Abstract
Data center networks (DCNs) connect hundreds and thousands of computers and, as a result of the exponential growth in their number of nodes, the design of scalable (compact) routing schemes plays a pivotal role in the optimal operation of the DCN. Traditional trends in the design of DCN architectures have led to solutions, where routing schemes and network topologies are interdependent, i.e., specialized routing schemes. Unlike these, we propose a routing scheme that is compact and generic, i.e., independent of the DCN topology, the word-metric-based greedy routing. In this scheme, each node is assigned to a coordinate (or label) in the word-metric space (WMS) of an algebraic group and then nodes forward packets to the closest neighbor to the destination in this WMS. We evaluate our scheme and compare it with other routing schemes in several topologies. We prove that the memory space requirements in nodes and the forwarding decision time grow sub-linearly (with respect to n, the number of nodes) in all of these topologies. The scheme finds the shortest paths in topologies based on Cayley graphs and trees (e.g. Fat tree), while in the rest of topologies, the length of any path is stretched by a factor that grows logarithmically (with respect to n). Moreover, the simulation results show that many of the paths remain far below this upper bound.
Highlights
D ATA Center Networks (DCN) connect hundreds of thousands of computers to store, manage, and disseminate information
We evaluate our scheme and compared it to other routing schemes in different topologies presented in Section II: the specialized routing schemes BCube Routing Protocol (BRP), DCell Fault-tolerant Routing protocol (DFR), two-level routing scheme and the routing schemes based on permutation-sort; the topologies BCube, DCell, Fat-tree, Jellyfish, Xpander, Slim fly, and five Cayley Graphs (CG), Hypercube, Bubble-sort, Star, Transposition and Butterfly
The scheme is based on an embedding process, where nodes are assigned to coordinates in the Word-Metric Space (WMS) of an algebraic group, and on a greedy forwarding strategy, where nodes forward packets to the closest neighbor to the destination in this WMS
Summary
D ATA Center Networks (DCN) connect hundreds of thousands of computers to store, manage, and disseminate information. This is the case of the traditional routing scheme K-shortest path [11] that has been proposed to work on Jellyfish [12], Xpander [13] and Slim fly [14] topologies. This scheme can not be considered compact because of its high memory space requirements. The information required to compute distances in the WMS is encoded using a Finite State Automaton (FSA) By proving that this FSA can be stored and processed in an efficient way, we guarantee that our scheme achieves scalable routing tables, node labels and forwarding decision time.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have