1 / 48

GENI: Global Environment for Network Innovations

GENI: Global Environment for Network Innovations. Larry Peterson Princeton University. March 10, 2006. GENI Initiative. Research Program e.g., FIND program in NeTS solicitation Experimental Facility proposal to MREFC program jointly from CISE and the research community

smackey
Télécharger la présentation

GENI: Global Environment for Network Innovations

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. GENI: Global Environment for Network Innovations Larry PetersonPrinceton University March 10, 2006

  2. GENI Initiative • Research Program • e.g., FIND program in NeTS solicitation • Experimental Facility • proposal to MREFC program • jointly from CISE and the research community • funds (cutting-edge) facility construction, not research • we currently have a “Conceptual Design” document • will eventually require Congressional approval

  3. Planning Group • Tom Anderson, Washington • Dan Blumenthal, UCSB • Dean Casey, Ngenet Research • David Clark, MIT • Deborah Estrin, UCLA • Larry Peterson, Princeton (Chair) • Dipankar Raychaudhuri, Rutgers • Mike Reiter, CMU • Jennifer Rexford, Princeton • Scott Shenker, Berkeley • John Wroclawski, USC/ISI

  4. Project Summary • GENI is an open, large-scale, realistic experimental facility that will revolutionize research in global communication networks. • A central goal of GENI is to change the nature of networked and distributed systems design; to move from a fully empirical to a more rigorously understood design process.

  5. Change the Process GENI Better Internet Intellectual Merit Broader Impact Increased fundamental understanding

  6. Broader Impact • Lead to artifacts that provide value to society • An Internet for tomorrow • more secure, available, manageable, usable • suited for computing in the next decade • Catalyze the distributed digitized world • personal control of your personal life • real time sensing of the planet • Lead to artifacts that provide value to science • Enhanced services that improve the scientific process

  7. Intellectual Merit • GENI will allow us to experimentally answer questions about complex systems, giving us an increased understanding about their dynamics, stability, evolvability, emergent behaviors, and so on. [Science] • GENI will allow us to evaluate alternative architectural structures, and reconcile the contradictory goals a network architecture must meet.[Architecture] • GENI will help us evaluate engineering tradeoffs, and test theories about how different architectual elements might be designed. [Engineering]

  8. Outline • Project Rationale • scientific case • need for a new facility • Facility Design • physical substrate • software management framework • Construction Plan • management organization • development teams

  9. Research Scope • Security and robustness • New network technologies • wireless • optical transport • New computing technologies • sensors, embedded systems, consumer electronics • New distributed applications and services • Service in times of crisis • Network management • Economic incentives

  10. Research Scope (cont) • Theoretical underpinnings • science of design • Architectural limits • overlays versus new core • Analysis and modeling • instrumentation built-in from the start • Cross-boundary opportunities • network, distributed systems, wireless, optical • Interdisciplinary opportunities • privacy, legal, societal, economic issues

  11. Research Landscape • Industry alone will not solve the problem • no incentive for architectural work • Academic “research as usual” is not enough • not being “Internet compatible” is viewed as risky • Architectural research starting to be done • vision papers, preliminary designs, feasibility studies • Strategic choices • clean slate is a process, not an end goal • leverage different aspects of today’s architecture • work across traditional layers (not just “new IP”) • multiple “tech transfer” paths

  12. Need for Infrastructure Goal: Seamless conception-to-deployment process Deployment Analysis Simulation / Emulation Experiment At Scale (models) (code) (results) (measurements)

  13. Existing Tools • Simulators • ns • Emulators • Emulab • WAIL • Wireless Testbeds • ORBIT • Emulab • Wide-Area Testbeds • PlanetLab • RON • X-bone • DETER

  14. Today’s Tools Have Limitations • Simulation based on simple models • topologies, admin policies, workloads, failures… • Emulation (and “in lab” tests) are similarly limited • only as good as the models • Traditional testbeds are targeted • often of limited reach • often with limited programmability • Testbed dilemma • production: real users but incremental change • research: radical change but no real users

  15. PlanetLab • New paradigm for network research infrastructure • Support experimental validation of new services • simultaneously support real users and clean slate designs • allow a thousand flowers to bloom • Provide plausible deployment path • Key ideas • Virtualization • multiple architectures on a shared infrastructure • Programmable • virtually no limit on new designs • Opt-in on a per-user / per-application basis • attract real users • demand drives deployment / adoption

  16. PlanetLab (cont) • 645 machines spanning 310 sites and 35 countries • nodes within a LAN-hop of > 3M users • Supports distributed virtualization • each of 375 network services running in their own slice • Carries real user traffic • 3-4 TB per day / connects to over 1M unique IP address each day

  17. Slices

  18. Slices

  19. Client Server User Opt-in Proxy

  20. Assessment • Borrow ideas from PlanetLab • virtualization (slice abstraction) • support for user opt-in • Improve upon this start • better provisioned • gentler learning curve for users • Incorporate advantages of other testbeds • containment • reproducibility • heterogeneity (more diverse set of components) • Extend in completely new ways • can be extended to incorporate new technologies • support for federation

  21. Summary of Requirements • Architecture/Service Neutrality • substrate is programmable • Shared across multiple experiments • substrate is virtualizable (sliceable) • Attracts real users • low barrier-to-entry for user opt-in • global reach • connected to legacy Internet • Sustainable / Unbounded Potential • federation across multiple countries and communities • Diverse and Extensible • defines interfaces that permit new technologies

