Paradyne 8540, 8546, HOTWIRE DSLAMж Network Configuration Manual

Download

Page 1

HOTWIRE DSLAM

FOR 8540 AND

8546 DSL CARDS

NETWORK CONFIGURATION GUIDE

Document No. 8000-A2-GB21-20

November 1997

Page 2

Notice

This publication is protected by federal copyright law. No part of this publication may be copied or distributed, transmitted, transcribed, stored in a retrieval system, or translated into any human or computer language in any form or by any means, electronic, mechanical, magnetic, manual or otherwise, or disclosed to third parties without the express written permission of Paradyne Corporation, 8545 126th Avenue North, P.O. Box 2826, Largo, Florida 33779-2826.

Paradyne Corporation makes no representation or warranties with respect to the contents hereof and specifically disclaims any implied warranties of merchantability or fitness for a particular purpose. Further, Paradyne Corporation reserves the right to revise this publication and to make changes from time to time in the contents hereof without obligation of Paradyne Corporation to notify any person of such revision or changes.

Changes and enhancements to the product and to the information herein will be documented and issued as a new release to this manual.

Warranty, Sales, and Service Information

Contact your sales or service representative directly for any help needed. For additional information concerning warranty, sales, service, repair, installation, documentation, or training, use one of the following methods:

 Via the Internet: Visit the Paradyne World Wide W eb site at http://www.paradyne.com  Via Telephone: Call our automated call system to receive current information via fax or to speak with a

company representative.

— Within the U.S.A., call 1-800-870-2221 — International, call 813-530-2340

T rademarks

All products and services mentioned herein are the trademarks, service marks, registered trademarks or registered service marks of their respective owners.

Printed on recycled paper

November 1997

8000-A2-GB21-20

Page 3

Contents

About This Guide

 Document Purpose and Intended Audience v. . . . . . . . . . . . . . . . . . . . . . . . .

 Document Summary vi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Product-Related Documents vii. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 Introduction to the HotWire DSLAM

 What is the HotWire DSLAM? 1-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 HotWire DSLAM Components 1-4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

HotWire DSLAM Chassis 1-4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

MCC Card 1-5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

DSL Cards 1-6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 What is an RTU? 1-6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5170 RTU 1-6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5171 Remote PC NIC 1-7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5216 and 5246 RTUs 1-8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5446 RTU 1-10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Data Rates 1-11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Overview of the HotWire DSLAM Network Model 1-12. . . . . . . . . . . . . . . . . . . .

 Understanding the Domain Types 1-16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 Service Domain Features

 Overview 2-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Protocols 2-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Proxy ARP (Theory of Operation) 2-2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Filtering 2-4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8000-A2-GB21-20

Scenario 1: Without Proxy ARP 2-2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Scenario 2: With Proxy ARP 2-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

November 1997

Page 4

Contents

3 Management Domain Features

 Overview 3-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Network Management Systems — SNMP and DCE Manager 3-1. . . . . . . . .

 Applications for Management Domain 3-2. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ping 3-2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TraceRoute 3-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TFTP Client 3-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

T elnet 3-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4 Components of the Network Model

 Overview 4-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Service Domain Components 4-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Proxy ARP 4-4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Management Domain Components 4-5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Discovering Devices on the Network (Discovery) 4-6. . . . . . . . . . . . . . . .

MCC Card Proxy ARP 4-7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5 IP Address Allocation

 Overview 5-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Port Naming Convention 5-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Assigning IP Addresses 5-2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Host Addressing 5-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Structured Subnet Addressing 5-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Management IP Address Allocation 5-7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Peer IP Addresses 5-8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Service IP Address Allocation 5-10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Dynamic IP Addressing 5-11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Recording Your Configuration Settings 5-11. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6 IP Routing

 Overview 6-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Routing Table 6-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Static Routes for Static IP Addressing 6-2. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

MCC Card Static Route Example 6-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

DSL Card Static Route Example 6-4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Dynamic Routes for Dynamic IP Addressing 6-5. . . . . . . . . . . . . . . . . . . . . . . .

