Barak Gafni

Dragonfly+: Low Cost Topology for Scaling Datacenters

Dragonfly topology was introduced by Kim et al. [1] aiming to decrease the cost and diameter of the network. The topology divides routers into groups connected by longlinks. Each group strives to implement high-radix virtual router, connected by a completely-connected topology. In this paper, we propose an extended Dragonfly+ networkin which routers inside the group are connected in Clos-liketopology. Dragonfly+ is superior to conventional Dragonfly due to the significantly larger number of hosts which it is able to support. In addition, Dragonfly+ supports similar or better bi-sectional bandwidth for various traffic patterns, and requires smaller number of buffers to avoid credit loop deadlocks in lossless networks. Moreover, we introduce a novel Fully Progressive Adaptive Routing algorithm with remote congestion notifications. To support our proposal we present analytical analysis and simulations.