On-the-fly adaptive routing for dragonfly interconnection networks
Adaptive deadlock-free routing mechanisms are required to handle variable traffic patterns in dragonfly networks. However, distance-based deadlock avoidance mechanisms typically employed in Dragonflies increase the router cost and complexity as a function of the maximum allowed path length. This pap...
| Autores: | , , , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2015 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/82765 |
| Acceso en línea: | https://hdl.handle.net/2117/82765 https://dx.doi.org/10.1007/s11227-014-1357-9 |
| Access Level: | acceso abierto |
| Palabra clave: | Routing (Computer network management) Multiprocessors Computational complexity Interconnection network Dragonfly network OFAR Adaptive routing Deadlock avoidance Encaminadors (Xarxes d'ordinadors) Multiprocessadors Complexitat computacional Àrees temàtiques de la UPC::Informàtica::Arquitectura de computadors |
| Sumario: | Adaptive deadlock-free routing mechanisms are required to handle variable traffic patterns in dragonfly networks. However, distance-based deadlock avoidance mechanisms typically employed in Dragonflies increase the router cost and complexity as a function of the maximum allowed path length. This paper presents on-the-fly adaptive routing (OFAR), a routing/flow-control scheme that decouples the routing and the deadlock avoidance mechanisms. OFAR allows for in-transit adaptive routing with local and global misrouting, without imposing dependencies between virtual channels, and relying on a deadlock-free escape subnetwork to avoid deadlock. This model lowers latency, increases throughput, and adapts faster to transient traffic than previously proposed mechanisms. The low capacity of the escape subnetwork makes it prone to congestion. A simple congestion management mechanism based on injection restriction is considered to avoid such issues. Finally, reliability is considered by introducing mechanisms to find multiple edge-disjoint Hamiltonian rings embedded on the dragonfly, allowing to use multiple escape subnetworks. |
|---|