Deadlock-Free Message Routing in Multiprocessor Interconnection Networks
IEEE Transactions on Computers
A Protocol for Deadlock-Free Dynamic Reconfiguration in High-Speed Local Area Networks
IEEE Transactions on Parallel and Distributed Systems
Interconnection Networks: An Engineering Approach
Interconnection Networks: An Engineering Approach
A New Methodology to Computer Deadlock-Free Routing Tables for Irregular Networks
CANPC '00 Proceedings of the 4th International Workshop on Network-Based Parallel Computing: Communication, Architecture, and Applications
Fast Dynamic Reconfiguration in Irregular Networks
ICPP '00 Proceedings of the Proceedings of the 2000 International Conference on Parallel Processing
Deadlock-Free Dynamic Reconfiguration Schemes for Increased Network Dependability
IEEE Transactions on Parallel and Distributed Systems
IEEE Transactions on Computers
Fast Routing Computation on InfiniBand Networks
IEEE Transactions on Parallel and Distributed Systems
A proposal for managing ASI fabrics
Journal of Systems Architecture: the EUROMICRO Journal
A model for the development of AS fabric management protocols
Euro-Par'06 Proceedings of the 12th international conference on Parallel Processing
Simple deadlock-free dynamic network reconfiguration
HiPC'04 Proceedings of the 11th international conference on High Performance Computing
Routing for the asi fabric manager
IEEE Communications Magazine
Implementing a Change Assimilation Mechanism for Source Routing Interconnects
Euro-Par '09 Proceedings of the 15th International Euro-Par Conference on Parallel Processing
Efficient network management applied to source routed networks
Parallel Computing
Hi-index | 0.24 |
Advanced Switching Interconnect (ASI) is a high-speed serial interconnect embodied in the Dolphin Express family of interconnect products. In order to support high availability, the ASI specification established a management infrastructure, which is in charge of maintaining network operation after the occurrence of a topological change. When such a change occurs, the management mechanism discovers the new topology, calculates a set of valid routing paths, and distributes them to endpoints within the fabric. Several implementations for such a management mechanism have been proposed that use a centralized approach. These solutions can have negative effects with respect to network service availability. With the aim of eliminating these potential negative effects, this paper proposes a distributed solution for the computation of new paths. The distributed solution is evaluated for management entities with different performance capabilities, and for a range of traffic patterns and load levels. Our results show that the new distributed solution significantly reduces the change assimilation time and the negative impact on the network service when it is compared to a centralized solution.