Wide-area route control for online services

Computer networks;Route control

Valancius, Vytautas ; Computing

University:Georgia Institute of Technology

Department:Computing

关键词: Computer networks;    Route control;   

Others  :  https://smartech.gatech.edu/bitstream/1853/43623/1/vytautas_valancius_201205_phd.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

Accelerated by on-demand computing, the number and diversity of theInternet services is increasing. Such online services often have uniquerequirements for the underlying wide-area network: For instance, onlinegaming service might benefit from low delay and jitter paths to client,while online data backup service might benefit from cheaper paths.Unfortunately, today's Internet does not accommodate fine-grained,service-specific wide-area route control. In this dissertation, I achievethe following goals: 1) improve the access to the routes, 2) quantify the benefits of fine-grained route control, and 3) evaluate theefficiency of current payment schemes for the wide-area routes.* Improving access to wide-area route control. Online servicesface significant technological and procedural hurdles in accessing the routes: Each service in need to control the Internetroutes, has to obtain own equipment, Internet numbered resources, andestablish contracts with upstream ISPs. In this dissertation, I proposeand describe implementation and deployment of a secure and scalablesystem which provides on-demand access to the Internet routes. Insetting such as cloud data center, the system can support multipleonline services, providing each service with an illusion of directconnectivity to the neighboring Internet networks, which, for allpractical purposes, allows services to participate fully in theInternet routing.* Quantifying the benefits of fine-grained route control. Evenif online services are presented with wide-area route choice, it is notclear how much tangible benefit such choice provides. Most modern OnlineService Providers (OSP) rely primarily on the content routing toimprove network performance between the clients and the replicas.Inthis dissertation, I quantify the potential benefit the OSPs can gain ifthey perform a joint network and content routing. Among other findings,I find that by performing joint content and network routing, OSPs canachieve 22% larger latency reduction than can be obtained by contentrouting alone.* Modeling and evaluating the efficiency of the current paymentschemes for wide-area routes. Finally, increasing diversity andsophistication of the online services participating in the Internetrouting poses a challenge to payment models used in today'sInternet. Service providers today charge business customers a blendedrate: a single, "average" price for unit of bandwidth, without regardto cost or value of individual customer's flows. In my dissertation, Iset to understand how efficient this payment model is and if moregranular payment model, accounting for the cost and value of differentflows could increase the ISP profit and the consumer surplus. I developan econometric demand and cost model and map three real-world ISP datasets to it. I find that ISPs can indeed improve the economic efficiencywith just a few pricing tiers.