Designing Wide Area Networks and Internetworks: A Practical Guide: A Practical Guide

Description

• Copyright 2000
• Dimensions: 7-3/8" x 9-1/4"
• Pages: 408
• Edition: 1st

• Book
• ISBN-10: 0-201-69584-7
• ISBN-13: 978-0-201-69584-7

Speedy, reliable, and secure communications are essential for maintaining an organizationis competitiveness, and Wide Area Networks and Internetworks are quickly proliferating in order to meet this need. Building such a network, however, can be a daunting task; a large investment is required, and organizations must navigate through a dizzying array of technological and design options. /P>

Designing Wide Area Networks and Internetworks clarifies this complex task by outlining a top-down, step-by-step process for constructing a WAN or internetwork that is effective for your organization. This book will guide you through the steps of determining requirements, designing the network structure, choosing appropriate technologies, and evaluating results. The authoris practical approach distills exactly what you need to know about networking theory and technological background in order to accomplish a given task.

On the financial side, it is important to note that the difference between a good design and a poor one can represent many millions of dollars per year. This book presents a quantitative, business-oriented approach to network design. It focuses on the economic and performance characteristics of various network technologies and carrier service options, and explains the conditions for which each is optimal.



Downloads

Example(s)

Click below for Example(s) related to this title:
spreadsheets.zip



Sample Content

Downloadable Sample Chapter

Click below for Sample Chapter related to this title:
marcusch14.pdf

Table of Contents

Preface.

I. BEGINNING THE INTERNETWORK DESIGN PROCESS.

1. Networks and Internetworks Technology Today.

Data Communications and Organizational Competitiveness.

Why Build an Internetwork?

Why Is Building an Internetwork so Challenging?

Challenge 1: Lack of a Coherent, Comprehensible Design Process for Wide Area Internetworks.

Challenge 2: Lack of a Holistic, Integrated View of the Wide Area Internetwork.

Challenge 3: Choosing from among a Potpourri of Data Link Technologies and Carrier Service Options.

Enterprise Networks, Intranets, and Extranets.

Internetworking.

Intranets.

Enterprise Networks and Enterprise Internetworks.

The Internet.

Extranets.

Virtual Private Networks (VPNs) and Internet Virtual Private Networks (iVPNs).

What Changes When You Cross a Street?

Evolution and Change.

The Real World Impinges.

2. The Internetwork Design Process: Overview.

Phase 1: Definition of Requirements.

Phase 2: Preliminary Design.

Phase 3: Management Review.

Phase 4: Final Design.

3. Definition of Requirements.

Identify Access Locations.

Identify Application Requirements.

Internet versus Non-Internet Protocols.

Transaction Processing.

Realtime Voice or Video.

Identify Traffic Patterns.

Determine Performance Requirements.

Throughput and the Traffic Matrix.

Delay and Latency.

Variability of Delay.

Packet Drops.

Misrouted Data.

Availability.

Determine Security Requirements.

Consider Cost.

II. THE PRELIMINARY DESIGN PHASE.

4. Preliminary Design: Overview.

Consider the Existing Network and Its Users.

Steps in the Preliminary Design Phase.

Select the Major Transmission Technologies.

Determine the Levels of Hierarchy.

Identify Possible Backbone Locations.

Create a Preliminary Access Design.

Create Access Homing Designs.

Create a Preliminary Backbone Topological Design.

Create a Preliminary Technical Design.

Demonstrate That Your Design Meets the Requirements.

Analyze Your Costs.

5. Circuit Costs and Cost Optimization Techniques.

Elements of Internetwork Costs.

Elements of Circuit Costs.

Cheaper by the Dozen.

Concentration and Aggregation.

Statistical Multiplexing.

Distance-Sensitive Tariffs.

Distance-Insensitive Tariffs.

Components of IXC Frame Relay Charges.

Applications of Distance-Insensitive Tariffs.

Distance-Sensitive versus Distance-Insensitive Tariffs.

The Bent Straw.

6. Selecting Transmission Technologies.

Roles of Each Transmission Technology.

