Jellyfish: Networking Data Centers Randomly

Abstract

Industry experience indicates that the ability to incrementally expand data centers is essential. However, existing high-bandwidth network designs have rigid structure that interferes with incremental expansion. We present Jellyfish, a high-capacity network interconnect, which, by adopting a random graph topology, yields itself naturally to incremental expansion. Somewhat surprisingly, Jellyfish is more cost-efficient than a fat-tree: A Jellyfish interconnect built using the same equipment as a fat-tree, supports as many as 25% more servers at full capacity at the scale of a few thousand nodes, and this advantage improves with scale. Jellyfish also allows great flexibility in building networks with different degrees of oversubscription. However, Jellyfish's unstructured design brings new challenges in routing, physical layout, and wiring. We describe and evaluate approaches that resolve these challenges effectively, indicating that Jellyfish could be deployed in today's data centers.

14 pages, 12 figures

Cited by in corpus (25)

• Hamiltonian Properties of DCell Networks
• Reliability and Survivability Analysis of Data Center Network Topologies
• Exploiting Path Diversity in Datacenters using MPTCP-aware SDN
• A Survey and Evaluation of Data Center Network Topologies
• Expanding across time to deliver bandwidth efficiency and low latency
• Accurate and Fast Federated Learning via IID and Communication-Aware Grouping
• An In-Depth Analysis of the Slingshot Interconnect
• Improving the Resilience of Fast Failover Routing: TREE (Tree Routing to Extend Edge disjoint paths)
• ReNets: Toward Statically Optimal Self-Adjusting Networks
• METTEOR: Robust Multi-Traffic Topology Engineering for Commercial Data Center Networks
• Online Dynamic B-Matching With Applications to Reconfigurable Datacenter Networks
• A Flat and Scalable Data Center Network Topology Based on De Bruijn Graphs
• COUDER: Robust Topology Engineering for Optical Circuit Switched Data Center Networks
• The Effect of Network Topology on Credit Network Throughput
• MIPS: Instance Placement for Stream Processing Systems based on Monte Carlo Tree Search
• POTUS: Predictive Online Tuple Scheduling for Data Stream Processing Systems
• Slim Fly: A Cost Effective Low-Diameter Network Topology
• Solving Large-Scale Granular Resource Allocation Problems Efficiently with POP
• Online VNF Chaining and Predictive Scheduling: Optimality and Trade-offs
• Predictive Switch-Controller Association and Control Devolution for SDN Systems
• Multi-Path Routing on the Jellyfish Networks
• De-Fragmenting the Cloud
• On Efficient Data Transfers Across Geographically Dispersed Datacenters
• Managing Recurrent Virtual Network Updates in Multi-Tenant Datacenters: A System Perspective
• DeltaPath: dataflow-based high-performance incremental routing