How Does Dynamic IP Addressing Work? 6-6. . . . . . . . . . . . . . . . . . . . . .

 General DHCP Relay Agent Configuration 6-8. . . . . . . . . . . . . . . . . . . . . . . . . .

November 1997

8000-A2-GB21-20

Page 5

7 IP Filtering

Contents

 Notes to the Authentication Server Administrator 6-8. . . . . . . . . . . . . . . . . . . .

RADIUS Authentication 6-9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

XT ACACS Authentication 6-9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Source-Based Routing 6-10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Without Source-Based Routing 6-10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

With Source-Based Routing 6-10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Overview 7-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 What is a Filter? 7-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Security Advantages 7-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Management Traffic Leakage 7-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Service Security 7-4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Service Security Filtering Scenario 7-5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8 SNMP Agent

 Overview 8-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 MIB Compliance 8-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Supported Traps 8-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 General SNMP Agent Configuration 8-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9 Packet Walk-Throughs

 Overview 9-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Packet Walk-Through Using an 8540 DSL Card 9-1. . . . . . . . . . . . . . . . . . . . .

Service Domain Packet Walk-Through 9-1. . . . . . . . . . . . . . . . . . . . . . . . .

Management Domain Packet Walk-Through 9-3. . . . . . . . . . . . . . . . . . . .

 Packet Walk-Through Using an 8546 DSL Card 9-3. . . . . . . . . . . . . . . . . . . . .

Service Domain Packet Walk-Through 9-3. . . . . . . . . . . . . . . . . . . . . . . . .

Management Domain Packet Walk-Through 9-4. . . . . . . . . . . . . . . . . . . .

8000-A2-GB21-20

November 1997

iii

Page 6

Contents

A Network Configuration Worksheets

 Overview A-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Summarizing the Network Configuration A-1. . . . . . . . . . . . . . . . . . . . . . . . . . .

 Management Domain Configuration Worksheets A-2. . . . . . . . . . . . . . . . . . . .

TASK 1: Assign an IP Address to the MCC Card A-3. . . . . . . . . . . . . . . . .

TASK 2: Clear NVRAM A-5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TASK 3: Assign an IP Address to the Backplane (s1b) A-6. . . . . . . . . . . .

TASK 4: Assign IP Addresses to the DSL Cards A-7. . . . . . . . . . . . . . . . .

TASK 5: Create a Default Route A-9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TASK 6: Reset the MCC Card A-11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TASK 7: (When Using an 8546 DSL Card) Configure

the HotWire 5446 RTU Management Domain IP Addresses A-12. . . . . . .

TASK 8: Create a Static Route to an NMS A-14. . . . . . . . . . . . . . . . . . . . . .

 Service Domain Configuration Worksheets A-16. . . . . . . . . . . . . . . . . . . . . . . . .

TASK 1: Assign IP Addresses to the DSL Card

LAN Interface (e1a) A-17. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TASK 2: Reset the DSL Card A-19. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TASK 3: Create Static Routes to End-User Systems A-20. . . . . . . . . . . . .

TASK 4: Create a Default Route or Source Route A-22. . . . . . . . . . . . . . . .

TASK 5: Define DHCP Relay Features to Enable Dynamic

IP Address Configuration A-24. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

B IP Filtering Configuration Worksheets

 Overview B-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Summarizing How to Define a Filter B-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Filtering Configuration Worksheets B-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Defining the Filter and Rules B-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Binding the Filter B-7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

C SNMP Configuration Worksheets

 Overview C-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 Summarizing the General SNMP Agent Configuration C-1. . . . . . . . . . . . . . .

 SNMP Agent Configuration Worksheets C-2. . . . . . . . . . . . . . . . . . . . . . . . . . . .

Defining a Community and Enabling Traps C-2. . . . . . . . . . . . . . . . . . . . . .