Dedicated Circuits (Leased Lines).

Parallel Circuits and Inverse Multiplexing.

Channelized Circuits.

Circuit Switching.

Analog Dialup.

ISDN.

ISDN BRI for Residential Consumer Access.

Concentration of Customers Who Use ISDN BRI or POTS.

Switched Backup for Leased Line Access.

Frame Relay.

Frame Relay Service Definition.

Frame Relay User Network Interface (UNMI).

Typical Frame Relay Applications in Wide Area Internetworks.

ATM.

ATM Service Definition and Protocol Interface.

Typical ATM Applications in Wide Area Internetworks.

Applications of ATM.

Switched Multimegabit Data Service (SMDS).

Direct Use of SONET.

IP Data over Cable TV Infrastructure.

ADSL and VDSL.

CDPD and Other Wireless Technologies.

Crytographic Encapsulation over the Public Internet: The iVPN.

Security and Management Implications of VPNs.

Economic Implications of the iVPN.

Suitability of an iVPN.

7. Determining the Levels of Hierarchy.

The Concept of Hierarchy.

Partitioning the Internetwork.

What Constitutes a Hierarchical Layer?

How Many Layers?

Hierarchy and Network Topological Design.

8. Identifying Backbone Locations.

Higher-Capacity Circuits and Services.

Higher Availability.

Additional Costs.

Geographic Expanse.

9. Creating a Preliminary Access Design.

Dedicated (Business or Organizational) Access.

Access with Leased Lines.

Access with Multiplexing to Achieve Concentration.

Access with Fast Packet Services.

Switched (Residential) Access.

Bypassing the Carriers.

10. Developing Strategies for Access Homing.

Single Homing.

Dual Homing.

The String of Pearls.

11. Creating a Preliminary Backbone Topological Design.

Fundamental Topologies.

Star.

Full Mesh.

Ring.

Double Star.

Partial Mesh.

Hybrid Topologies.

Summary of Network Topologies.

Creating the Backbone Design.

Traffic Characterization.

Physical Star Network.

Logical Star Network.

Logical Full Mesh Network.

Partial Mesh Network.

12. Naming, Addressing, and Routing.

Naming and the Domain Name System (DNS).

Origin of Domain Names.

Servers and Resolvers.

DNS Name Space.

DNS Records.

Primary and Secondary DNS.

Zones and Delegation.

The IN-ADDR.ARPA Domain.

More Exotic Uses of the Domain Name System.

Addressing.

Networks and Subnetworks.

Classless Inter-Domain Routing (CIDR).

Developing the Address Plan.

Security and the Addressing Plan.

Internet Protocol Version 6 (IPv6).

Routing.

Interior Routing.

Host Support for Redundancy.

Exterior Routing.

13. Security.

Threats to Network Security.

Fundamentals of Network Security.

Security Services.

Security Solutions.

Encryption.

Authentication of Dialup Users (PAP, CHAP, and RADIUS).

Link Layer Encryption.

Network Layer Security (IPSEC).

Session Layer Security: The Secure Socket Layer (SSL).

Application Layer Security (PGP, Kerberos, and X.509 Certificates).

Intrusion Detection.

Decoys.

Firewalls.

Perimeter Security.

Limitations of Perimeter Security.

A Castle with Inner and Outer Walls.

14. The Public Internet: Unique Design Considerations.

Evolution of Today’s Internet.

National Service Providers and Internet Service Providers.

Network Access Points.

Transit Service among Regional ISPs.

The Very High Speed Backbone Network Service (vBNS).

The Route Arbiter (RA) Project.

Structure of the Internet Today.

Direct and Shared Interconnections (Public and Private Peering).

Traffic Characterization in Light of Shortest-Exit Routing.

International Internet Traffic Flows.

Traffic Statistics.

Internet Access.

15. Network Management.

Network Management versus System Management.

Network Management Protocols.

System Management.

Operational Tasks.

Fault Management.

Configuration Management.

Accounting.

Performance Management.

Security Management.

