High-Performance Switch/Router Design:
Theory and Practice

The rapid evolution of high-speed switches, routers, and cross-connects over the past few years has lead to several advances in the theory and practice of high-performance switch architectures. With the rapid convergence of packet and circuit-switched services, it is crucial for designers and systems engineers to grasp and appreciate this evolution and new design paradigms. Whether it is lookups and packet classification or software design in the control plane, a number of new best practices have shaped how modern systems are architected.

Understanding the major developments of the last decade in this field, and having insights into future trends is crucial for building advanced software, chips, systems, and networks.

With this goal, this workshop covers:

• Switch architectures & fabrics: features, properties, applicability, practical realizations
• Analysis of data path processing: through a router, cross-connect, hybrid packet/TDM switch
• Comparative evaluation: Cisco Catalyst 6K family (large enterprise) versus the Juniper M40/160 (metro/core) and Gibson T640 (core); discussion of the key architectural aspects of edge/metro boxes, such as Unisphere/Juniper's ERX and Foundry's BigIron family.
• Forwarding, lookups, packet classification: methods, hardware realizations, and performance issues
• Scheduler design: theoretical limits, practical algorithms and implementations
• Output scheduling: fair queueing algorithms, practical considerations, QoS guarantees
• Considerations in modern router design: high-availability architectures, scalability, building services
• Techniques for building very high-capacity switches (time permitting): parallelism, optical fabrics (putting optics inside of routers)

A unique aspect of the workshop is that the exposition of the material is modulated based on the business and technial interests of the organization it is deliverd at. Thus, delivery at a software or systems company may involve greater emphasis and discussion (as required by the company or attendee questions) of software and/or algorithmic aspects, while delivery at a chip or components company may involve greater emphasis on hardware implementation complexities.

Audience: The workshop is targeted at system and network architects, advanced software or hardware development, system engineers, strategic marketing, management and application engineers.

Category: Intermediate to Advanced

Expected background: This is a second-level workshop that assumes that the audience is conversant with concepts in switching and routing, is familiar with IP routing, and understands basic switch architectures (for example shared memory, cross-bar, and multi-stage architectures), so that the workshop can focus primarily on the practical and theoretical details of the operation, performance, implementation, and limitations of these architectures.

Workshop Duration: Expected duration is 2 days (with 6 hours of instruction and Q A each).

Workshop Goals: To help the audience walk away with a detailed understanding of the operation of canonical high-speed switch architectures, an appreciation of some key trade-offs in the design of such switch cores, and an understanding of the architecture of the systems and networks where these switches fit. The attendees leave with an intuitive understanding of the concepts and technologies involved, and the ability to rapidly apply their learning to the development of software, chips, hardware and systems, or use this knowledge for making intelligent deployment decisions for carrier networks.

  Go Back

Advanced IP/MPLS and Optical Networks:
Concepts and Standard

The suite of protocols that defines Multi-Protocol Label Switching (MPLS) has been enhanced to generalize its applicability to the control of optical networks, with one area of prime consideration being the use of these generalized MPLS protocols in upgrading the control plane of optical networks. A major advantage of the MPLS architecture for use as a general network control plane is its clear separation between the data forwarding plane, the signaling (connection control) plane, and the routing (topology discovery/resource status) plane.

A through understanding of MPLS technology, the issues involved in using it to control optical networks, and the protocol and system enhancements needed to enable this is crucial for building advanced transport and hybrid packet/TDM systems, and for building modern transport networks.

With this goal, this workshop covers:

• MPLS technology: Brief introduction, operation, and advantages
• Agile optical networks: motivation and basic concepts
• Control plane architectures and system design trade-offs: possible architectures for the control plane, their pros and cons
• Differences between datagram and optical routing: service impact, standardization
• MPLS control plane for non-packet technologies: Enhancements to IP routing (OSPF) and signaling (RSVP-TE) , how that impacts system and network operation
• Overview of optical control plane standards: discovery, intra-domain routing, signaling, and inter-domain routing, and the UNI (User-Network Interface), with a look at work ongoing at the ITU, IETF, and OIF.
• Dynamic path establishment in multi-layer networks: issues and operation

The workshop adds value by providing attendees a through overview of the enabling technologies and the state of standardization of the control plane for integrated optical networks. It highlights the protocol changes and the system enhancements needed to build advanced systems for such networks, while at the same time explaining their impact on network operation, which is crucial for deploying new network infrastructure. The workshop is particularly useful for service provider organizations and system/software vendors seeking to sort through the maze of options and standards being worked on today, because it crystallizes the key issues, and clearly outlines their impact at the protocol, system, and network level.

