{{{ #!html

GENI Engineering Conference 8

July 20, 2010 UCSD, San Diego, CA

}}} [[PageOutline(2, Table of Contents)]] '''Live demonstrations, posters, and presentations at GEC8 highlight results from spiral 2 projects. See project descriptions, posters and presentations here. ''' '''Please enter your project demonstration information here. Attach posters and presentations here or on linked pages.''' ---- == BGP Multiplexer == ''Demo Participants:'' Vytautas Valancius, Nick Feamster, Jennifer Rexford, and Akihiro Nakao [[BR]] BGP-Mux is a system that enables wide-area route control for networks inside GENI. BGP-Mux allows scarce resources, such as IP prefixes and AS numbers, to be shared among experimental virtual networks. This demo will showcase the automated management of such resources. == Cognitive Radio == ''Demo Participants:'' Ivan Seskar, Peter Wolniansky, Dirk Grunwald [[BR]] Demo cognitive radio platform doing spectrum sensing == Data-Intensive Cloud Control for GENI: Processing Weather Radar Data on Amazon EC2 == ''Demo Participant:'' Emmanuel Cecchet [[BR]] This demo will show how Orca can be used to manage Amazon Web Services (AWS) resources including servers (EC2) and storage (S3). Data from a weather radar will be uploaded to S3 and processed by an EC2 server to produce images. A limited budget of AWS resources will be allocated and the processing will stop as soon as the budget limit has been reached. This will showcase the !DiCloud? proxy aggregate manager for Orca. == GENI Pathlets == ''Demo Participants:'' Mark Berman, Brighten Godfrey [[BR]] Source-controlled routing is a promising architectural approach to improve reliability, efficiency, and overall flexibility of the network. We will demonstrate a streaming media application weathering "storms" on the network via source-controlled adaptive routing, running on top of pathlet routing, a flexible interdomain routing and forwarding substrate. == GENI !NetServ == ''Demo Participants:'' Jan Janak, Mark Berman [[BR]] We present !NetServ, an extensible platform for deploying in-network services. !NetServ-enabled network nodes provide a common execution environment, where network services implemented as modules can be dynamically installed and removed. Our demonstration shows how !NetServ can help solve some of the issues faced by content and real-time service providers. We show how !NetServ-enabled nodes can help content providers disseminate content and how providers of real-time services, such as voice and video calls, can use !NetServ-enabled nodes to make their systems more scalable and reliable. GENI allows us to deploy !NetServ?-enabled nodes in the network. == GENI VIOLIN == ''Demo Participants:'' Pradeep Padala (DOCOMO USA Labs), Ardalan Kangarlou (Purdue University), Robert Lantz (DOCOMO USA Labs) [[BR]] We will show how VIOLIN (http://friends.cs.purdue.edu/projects/violin/index.shtml) can be used for suspending and resuming GENI experiments. This project is part of the GENI-alpha demos planned for GEC9. Our demo will show starting an Emulab experiment, suspending it, and resuming it from a checkpoint. The suspend/resume feature can be used for various purposes including fault tolerance, resource management, and debugging. In GEC8 demo, we showed how to create a VIOLIN with 4 VMs on 4 different nodes (on Utah Emulab), and ran a distributed Mandelbrot application written using MPI. The application is snapshotted, and when a failure happens, it is restored from a previously stored consistent snapshot. Please see the demo video (to be uploaded shortly) for more details. [http://groups.geni.net/geni/attachment/wiki/GEC8DemoSummary/GENI-VIOLIN-GEC8-poster.pdf GEC8 Poster] [[BR]] [http://groups.geni.net/geni/attachment/wiki/GEC8DemoSummary/GENI-VIOLIN-GEC8-slides.pdf GEC8 Slides] GMOC Meta-Operations Center == GpENI == == iGENI Distributed Network Research Infrastructure == ''Demo Participant:'' Joe Mambretti [[BR]] The iGENI dynamic provisioning demonstration will showcase capabilities for large scale (national) multiple domain dynamic L1/L2 path provisioing involving multiple sites, using specialized