Self-organizing publish/subscribe
DSM '05 Proceedings of the 2nd international doctoral symposium on Middleware
Modeling the communication costs of content-based routing: the case of subscription forwarding
Proceedings of the 2007 inaugural international conference on Distributed event-based systems
Modeling event-driven service-oriented systems using the palladio component model
Proceedings of the 1st international workshop on Quality of service-oriented software systems
Stochastic Analysis of Hierarchical Publish/Subscribe Systems
Euro-Par '09 Proceedings of the 15th International Euro-Par Conference on Parallel Processing
Content-Based Routing in Networks with Time-Fluctuating Request Rates
NET-COOP '09 Proceedings of the 3rd Euro-NF Conference on Network Control and Optimization
Reconfiguring self-stabilizing publish/subscribe systems
DSOM'06 Proceedings of the 17th IFIP/IEEE international conference on Distributed Systems: operations and management
Performance modeling and analysis of message-oriented event-driven systems
Software and Systems Modeling (SoSyM)
| Hi-index | 0.00 |
Publish/subscribe is becoming increasingly popular as it provides means for decoupled communication. One important issue for increasing the success of publish/subscribe middleware is to make them fault tolerant. Classical faulttolerance mechanisms apply redundancy to mask certain faults. However, if a fault cannot be masked, it is not guaranteed that the system ever returns to normal operation. In contrast to that, self-stabilizing systems recover from arbitrary transient faults provided that faults do not continue to occur until the system is stable again. However, while the system stabilizes, it may not exhibit the desired behavior. In this paper, we present the first comprehensive analysis of publish/subscribe systems including self-stabilization, giving an alternative to extensive simulations. The analysis is based on continuous time birth-death Markov Chains and investigates the characteristics of publish/subscribe systems in equilibrium. We give closed analytical solutions for the sizes of routing tables, for the overhead required to keep the routing tables up-to-date, and for the leasing overhead required for self-stabilization. To judge the efficiency of selfstabilizing routing, we compare it to flooding which is the na篓ive implementation of a self-stabilizing publish/subscribe system.