Network Management in Perspective.

16. Validation of the Design against the Requirements.

Performance.

Measures of Performance.

Tools for Evaluating Performance.

Bandwidth.

Queuing Models, Delay, and Latency.

Propagation Delay.

Verification of Functionality.

Security.

Availability.

Availability as a Function of MTBF and MTTR.

Computing Availability for Multiple Components.

More Complex Analyses.

III. THE MANAGEMENT REVIEW PHASE.

17. Management Review: Financial Analysis.

Cost Categories.

Circuits and Services.

Equipment.

Labor.

Facilities.

Ancillary Costs Associated with New Services.

Expenses over Time.

Revenues: The Benefits.

Profitability: Calling the Shots.

Hazards in Estimating Profitability.

Dealing with an Insufficient Profitable Network.

The Budget.

IV. THE FINAL DESIGN PHASE.

18. Final Design Overview.
19. Selection of Carriers and Vendors.

The Carriers.

Interexchange Carriers (IXCs).

Local Exchange Carriers (LECs).

Carriers of Cellular and Wireless Communications.

Nontraditional Carriers.

The Equipment Vendors.

Routers.

Terminal Servers.

CSU/DSUs.

WAN Switches: Frame Relay, SMDS, and ATM.

The Request for Proposal (RFP).

Business Decisions.

Types of RFPs.

Hints in Constructing an RFP.

Index. 0201695847T04062001

Preface

In the past few years, internetworking has exploded on the scene, thrusting itself into the awareness of the general public. A great many major corporations are finding it necessary to construct private internetworks, or intranets, to facilitate interactions internal to the organization, and many corporations are struggling to develop ways to securely use the public Internet to exchange business information with external customers and suppliers. Telephone companies and Internet service providers (ISPs) are moving quickly to provide cheap and reliable Internet access anywhere in the world.

Unfortunately, the design process for networks of this type is anything but straightforward. Any network that attempts to span a significant area today will represent a significant investment. A comprehensive network covering the entire United States could easily cost tens of millions of dollars a year to operate. Much of this cost is in circuits and services from the telephone companies and in labor.

Routing technology is well understood by a small but steadily growing cadre of gurus. But an understanding of routing technology is not, in and of itself, sufficient to create a cost-effective large-scale internetwork. The difference between a good design and a poor one can represent many millions of dollars of cost a year, one way or the other. For this reason, cost analysis can be even more important to the design process than is technology analysis.

Launching a new internetwork today is a daunting task. There are three primary reasons for this:

0. Lack of a well-established design process
1. Compartmentalization of knowledge within most organizations or its unavailability to the design team
2. Overwhelming number of alternative techniques, technologies, and vendors and carriers available

In attempting to fill in all three of these gaps, this book provides a workable network design process and presents and synthesizes necessary aspects of the underlying topics in the context of that process. The book attempts to present the economic and performance characteristics of the various technology and carrier service options in a straightforward way, to explain the domain of conditions over which each is optimal, and the options for intermixing technologies and services.

In this book, I have attempted to cover the necessary concepts and skills in an integrated way and to present them at the point in the design process at which they would be used in practice. I did not feel it necessary to duplicate material that is already widely available in the literature; rather, I have provided pointers to suitable references.

Audience

This book is intended for people who design computer internetworks that extend beyond a single building or a single campus. The book will be of particular interest to network designers and planners, network architects, and engineers who design, build, and support wide area internetworks, including people who belong to the information technology or telecommunications staff of a large corporation. The book will also be of value to data network sales and sales support personnel who work for carriers, Internet service providers (ISPs), and systems integrators. Finally, this book will be of interest to anyone wanting to better understand the design process and the underlying cost models for large-scale computer internetworks.

I have attempted to address the needs both of people in large enterprises that use networks and those who build carrier service provider and independent ISP networks to serve broader constituencies. The basic design principles are much the same in both cases, but service providers may tend to work with different transmission technologies and to think about their economics differently.

