MPLS and GMPLS: Principles, Implementation, and Advanced Concepts

Adrian Farrel, Old Dog Consulting
Zafar Ali, Cisco Systems, Inc.
Mallik Tatipamula, Cisco Systems, Inc.

MPLS and GMPLS are fundamental building blocks for today's evolving networks. Switch and router manufacturers as well as Service Providers are looking to MPLS and GMPLS to leverage new revenue streams, to provide new services, and to extract better performance and capacity from their existing equipment. In order to offer competitively differentiated services, it is necessary to understand the basic principles of the protocols, how they are implemented and deployed, and their advanced capabilities.

This tutorial will start by reviewing the fundamental concepts in MPLS, MPLS-TE and GMPLS, and will highlight the key components of these protocols. It will then describe some aspects of the software implementation of the protocols drawing on the presenters' wide experience of different implementations, switches and deployments to highlight some of the common pitfalls and misconceptions.

The final section of the tutorial will focus on advanced applications and new developments within MPLS-TE and GMPLS. It will cover:

  • Extending the existing protocols to provide end-to-end service across multiple network domains.
  • Protection and recovery schemes that can make the network robust against all manner of failures.
  • Operations and Management features that are essential to ensure the correct functioning of the network.
  • New developments including Error and Alarm Reporting, Layer One VPNs, and Interworking with the ITU-T's ASON architecture.

    Outline

    1. Principles of MPLS and MPLS-TE

  • Labels, label swapping and data plane operations
  • Building block protocols
        o RSVP-TE
        o Routing protocols for Traffic Engineering (IGPs)

    2. Extending the concepts to GMPLS

  • Basic functions
        o Generalized labels
        o Label management
        o Bi-directional services
  • The protocols
        o Extensions to RSVP-TE
        o Routing protocols for GMPLS
        o Link Management Protocol

    3. Fundamental concepts

  • Out of Band Signaling
  • Forwarding Adjacencies
        o LSP Hierarchies
        o LSP Stitching

    4. Implementing and Deploying MPLS-TE and GMPLS

  • What software components do you need?
  • Some common implementation issues
        o Where is my label space?
        o Oh, do I need to be able to manage the product?
        o When is routing not routing?
        o You want how many LSPs?
        o How not to distribute the processing load.
  • - Some real deployment scenarios

    5. Inter-domain Traffic Engineering

  • What is a domain? IGP areas, ASs and beyond.
  • GMPLS Switching Regions
  • Using Forwarding Adjacencies
  • Path Computation Elements

    6. Components of MPLS/ GMPLS High Availability

  • Control Plane Resilience and Non-stop Forwarding
        o Restart/ Recovery Procedures
  • Data Plane Resilience
        o Recovery (Protection and Restoration)
        o MPLS Protection Techniques
           - Fast Reroute (link/ Node Protection)
           - End-to-end protection
        o GMPLS Protection Techniques
           - Link-Level Protection
           - The Notify Message and Repair
           - End-to-end Protection
           - Segment Protection

    7. Operations and Management (A&M)

  • Objectives and Requirements
  • Management Information Bases
  • Connectivity Testing
        o LSP Ping
        o Trace Route
        o LSP Self-Test
        o Using BFD with MPLS
  • OAM Issues for GMPLS
        o Hierarchies and Tunnels
        o The Tunnel Trace Protocol (GTTP)

    8. Future Work

  • Alarm and Error Reporting
        o Crankback
        o Signaling alarm information in GMPLS
  • Layer One VPNs (L1VPN)
  • The ITU-T's ASON Architecture
        o Reference Points
        o The Call and Connection Models
        o Integration with GMPLS
        o Providing the function with GMPLS
  • Point-to-Multipoint Traffic Engineering
        o An overview of the problems space

    Speaker Bios.

    Adrian Farrel runs an Internet protocols consultancy, Old Dog Consulting, and specialises in MPLS and GMPLS. Until recently he was Director of Protocol Development for Movaz Networks, a manufacturer of lambda switching equipment, and before that he worked as Development Manager and Architect for the British software house Data Connection Ltd., where he was responsible for the MPLS and GMPLS products.

    Adrian is currently co-chair of the IETF's Common Control and Measurement Plane (CCAMP) working group that is responsible for the development of the GMPLS protocol suite. He is the author of The Internet and Its Protocols: A Comparative Approach, published by Morgan Kaufmann

    Zafar Ali is a Technical Leader at Cisco Systems, Inc. where he leads software protocol development for Cisco Systems high speed, carrier-class routing business unit. Most recently Zafar has been focused on designing and developing GMPLS protocols and high availability solutions. Prior to joining Cisco, Zafar worked at Nortel Networks and Hughes Network Systems.

    Zafar Ali received his Ph. D. and MS in Electrical and Computer Engineering from Purdue University, West Lafayette, Indiana. He obtained his Bachelor of Engineering in Electrical Engineering from NED University, where he was awarded the University Gold Medal.

    Mallik Tatipamula is a Senior Product Manager at Cisco Systems, Inc. where he is responsible for Emerging technologies in Routing Technologies Group. Mallik has over 14 years of experience in telecom and networking. He closely works with service providers and national research networks around the world in deploying advanced technologies (IPv6, GMPLS) in their next generation networks.