signaling and implementation techniques. == Kansei Wireless Sensor Networks == ''Demo Participant:''Wenjie Zeng [[BR]] The KanseiGenie team will demonstrate its capability of data plane stitching, in addition to its demonstrated control plane stitching capability, on two federated sensor network testbeds, the Kansei testbed located the Ohio State University and the NetEye testbed located at the Wayne State University. Near real-time round-trip delay between the two sites will be measured, collected, and visualized. K-GENI Operational Linkage between GENI and ETRI/KISTI-Korea for International Federation == LAMP Leveraging and Abstracting Measurements with perfSONAR == ''Demo Participants:'' Ezra Kissel,Martin Swany, Guilherme Fernandes [[BR]] perfSONAR is an established framework that enables network performance information to be gathered and exchanged in a multi-domain, federated environment. Leveraging and Abstracting Measurements with perfSONAR (LAMP) is a project to create an instrumentation and measurement (I&M) system, based on perfSONAR, for use by experimenters on GENI. This demo will present the current status of the LAMP I&M system by demonstrating its use on ProtoGENI. Demonstrate monitoring of a running experiment. == LEARN Programmable Measurements over Texas-based Research Network == ''Demo Participant:'' Deniz Gurkan [[BR]] Milestone S2.d: Complete initial integration of the LEARN network into the ORCA control framework (GENI Cluster D), to enable GENI researchers to utilize the LEARN network for L2 (VLAN) transport between a limited number of sites, e.g., University of Houston and Rice University. (7/20/10, GEC8) [[BR]] ''STATUS: Pending hardware deployment. Cisco 3400 switches are purchased.'' Milestone S2.k: Based upon results of Data Plane Measurements project, collaborate with GPO and other projects in Cluster D, to establish a list of measurement handlers that are needed for commercial transport and measurement equipment that will be used in GENI. (7/20/10, GEC8) [[BR]] ''STATUS: Submitted a recommendation list for all measurement handlers for GENI substrate.'' [http://groups.geni.net/geni/attachment/wiki/LEARN/GENI_MS2K_LEARN_Jul20.pdf milestone S2.k report] [http://groups.geni.net/geni/attachment/wiki/GEC8DemoSummary/GENI%20GEC8%20LEARN%20poster.pdf poster] == netKarma GENI Provenance Registry == ''Demo Participant:'' Chris Small [[BR]] NetKarma: a provenance registry to capture the metadeta describing the experiment conditions, time ordering, and relationships within the experiment and across the experiment and infrastructure layer == OpenFlow at Clemson University: OF Management and Inter-Campus Access Control == ''Demo Participant:'' Kuang-Ching Wang [[BR]] The demo will consists of three parts: 1. Campus Ethernet deployment and user management process. 1. Inter-campus access control demo jointly with GT OpenFlow team. 1. OpenFlow operation on wireless mesh network. == OpenFlow at Georgia Tech: RESONANCE Access Control == ''Demo Participants:'' Russell Clark, Nick Feamster [[BR]] We will demonstrate a network policy management system including network access control for wired and wireless users. * RESONANCE - an OpenFlow Network Access Control for campus users * Inter-campus access control demo jointly with Georgia Tech and Clemson University == OpenFlow at Indiana University == ''Demo Participant:'' Chris Small [[BR]] Demonstration of the deployment of OpenFlow in the IU campus network and connectivity to backbone and other campus networks. The demonstration will also highlight tools used to monitor and manage OpenFlow networks == OpenFlow at Stanford: Packet and Circuit Convergence == ''Demo Participants:'' Saurav Das, Yiannis Yiakoumis [[BR]] 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. == Expedient == ''Demo Participant:'' Jad Naous [[BR]] We started by asking the question, "Is there a simpler control framework that has a milder learning curve and that is less complex and brittle than today's GENI control frameworks?". Expedient provides a pluggable control framework that leaves the specification of APIs to the resource developers