaptive Systems (TAAS)
Timeout-based adaptive consensus: improving performance through adaptation
Proceedings of the 27th Annual ACM Symposium on Applied Computing
Hi-index | 0.00 |
Unreliable failure detectors have been an importantabstraction to build dependable distributed applicationsover asynchronous distributed systems subject to faults.Their implementations are commonly based on timeoutsto ensure algorithm termination. However, for systemsbuilt on the Internet, it is hard to estimate this time valuedue to traffic variations. Thus, different types ofpredictors have been used to model this behavior andmake predictions of delays. In order to increase thequality of service (QoS), self-tuned failure detectorsdynamically adapt their timeouts to the communicationdelay behavior added of a safety margin. In this paper,we evaluate the QoS of a failure detector for differentcombinations of communication delay predictors andsafety margins. As the results show, to improve the QoS,one must consider the relation between the pairpredictor/margin, instead of each one separately.Furthermore, performance and accuracy requirementsshould be considered for a suitable relationship.