Jupiter Rising: A Decade of Clos Topologies and Centralized Control in Google's Datacenter Network

@article{Singh2015JupiterRA,
  title={Jupiter Rising: A Decade of Clos Topologies and Centralized Control in Google's Datacenter Network},
  author={Arjun Singh and Joon Suan Ong and Amit Agarwal and Glen Anderson and Ashby Armistead and Roy Bannon and Seb Boving and Gaurav Desai and Bob Felderman and Paulie Germano and Anand Kanagala and Hong Liu and Jeff Provost and Jason Simmons and Eiichi Tanda and Jim Wanderer and Urs H{\"o}lzle and Stephen Stuart and Amin Vahdat},
  journal={Proceedings of the 2015 ACM Conference on Special Interest Group on Data Communication},
  year={2015}
}
  • Arjun SinghJ. Ong Amin Vahdat
  • Published 17 August 2015
  • Computer Science
  • Proceedings of the 2015 ACM Conference on Special Interest Group on Data Communication
We present our approach for overcoming the cost, operational complexity, and limited scale endemic to datacenter networks a decade ago. Three themes unify the five generations of datacenter networks detailed in this paper. First, multi-stage Clos topologies built from commodity switch silicon can support cost-effective deployment of building-scale networks. Second, much of the general, but complex, decentralized network routing and management protocols supporting arbitrary deployment scenarios… 

Jupiter evolving: transforming google's datacenter network via optical circuit switches and software-defined networking

It is shown that the combination of traffic and topology engineering on direct-connect fabrics achieves similar throughput as Clos fabrics for the authors' production traffic patterns, and OCS achieves 3x faster fabric reconfiguration compared to pre-evolution ClosFabric.

Stardust: Divide and Conquer in the Data Center Network

Stardust is a distributed solution that attends to the scale limitations of network-switch design, while also offering improved performance and power savings compared with traditional solutions.

Orion: Google's Software-Defined Networking Control Plane

Orion was designed around a modular, micro-service architecture with a central publish-subscribe database to enable a distributed, yet tightlycoupled, software-defined network control system, and today enables Google’s Software-Defined Networks.

Self-Adjusting Ego-Trees Topology for Reconfigurable Datacenter Networks

A novel algorithm, G REEDY E GO T REES, that dynamically changes the network topology and greedily builds ego trees for nodes in the network, where nodes cooperate to help each other, taking into account the global needs of the network.

A Tale of Two Topologies: Exploring Convertible Data Center Network Architectures with Flat-tree

The flat-tree prototype architecture is proposed, which can be implemented as a Clos network and later be converted to approximate random graphs of different sizes, thus achieving both Clos-like implementation simplicity and random-graph-like transmission performance.

Improved Scalability of Demand-Aware Datacenter Topologies With Minimal Route Lengths and Congestion

This paper studies the algorithmic problem of how to design low-degree and hence scalable datacenter networks optimized toward the current traffic they serve and provides a novel network design which provides asymptotically minimal route lengths and congestion.

Expander Datacenters: From Theory to Practice

This paper examines if expanders can be effective for the technology and environments practical in today's data centers, including the use of traditional protocols, at both small and large scale while complying with common practices such as over-subscription.

Sirius: A Flat Datacenter Network with Nanosecond Optical Switching

Sirius, an optically-switched network for datacenters providing the abstraction of a single, high-radix switch that can connect thousands of nodes---racks or servers---in a datacenter while achieving nanosecond-granularity reconfiguration, is proposed.

Improving Fault Tolerance and Performance of Data Center Networks

This thesis is that a key part of the answer is the network's wiring topology: topology has foundational cross-layer effects, and a small amount of intentional asymmetry in the topology can help data center networks meet that challenge.

dcPIM: near-optimal proactive datacenter transport

Datacenter Parallel Iterative Matching (dcPIM) is a proactive data-center transport design that simultaneously achieves near-optimal tail latency for short flows and near-optimal network utilization,
...

References

SHOWING 1-10 OF 41 REFERENCES

A scalable, commodity data center network architecture

This paper shows how to leverage largely commodity Ethernet switches to support the full aggregate bandwidth of clusters consisting of tens of thousands of elements and argues that appropriately architected and interconnected commodity switches may deliver more performance at less cost than available from today's higher-end solutions.

VL2: a scalable and flexible data center network

VL2 is a practical network architecture that scales to support huge data centers with uniform high capacity between servers, performance isolation between services, and Ethernet layer-2 semantics, and is built on a working prototype.

B4: experience with a globally-deployed software defined wan

This work presents the design, implementation, and evaluation of B4, a private WAN connecting Google's data centers across the planet, using OpenFlow to control relatively simple switches built from merchant silicon.

Jellyfish: Networking Data Centers Randomly

Jellyfish is a high-capacity network interconnect which, by adopting a random graph topology, yields itself naturally to incremental expansion, and is more cost-efficient than a fat-tree.

Dcell: a scalable and fault-tolerant network structure for data centers

Results from theoretical analysis, simulations, and experiments show that DCell is a viable interconnection structure for data centers and can be incrementally expanded and a partial DCell provides the same appealing features.

Data Center Switch Architecture in the Age of Merchant Silicon

This paper describes how to save cost and power by repackaging an entire data center network as a distributed multi-stage switch using a fat-tree topology and merchant silicon instead of proprietary ASICs.

BCube: a high performance, server-centric network architecture for modular data centers

Experiments in the testbed demonstrate that BCube is fault tolerant and load balancing and it significantly accelerates representative bandwidth-intensive applications.

Scale-Out Networking in the Data Center

Through the UCSD Triton network architecture, the authors explore issues in managing the network as a single plug-and-play virtualizable fabric scalable to hundreds of thousands of ports and petabits per second of aggregate bandwidth.

KLAT2's Flat Neighborhood Network

This paper describes a novel "Flat Neighborhood" network topology that was designed by a genetic algorithm (GA) and how this new network architecture was derived, how it is used, and how it performs.

OpenFlow: enabling innovation in campus networks

This whitepaper proposes OpenFlow: a way for researchers to run experimental protocols in the networks they use every day. OpenFlow is based on an Ethernet switch, with an internal flow-table, and a