This book is written for the practitioner, not for the theorist. It attempts to provide the basic underpinnings of network design in a simple, direct, and intuitive way. Technologies are presented at a level of detail consistent with the needs of the network designer, which is to say that many interesting but irrelevant details are omitted.

I had the good fortune to study discrete event simulation at Columbia under the late Ed Ignalls. He taught me two important things, both of which are reflected in this text.

• Take the time to understand your problem.
• Use that understanding to simplify your model by eliminating those factors that don't really matter.

Prerequisite Background

I have assumed that you already have a general familiarity with communication protocols in general and with TCP/IP in particular. If that is not the case, I would encourage you to read a good text, such as Andrew Tanenbaum's Computer Networks,1 before you tackle this book.

Mathematical and statistical techniques are presented in terms of their application to the network design problem. No derivations appear in this book. At the same time, a basic understanding of probability and statistics is fundamental to the design process, as these data networks are all based in some degree on statistical multiplexing. I have attempted to provide a nuts-and-bolts review of any necessary aspects of statistics; nonetheless, if you are serious about working in this field, and if you have not already taken the equivalent of a semester course in probability and statistics, you would be well advised to do so.

You do not need a Ph.D. in mathematics to do good work in this field, but you do need to be reasonably good at working with figures. Much network design and most network cost analysis today is done not with sophisticated design tools but with commonplace spreadsheeting tools, such as Microsoft Excel. You will need to be facile with a spreadsheet. With that in mind, a number of illustrative spreadsheets are included. If you have a PC or compatible device with a decent spreadsheeting program, you will find it helpful to work through the examples that we have provided on the Web site for this book (http://www.awl.com/cseng/titles/0-201-69584-7). The formatted versions of the sample spreadsheets are suitable for use by Microsoft Excel; in addition, text versions are provided that should work adequately with any spreadsheeting product that can accommodate comma-delimited text.

This book takes an unabashedly U.S.-centric outlook, at least in terms of the examples that are offered. Yes, I realize that there is a large world out there--as a matter of fact, I spent several years working in Europe. The focus on the United States was intentional and was not based solely on the consideration that internetworking is most advanced in the United States. The deregulation of the carriers is also most advanced at present in the United States. It is therefore the United States that poses the greatest opportunity and the greatest challenge for the designer. Deregulation is, however, a worldwide trend, and its pace is accelerating. The techniques that are presented have broad applicability, and the holistic approach that I have emphasized will become increasingly necessary as deregulation progresses.

Organization of This Book

The book begins with Chapter 1, which is a look at internetworking today. Chapter 2 then provides an overview of a suggested network design process, with its four major elements, or phases: definition of requirements, preliminary design, management review, and final design.

Chapter 3 provides more detail on the first phase, definition of requirements, discussing the identification of access locations, performance, availability, and security requirements. Protocol requirements must also be considered in this phase, as well as any needs for realtime voice or video or for transaction processing--needs that typically impose a requirement for deterministic delay. Unit cost requirements must also be considered. These three chapters compose Part I.

Part II drills down into the second phase: preliminary design. It opens with a discussion of the structure of carrier tariffs, as well as the trade-offs and combinations that are possible among them, as this permeates many of the tasks in this phase. Individual chapters then expand on the more complex aspects of the task: selecting an appropriate transmission technology or set of technologies; determining how many levels of hierarchy to use; identifying the number and placement of backbone locations; creating preliminary access, concentration, and backbone topological designs; and laying out the main engineering aspects of the design: routing, naming, addressing, network management, and security. This part concludes with a lengthy chapter on design validation: Have you met your design requirements?

Part III discusses the third phase: management review. Once the preliminary design is complete and the costs are understood, it is usually necessary to take the associated business case to management for review, typically including in-depth financial analysis and a go/no-go decision.

Part IV presents the fourth phase: final design. It is at this point that you will select carriers and equipment vendors. In many cases, you will use a Request for Proposal (RFP) process to obtain competitive price quotations. Once you have made your choices, you will revisit your design assumptions and will fill in the missing details in your design.

0201695847P04062001



Updates

Submit Errata