Audience: The workshop is targeted at system and network architects, advanced hardware and/or software engineers, strategic marketing, executives, system engineers, application engineers, and network engineers who wish to understand some of the key control plane issues and technology behind building modern transport networks. The workshop will be useful for companies involved in providing systems, software, or chips for wide-area networks, optical networks, and the carrier/ISP services spaces.

Category: Introductory to Intermediate.

Expected background: This is an intermediate-level workshop that assumes that the audience has some familiarity with SDH/SONET signal structure, and some understanding of IP routing and signaling protocols. The workshop will focus primarily on the operation and advantages of MPLS technology, and on the adaptations needed to apply it to the control of SDH/SONET and optical networks, plus the state of relavant international standards.

Workshop Duration: Expected duration is 1.5 days (with 8 hours of instruction plus 2-3 hours of Q A).

Workshop Goals: To enable the attendees to:

• Understand the key issues in the dynamic control of modern transport networks
• Grasp the fundaments of MPLS technology
• Appreciate why MPLS is useful for controlling transport networks
• Comprehend the applicability of MPLS technology for modernizing the control plane of SDH/SONET and optical networks
• Recognize what such control entails
• Focus on how MPLS has been enhanced to apply to transport networks, and how it operates in this context. The attendees will also have a
• See the state of relevant international standards, and of the functionality provided by the basic components of the optical control plane currently undergoing standardization.

  Go Back

MPLS:
Technology for the Next-Generation Internet - Protocols and Carrier Applications for Convergence

Multi-Protocol Label Switching (MPLS) has evolved into a key technology for efficiently operating and managing IP networks, and for facilitating network convergence. Indeed, several carriers today deploy MPLS as an integral part of their network engineering strategy, and most major vendors ship routers with MPLS support.

A major advantage of the MPLS architecture is its clear separation between the data forwarding plane, the signaling (connection control) plane, and the routing (topology discovery/resource status) plane, and its ability to admit various path setup and QoS paradigms with no change in the forwarding mechanisms (which is where it scores over vanilla IP routing).

A thorough understanding of MPLS technology, the issues involved in using it to enhance IP networks, and the protocol and system modifications needed to enable this, is crucial for building and deploying modern converged data networks, and for designing advanced packet or hybrid packet/TDM systems.

With this goal, this workshop focuses on:

• Introduction to MPLS: initial history, and current provider motivations for MPLS development
• Fundamental concepts: MPLS control & forwarding, forwarding equivalence class (FEC)
• System architecture: MPLS node architecture, LSP, LER, LSR .
• MPLS architecture: detailed operation of the forwarding and control components
• Frame-mode and cell-mode MPLS: MPLS over Ethernet and ATM respectively.
• Control protocols: signaling (RSVP-TE) and routing protocol (IGPs): function & operation
• Advanced concepts: constraint-based routing (CBR), CSPF , Diff-serv and MPLS, Diff-serv aware MPLS: how they fit in, their role in an integrated network engineering strategy
• Key applications: intra- and inter-domain traffic engineering and network optimization, fast re-route (restoration), and virtual private networks (VPNs)
• Contemporary issues: introducing MPLS in a provider network - approaches and trade-offs, MPLS equipment - key features, role of MPLS in metro/access networks (discussed throughout the workshop)

Audience: The workshop is targeted at system and network architects, advanced hardware and/or software engineers, strategic marketing, executives, system engineers, application engineers, and network engineers who wish to understand some of the key MPLS issues and technology behind building modern IP networks.

Category: Introductory to Intermediate.

Expected background: This is a beginning to intermediate-level workshop that assumes that the audience has some understanding of IP routing and signaling protocols. The workshop will focus primarily on the operation, protocols, advantages, and uses of MPLS technology, the state of relevant international standards.

Workshop Duration: Expected duration is 2 days (with 6 hours of instruction and Q A each).

Workshop Goals: To enable the attendees to:

• Acquire sound knowledge of MPLS technology
• Comprehend the key issues in the applicability of MPLS technology for modernizing the control and management of IP networks
• Understand its usefulness in IP networks
• Appreciate what such control entails, and how IP had been enhanced to apply to MPLS
• Grasp the state of the relevant international standards
• Cultivate insights into the basic components of MPLS and the key applications of MPLS

  Go Back

The Keys to Converged IP Networks:
Traffic Engineering and Quality-of-Service - Techniques, Carrier Current Practices, and Algorithms

A fundamental impetus at every major carrier anywhere on the globe today is network convergence, the notion of eventually putting multiple types of traffic - packet- and circuit-switched - on a common packet-based backbone that uses the Internet Protocol (IP). Whether it be ATT in the US, France Telecom or BT in Europe, Optus in Australia, KT in Korea, NTT in Japan, or BSNL/MTNL or Reliance in India, each of these carriers is in the process of formulating strategies for efficient network operation and deployment of profitable packet services.