themselves. It simplifies authorization and authentication, and allows resource providers to choose to delegate these operations to Expedient if they want to. It allows developers to design rich user interfaces by providing access to an object-oriented database that describes resources and aggregates. The demo shows federation of PlanetLab and OpenFlow resources at Stanford with similar resources at BBN, and the ability to create a slice that spans these resources through a user interface plugin. Links: * [http://yuba.stanford.edu/~jnaous/expedient/ Expedient Website] * [http://yuba.stanford.edu/~jnaous/expedient/gec8-poster.pdf Poster] == OpenFlow at Stanford: Aster*x == ''Demo Participant:'' Nikhil Handigol, Srini Seetharaman [[BR]] 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. Collaborating campuses: UWash, Indiana, Princeton and BBN == OpenFlow at University of Washington: Aster*x == ''Demo Participant:''Vjekoslav Brajkovic [[BR]] [http://groups.geni.net/geni/ticket/449 Aster*x: Load-Balancing OpenFlow Demo] == OpenFlow at University of Wisconsin: Mobile Application Migration in Enterprise Metworks == ''Demo Participants:'' Aaron Gember, Aditya Akella [[BR]] Smartphones are becoming an important element of enterprise computing, but new applications are putting increased demands on limited hardware resources. We propose an offloading framework that takes advantage of local resources in enterprise networks to simultaneously address the resource, performance, and security demands of mobile applications. We demonstrate the application migration component of our in-progress offloading system prototype. An application is launched on a Google Android smartphone, then migrated to an idle desktop (laptop) over an OpenFlow network. We also demonstrate offloading applications to a remote cloud by offloading a running application from the smartphone to the University of Wisconsin-Madison campus OpenFlow network. == OpenFlow at University of Wisconsin: Packet Level Redundancy Elimination (RE) on Network Elements == ''Demo Participants:'' Ashok Anand, Aditya Akella, Athula Balachandran, Srinivasan Seshan [[BR]] Internet traffic is increasing at a tremendous rate. We propose a framework of deploying packet level redundancy elimination (RE) on network elements (e.g., routers) ) and show that it can serve as an effective way to improve network efficiency without expensive upgrading of network links to higher bandwidths. In the demo, we first show the current scenario of Internet video transmission in terms of link loads.We then show the impact of RE, and how it reduces link loads by eliminating redundant content. We have RE enabled software routers, which are configured by a Nox controller. == OKGems GENI-Federated Cyber-Physical System with Multi-Modalities == ''Demo Participant:'' Andy Li [[BR]] Three major features are: 1. Create slices to do experiments on a subset of remote sensor networks (8 * 8 lab sensor grid, totally 64 nodes); 1. Create slices to do limited experiments on hybrid sensor networks in lab and in the farmland; 1. Remote controllable robots (iRobot, Surveyor, and Robotic Arms) with remote videos via cameras (possibly robot-carried cameras, controllable too). == OnTimeMeasure Centralized and Distributed Measurement Orchestration Software == ''Demo Participants:'' Prasad Calyam, kzhu@osc.edu [[BR]] We will demonstrate the capabilities of our OnTimeMeasure software that was recently released to early experimenters [http://groups.geni.net/geni/attachment/wiki/GEC8DemoSummary/GEC8-OnTimeMeasureGENI-Poster.pdf?format=raw (Demo poster with screenshots)]. Specifically, we will show how early experimenters can setup and use the OnTimeMeasure software in their own ProtoGENI slices. We will also show the newly developed/enhanced features (e.g., researcher web-portal, user slice registration, distributed on-demand measurement, centralized on-going measurement, handling monitoring objectives, user customizable dashboard, admin interface). Lastly, we will show our early efforts in integrating the OnTimeMeasure software with other GENI projects such as Instrumentation