90 | | OpenFlow at Stanford: Aster*x |
| 95 | ''Demo Participants:'' Saurav Das, Yiannis Yiakoumis [[BR]] |
| 96 | Packet and Circuit switched networks are typically planned, operated and managed separately, leading to substantial management overhead, functionality & resource duplication, and increased Capex & Opex. We propose and demonstrate a converged network, where OpenFlow is used to control both switching technologies in a common way, allowing the service provider maximum flexibility in using the correct mix of technologies while designing and operating their networks. Specifically, we will demonstrate how circuit flow properties (guaranteed bandwidth and delay, no jitter, bandwidth-on-demand) can be used to provide differential treatment to different types of aggregated packet flows - voice, video and web traffic. |
| 97 | |
| 98 | == OpenFlow at Stanford: Expedia == |
| 99 | ''Demo Participant:'' Jad Naous [[BR]] |
| 100 | The Expedient Demo will show how we can reserve a slice across several sites running various aggregates. Expedient will show slice reservations across resources traditionally associated with GENI such as PlanetLab and OpenFlow and across resources outside the GENI sphere such as EC2. |
| 101 | |
| 102 | == OpenFlow at Stanford: Aster*x == |
| 103 | ''Demo Participant:'' Nikhi Handigol [[BR]] |
| 104 | Effective load-balancing systems for services hosted in unstructured networks need to take into account both the congestion of the network and the load on the servers. In this demonstration, we illustrate a comprehensive load-balancing solution that works well for such networks. The system we showcase, called Aster*x, tries to minimize response time by controlling the load on the network and the servers using customized flow routing. The demonstration shows how Aster*x works over a combined network and computation slice spanning multiple campuses across the country. Besides the base behavior, we show the effect of dynamically adding and removing computing resources to the system, increasing the request arrival rate, altering the CPU or network load of each request, and changing load-balancing algorithms. |