  22. Success Scenarios • Caveat: unexpected consequences (uses) • Change the research process • Possible outcomes (broad impact) • create new network services • converge on a single new architecture and deploy it • virtualization wins: plurality of architectures • architectural clarity for current incremental path • Deployment scenarios • transformational pressure • alternative “ISPs” emerge

  23. GENI Design • Key Idea • Slices embedded in a substrate of networking resources • Two central pieces • Physical network substrate • expandable collection of building block components • nodes / links / subnets • Software management framework • knits building blocks together into a coherent facility • embeds slices in the physical substrate

  24. National Fiber Facility

  25. + Programmable Routers

  26. + Clusters at Edge Sites

  27. + Wireless Subnets

  28. + ISP Peers MAE-West MAE-East

  29. Dynamic Configurable Swith Closer Look Sensor Network backbone wavelength backbone switch Customizable Router Internet Edge Site Wireless Subnet

  30. Summary of Substrate • Node Components • edge devices • customizable routers • optical switches • Bandwidth • national fiber facility • tail circuits (including tunnels) • Wireless Subnets • urban 802.11 • wide-area 3G/WiMax • cognitive radio • sensor net • emulation

  31. Management Framework Management Services • name space for users, slices, & components • set of interfaces (“plug in” new components) • support for federation (“plug in” new partners) GMC Substrate Components

  32. Hardware Components Substrate HW Substrate HW Substrate HW

  33. Virtualization Software Virtualization SW Virtualization SW Virtualization SW Substrate HW Substrate HW Substrate HW

  34. CM Virtualization SW Substrate HW Component Manager CM CM Virtualization SW Virtualization SW Substrate HW Substrate HW

  35. slice_spec (object hierarchy) CM CM Virtualization SW Virtualization SW Substrate HW Substrate HW GENI Management Core (GMC) Slice Manager GMC Resource Controller Auditing Archive node control sensor data CM Virtualization SW Substrate HW

  36. Federation GMC GMC . . .

  37. User Front-End(s) GUI Front-End (set of management services) provisioning service file & naming service information plane GMC GMC . . .

  38. Virtualization • Multiple levels possible • different level required by different experiments • different level depending on the technology • in the limit, we may need to slice “physical component array” • Example “base cases” • virtual server (socket interface / overlay tunnels) • virtual router (virtual line card / static circuits) • virtual switch (virtual control interface / dynamic circuits) • virtual AP (virtual MAC / fixed spectrum allocation) • Specialization • the ability to install software in your own virtual-*

  39. Distributed Services • Goals • Complete the GENI management story • Lower the barrier-to-entry for researchers (students) • Example focus areas • Provisioning (Slice Embedder) • Security • Information Plane • Resource Allocation • File & Naming • Topology Discovery • Development Tools • Utility Networking (Internet in a Slice)

  40. Containment & Auditing • Limits placed on slice “reach” • restricted to slice and GENI components • restricted to GENI sites • allowed to compose with other slices • allowed to interoperate with legacy Internet • Limits on resources consumed by slices • cycles, bandwidth, disk, memory • rate of particular packet types, unique addrs per second • Mistakes (and abuse) will still happen • auditing will be essential • network activity slice responsible user(s)

  41. Project Management • GENI Community Consortium (GCC) • Executive Committee (EC) • provides project oversight • appoints Project Director and Project Manager • directs sub-contract selection process • Technical Advisory Board (TAB) • provides technical leadership for the project • establishes a set of working groups • GENI Project Office (GPO) • serves as the “operational arm” of the effort • on the hook to deliver the facility

  42. Executive Committee Project Director Chief Architect Project Manager Technical Advisory Board GENI Project Office Systems Engineering Office (Chief Engineer) Research Coordination WG Finance Office Designs Facilities Architecture WG Legal Affairs Office Backbone Network WG Operations & Planning Office Wireless Subnet WG External Liaison Office Specifications Contracting & Supervision Distributed Services WG Education & Outreach Office Education & Outreach WG Contracted Development & Assembly Teams Management (cont)

  43. Working Groups • Research • usage policy / requirements / instrumentation • Clark (co-chair), Shenker (co-chair) • Architecture • define core modules & interfaces • Peterson (co-chair), Wroclawski (co-chair), Barford, Brassil, Schwab • Backbone • fiber facility / routers & switches / tail circuits / peering • Rexford (chair), Blumenthal, Casey, Lakshman, McKeown, Turner, Zhang

  44. Working Groups (cont) • Wireless • RF technologies / deployment • Raychaudhuri (chair), Bahl, Estrin, Evans, Gerla, Heidemann, Minden, Seshan • Services • edge sites / infrastructure & underlay services • Anderson (chair), Andersen, Kaashoek, Pai, Reiter, Roscoe, Stoica, Vahdat • Education • training / outreach / course development • TBD

  45. Consortium Executive Council & NSF Technical Advisory Board Policy, Evaluations, Authorizations GPO Execution Priorities Documentation Working Groups Requirements Specifications Prelim Budget Recommendations to TAB Feedback Analysis Architecture Action Implementation Ideas Contracts Research Community Expertise Development Teams Receive Contracts Build Network Workflow

  46. Development Teams Provisioning Service File & Naming Service … Information Plane Early Adopters Management Core Team Assembly Team HW SW Edge Devices Customizable Routers Optical Switches Wireless Subnets

  47. More Information www.geni.net

  48. Maturity DeployedFuture Internet Global ExperimentalFacility This chasm is a majorbarrier to realizing the Future Internet Small Scale Testbeds Simulation and Research Prototypes Foundational Research Time Chasm

More Related