Quantifying the Reliability of Proven SPIDER Group Membership Service Guarantees

Authors:
Elizabeth Latronico;Paul Miner;Philip Koopman
Affiliations:
Carnegie Mellon University, Pittsburgh, PA;NASA Langley Research Center, Hampton, VA;Carnegie Mellon University, Pittsburgh, PA
Venue:
DSN '04 Proceedings of the 2004 International Conference on Dependable Systems and Networks
Year:
2004

Citing 0
Cited 3

Reliable communication for DuST networks

ETFA'09 Proceedings of the 14th IEEE international conference on Emerging technologies & factory automation
Modeling the reliability of a group membership protocol for dual-scheduled time division multiple access networks

Computer Standards & Interfaces
A reliability evaluation of a group membership protocol

SAFECOMP'07 Proceedings of the 26th international conference on Computer Safety, Reliability, and Security

Quantified Score

Hi-index	0.00

Visualization

Abstract

For safety-critical systems, it is essential to quantify thereliability of the assumptions that underlie provenguarantees. We investigate the reliability of theassumptions of the SPIDER group membership service withrespect to transient and permanent faults. Modeling12,600 possible system configurations, the probability thatSPIDER's Maximum Fault Assumption will not hold for anhour mission varies from less likely than 10{-11} to more likelythan 10{-3}. In most cases examined, a transient faulttolerance strategy was superior to the permanent faulttolerance strategy previously in use for the range oftransient fault arrival rates expected in aerospace systems.Reliability of the Maximum Fault Assumption (upon whichthe proofs are based) differs greatly when subjected toasymmetric, symmetric, and benign faults. This case studydemonstrates the benefits of quantifying the reliability ofassumptions for proven properties.