Tools [http://groups.geni.net/geni/attachment/wiki/GEC8DemoSummary/OnTimeMeasure-InsTools-GEC8.pdf?format=raw (Integration screenshots)]. == ORCA/BEN: A Prototype GENI Control Plane (ORCA) and a Metro-Scale Optical Testbed (BEN) == ''Demo Participants:'' Ilia Baldine (RENCI), Yufeng Xin (RENCI), Anirban Mandal (RENCI), Namgon Kim(GIST), Anjing Wang (NCSU), Michael Wang (Columbia), Balagangadhar Bathula (Columbia) [[BR]] We demonstrate: * Capabilities of ORCA control framework to create cross-site slices using resources from multiple distributed resource providers (BEN, Duke University, NLR, StarLight facility, UMass Amherst). The slice is stitched from VMs on Duke campus, multiple VLAN segments and another VM at UMass Amherst that collects sensor feeds from ViSE testbed. In this demo ORCA actors (authorities, brokers and experiment controller) exchange NDL-OWL formatted substrate descriptions and requests to form the slice. * Capabilities of ORCA to provision OpenFlow slices: ORCA provisions OpenFlow slice on OF testbed located in Korea and connects them to VMs running in BEN network (North Carolina). The experiment demonstrates an OF controller capable of performing VLAN translation. * We show a [http://groups.geni.net/geni/attachment/wiki/GEC8DemoSummary/GEC8_IMF_Poster.pdf poster] and demo for GENI-IMF project demonstrating integration of physical layer measurement capability and programmatic in-stack (protocol stack) consumption of such measurement data for cross-layer control. Physical layer measurements from BEN network are relayed via pubsub IMF system to an application running experimental SILO modular networking protocol stack. The SILO cross-layer controller consumes the measurement data and is able to dynamically control the SOA (Silicon Optical Amplifier) on the link connecting the two parts of the application. This dynamic control allows the application to actively compensate for the loss of power on the link introduced by a programmable attenuator. == PrimoGENI Developing GENI Aggregates for Real-Time Large-Scale Network Simulation == ''Demo Participants:'' Miguel Erazo, Jason Liu [[BR]] We are going to show the new release of PrimoGENI which includes the creation of real-time simulation/emulation experiments on top of the Emulab/Protogeni environment, the use of OpenVZ for container creation that will host real applications, release of our new IDE to launch experiments. == S3MONITOR Scalable, Extensible, and Safe Monitoring of GENI == ''Demo Participant:'' Puneet Sharma [[BR]] This demo will show the Sensing Information Manager for the S3Monitor project. The manager acts as portal for researchers to add measurement requests and view archived data. == SPP Internet Scale Overlay Hosting == ''Demo Participant:'' John D. !DeHart [[BR]] We will demonstrate a fastpath and a slowpath on our installed SPP nodes with control and monitoring. == UMLPEN Programmable Edge Node == ''Demo Participant:'' Yan Luo [[BR]] A demonstration of VLAN connectivity to FIU, and federation with FIU PrimoGENI aggregate, and measurement capability of UMLPEN. We will establish an experiment of hybrid emulation and simulation using ProtoGENI control framework. == ViSE Sensor Virtualization and Slivering in an Outdoor Wide-Area Wireless Sensor/Actuator Network Testbed == ''Demo Participant:'' David Irwin [[BR]] ViSE's demonstration will show NOWCasting---short-term real-time forecasts 1 to 30 minutes in the future---visualizations of weather from ViSE sensors. The demonstration will utilize resources from the ViSE testbed. == WiMAX == ''Demo Participants:'' Gautam Bhanage, Ivan Seskar, Manu Gosain [[BR]] * Demo of an application including clients at multiple WiMAX sites, plus a WiFI connected client at demo site. * WiMAX sites include: Rutgers; BBN; and, if possible, NYU Polytechnic * In addition, there will be a research-oriented mobility demo conducted at Rutgers WINLAB site, that will demonstrate all of the mobility features that have been added to ORBIT * This demo covers WIMAX milestone S2.h and ORBIT milestone S2.g * This demo is an early version of the demo being planned for GEC9