This requires that carriers not only optimize network performance by better resource allocation to the inherently different traffic demands impinging on the network, but that they also use new and improved techniques to offer graded qualities-of-service (QoS) to different traffic. Thus, voice-over-IP (VoIP) or video-on-demand (VoD) traffic requires a rather different network operation strategy than does SAP traffic or best-effort email traffic. This is especially important with the packet infrastructure becoming a critical business resource for enterprises the world over.

The key to achieving this is traffic engineering (TE), defined as a set of processes that systematically aims to put traffic where there is capacity, and intelligent QoS strategies that determine the resources and priority to be accorded to different traffic types at different times.

Understanding the methods, protocols, practices, and algorithms for TE and QoS, and gaining insights into the network- and deployment-level issues that influence the design of chips, systems and software going into carrier networks is crucial for building both advanced networking chips, systems, and software, and the networks themselves.

With this goal, this workshop covers:

• The move towards convergence: drivers, business benefits, current and future roadmaps
• Drawbacks of traditional IGPs & proposed extensions: equal-cost multipath, metric tuning
• TE process model at a provider: measurement, modeling, analysis, optimization
• Components of Internet TE: MPLS, constraint-based routing (CBR), enhanced IGPs and signaling protocols, or alternatives discussed under TE methods" below
• Components of Internet QoS: service definitions, signaling and admission control, policing/shaping, QoS-based routing, QoS-based forwarding (buffer allocation, RED, queueing disciplines)
• TE methods in operational IP networks: evolution and best current practices; case studies of Sprint and Global Crossing
• On-line & off-line algorithms for TE: theoretical developments, practical approximations
• Considerations for inter-domain TE: controlling congestion and peering traffic across provider boundaries
• Practical techniques for providing QoS: what are some of the provider best-practices to offer QoS and tiered services; protocol details and practical issues
• Canonical service provider approaches to network design: fiber infrastructure, restoration strategies, backbone robustness, IGP convergence & network stability, queueing for voice.

The workshop adds value by providing attendees a comprehensive view of the traffic engineering process, algorithms, and applications, and key QoS techniques that have emerged as accepted best practices in the industry and have been adopted in some major provider networks. It provides insights into network-level and deployment-level issues that significantly influence the design of systems and software deployed by carriers. It also highlights some of the requirements that TE imposes on the design of chips, components, systems and software that goes into network elements.

The workshop is useful service provider organizations seeking to use new techniques to design, build, and deploy efficient, revenue-generating networks. It is also useful for software and hardware engineers working on advanced routing protocol, QoS, and/or MPLS implementations, or on system design, and can also be valuable for chip and component designers and system vendors, because it provides insights into the requirements that TE imposes on architecture of chips, processors, systems, and software.

Audience: The workshop is targeted at system and network architects, advanced hardware and/or software development, strategic marketing, executives, system engineers, and application engineers. The workshop will be useful for companies involved in providing systems, software, or chips for wide-area networks, local area networks, optical networks, storage, and carrier/ISP services spaces.

Category: Intermediate to Advanced.

Expected background: This workshop assumes that the audience has a basic understanding of data networking, and has familiarity with elementary IP routing and IP routing protocols. Knowledge of basic MPLS operation and the RSVP protocol will be helpful, though not necessary. The workshop will focus on the routing and signaling protocol modifications/enhancements needed for TE and QoS, on the the state of the art in TE and QoS deployment in carrier networks, and on the algorithms and tools to accomplish TE and provided tiered QoS.

Workshop Duration: Expected duration is 2 days (with 4 hours of instruction each plus 1-2 hours of discussion and QA each day).

Workshop Goals: The aim of the workshop is to enable the audience to:

• Appreciate the TE process
• Gain insights into how that impacts the design of systems and software
• Understand techniques for QoS and their contemporary applications
• Grasp thoroughly the basic steps involved in TE in service provider environments
• Acquire detailed knowledge of contemporary TE and QoS algorithms
• Apply their learning to products (software, chips, or systems) and networking problems (of network build out and deployment) that they are working on.

  Go Back