To be agile and cost effective, data centers should allow dynamic resource allocation across large server pools. In particular, the data center network should enable any server to be assigned to any service. To meet these goals, we present VL2, 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. VL2 uses (1) flat addressing to allow service instances to be placed anywhere in the network, (2) Valiant Load Balancing to spread traffic uniformly across network paths, and (3) end-system based address resolution to scale to large server pools, without introducing complexity to the network control plane. VL2's design is driven by detailed measurements of traffic and fault data from a large operational cloud service provider. VL2's implementation leverages proven network technologies, already available at low cost in high-speed hardware implementations, to build a scalable and reliable network architecture. As a result, VL2 networks can be deployed today, and we have built a working prototype. We evaluate the merits of the VL2 design using measurement, analysis, and experiments. Our VL2 prototype shuffles 2.7 TB of data among 75 servers in 395 seconds - sustaining a rate that is 94% of the maximum possible.

/pdf/vl2-a-scalable-and-flexible-data-center-network-dkbtxdift1.pdf

VL2: a scalable and flexible data center network

We explore the nature of traffic in data centers, designed to support the mining of massive data sets. We instrument the servers to collect socket-level logs, with negligible performance impact. In a 1500 server operational cluster, we thus amass roughly a petabyte of measurements over two months, from which we obtain and report detailed views of traffic and congestion conditions and patterns. We further consider whether traffic matrices in the cluster might be obtained instead via tomographic inference from coarser-grained counter data.

/pdf/the-nature-of-data-center-traffic-measurements-analysis-1db3il1pin.pdf

The nature of data center traffic: measurements & analysis

Current intra-domain Traffic Engineering (TE) relies on offline methods, which use long term average traffic demands. It cannot react to realtime traffic changes caused by BGP reroutes, diurnal traffic variations, attacks, or flash crowds. Further, current TE deals with network failures by pre-computing alternative routings for a limited set of failures. It may fail to prevent congestion when unanticipated or combination failures occur, even though the network has enough capacity to handle the failure.This paper presents TeXCP, an online distributed TE protocol that balances load in realtime, responding to actual traffic demands and failures. TeXCP uses multiple paths to deliver demands from an ingress to an egress router, adaptively moving traffic from over-utilized to under-utilized paths. These adaptations are carefully designed such that, though done independently by each edge router based on local information, they balance load in the whole network without oscillations. We model TeXCP, prove the stability of the model, and show that it is easy to implement. Our extensive simulations show that, for the same traffic demands, a network using TeXCP supports the same utilization and failure resilience as a network that uses traditional offline TE, but with half or third the capacity.

/pdf/walking-the-tightrope-responsive-yet-stable-traffic-1q735fzs5r.pdf

Walking the tightrope: responsive yet stable traffic engineering

IP networks today require massive effort to configure and manage. Ethernet is vastly simpler to manage, but does not scale beyond small local area networks. This paper describes an alternative network architecture called SEATTLE that achieves the best of both worlds: The scalability of IP combined with the simplicity of Ethernet. SEATTLE provides plug-and-play functionality via flat addressing, while ensuring scalability and efficiency through shortest-path routing and hash-based resolution of host information. In contrast to previous work on identity-based routing, SEATTLE ensures path predictability and stability, and simplifies network management. We performed a simulation study driven by real-world traffic traces and network topologies, and used Emulab to evaluate a prototype of our design based on the Click and XORP open-source routing platforms. Our experiments show that SEATTLE efficiently handles network failures and host mobility, while reducing control overhead and state requirements by roughly two orders of magnitude compared with Ethernet bridging.

/pdf/floodless-in-seattle-a-scalable-ethernet-architecture-for-4l1tnb9otx.pdf

Floodless in seattle: a scalable ethernet architecture for large enterprises

We present Dionysus, a system for fast, consistent network updates in software-defined networks. Dionysus encodes as a graph the consistency-related dependencies among updates at individual switches, and it then dynamically schedules these updates based on runtime differences in the update speeds of different switches. This dynamic scheduling is the key to its speed; prior update methods are slow because they pre-determine a schedule, which does not adapt to runtime conditions. Testbed experiments and data-driven simulations show that Dionysus improves the median update speed by 53--88% in both wide area and data center networks compared to prior methods.

/pdf/dynamic-scheduling-of-network-updates-1fmucv9mvs.pdf

Dynamic scheduling of network updates

This thesis proposes two-phase routing as a capacity efficient and robust strategy for handling highly variable traffic. The scheme allows preconfiguration of the network such that all traffic patterns permissible within the network's natural ingress-egress capacity constraints can be routed with bandwidth guarantees without requiring detection of traffic changes in real-time or reconfiguring the network in response to it. The scheme routes traffic in two phases---traffic entering the network is sent from the source to a set of intermediate nodes in predetermined split ratios that depend on the intermediate nodes, and then from the intermediate nodes to the final destination. The scheme has the desirable properties of supporting static optical layer provisioning in IP-over-Optical networks and indirection in specialized service overlay models unlike previous approaches---like direct source-destination path routing---for handling variable traffic. 
This thesis represents the first comprehensive study, problem formulation, and algorithm design for many aspects of two-phase routing. Our contributions can be grouped into three broad parts. First, we consider the problems of minimum cost network design and maximum throughput network routing for the scheme. We give a simple solution for minimum cost network design. For maximum throughput network routing, we design linear programming based and combinatorial algorithms. We show how the algorithms can handle a, total cost constraint for maximum throughput two-phase routing. This can be used to solve the link capacitated version of minimum cost two-phase routing. We establish theoretical bounds on the resource requirements of two-phase routing under throughput and cost models with respect to the optimal scheme that is allowed to make the routing dynamically dependent on the current traffic matrix. We also generalize the traffic split ratios to depend not only on the intermediate nodes but also on source and destination of traffic and solve the corresponding optimization problems. 
Second, we consider making two-phase routing resilient to network failures. Two-phase routing in IP-over-Optical networks can be protected against router node failures through redistribution of traffic split ratio for the failed router node to other intermediate nodes. We propose two different schemes for provisioning the optical layer to handle router node failures. We develop linear programming formulations for both schemes and a fast combinatorial algorithm for the second scheme so as to maximize network throughput. 
Third, we consider the application of two-phase routing to multi-hop Wireless Mesh Networks (WMNs). These networks have recently been of much research interest due to their lowered need for wired infrastructure support and due to envisaged new applications like community wireless networks. We extend our optimization framework for maximum throughput two-phase routing in wired networks to handle routing and scheduling constraints that are peculiar to WMNs and arise from the requirement to handle radio transmit/receive diversity and the phenomenon of wireless link interference. 
We evaluate various aspects of two-phase routing on actual ISP topologies using the developed algorithms. For the WMN application, we use randomly generated WMN topologies for the evaluations. (Copies available exclusively from MIT Libraries, Rm. 14-0551, Cambridge, MA 02139-4307. Ph. 617-253-5668; Fax 617-253-1690.) (Abstract shortened by UMI.)

/pdf/efficient-and-robust-routing-of-highly-variable-traffic-2x1bej9y9m.pdf

Sundipta Sengupta

Papers

Efficient and robust routing of highly variable traffic