Preventing Unauthorized Access C-6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Glossary

Index

November 1997

8000-A2-GB21-20

Page 7

About This Guide

Document Purpose and Intended Audience

This guide describes the HotWire Digital Subscriber Line Access Multiplexer (DSLAM), its internetworking features, and how it works for the HotWire 8540 and 8546 Digital Subscriber Line (DSL) cards. It also provides information on what you need to know before planning your network. Use this guide to:

 Obtain a basic understanding of the HotWire DSLAM  Understand how the DSLAM works within the network  Understand the network model, management domain, and service domain  Understand how to allocate Internet Protocol (IP) addresses  Understand dynamic IP addressing

NOTE:

The

DSL Sourcebook

technology and opportunities. Read the and why of DSL-based service deployment. The book is available by calling 1-800-PARADYNE.

This guide is intended for network planners, network administrators, and network maintainers. It is assumed that you have a basic understanding of internetworking protocols and their features. Specifically, you should have a basic familiarity with Simple Network Management Protocol (SNMP), Network Management Systems (NMSs), and the following internetworking concepts:

 TCP/IP applications  IP and subnet addressing  IP routing (also referred to as IP forwarding)

, written by Paradyne Corporation, is about DSL

DSL Sourcebook

for the what, how,

8000-A2-GB21-20

November 1997

Page 8

About This xxxx

Document Summary

Section Description

Chapter 1

Chapter 2

Chapter 3

Chapter 4

Chapter 5

Chapter 6

Chapter 7

Introduction to the HotWire DSLAM.

overview of the HotWire DSLAM and its components. It also briefly describes the network model and the domain types.

Service Domain Features.

are supported in the service domain.

Describes the features that

Management Domain Features.

that are supported in the management domain.

Components of the Network Model.

components of the service and management domains. These domains comprise the network model.

IP Address Allocation.

allocation schemes for the components that make up the network model. With these allocation schemes, IP addresses can be assigned statically or dynamically. It also describes the naming convention used for the HotWire DSLAM system ports.

IP Routing.

destination-based routing (static and dynamic routes) and source-based routing.

IP Filtering.

scenarios.

Provides information and examples of

Describes IP filtering advantages and

Describes the IP address

Provides an

Describes the features

Describes the

Chapter 8

Chapter 9

Appendix A

Appendix B

Appendix C

Glossary Defines acronyms and terms used in this document. Index Lists key terms, acronyms, concepts, and sections in

SNMP Agent.

(community configuration and trap configuration).

Packet Walk-Throughs.

data packets are routed through the service and management domains.

Network Configuration Worksheets.

worksheets to record your configuration settings.

IP Filtering Configuration Worksheets.

worksheets to help you define a filter for a specific interface on an MCC or DSL card.

SNMP Configuration Worksheets.

to help you set up community names and enable/disable the generation of alarms.

alphabetical order.

Describes the SNMP agent configuration

Provides examples of how

Provides

Provides worksheets

November 1997

8000-A2-GB21-20

Page 9

Product-Related Documents

Document Number Document Title

About This Guide

5020-A2-GN10

5030-A2-GN10

5034-A2-GN10

5100-A2-GB21

5100-A2-GB22 5216-A2-GN10

5246-A2-GN10

5446-A2-GN10

7700-A2-GB23

7800-A2-GB26 8000-A2-GB20

HotWire POTS Splitter Central Office Installation Instructions

HotWire 5030 POTS Splitter Customer Premises Installation Instructions

HotWire 5034 Indoor POTS Splitter Customer Premises Installation Instructions

HotWire 5171 Remote PC Network Interface Card User’s Guide

HotWire 5170 Remote Termination Unit User’s Guide HotWire 5216 Remote Termination Unit (RTU)

Customer Premises Installation Instructions HotWire 5246 Remote Termination Unit (RTU)

Customer Premises Installation Instructions HotWire 5446 Remote Termination Unit (RTU)

Customer Premises Installation Instructions DCE Manager for HP OpenView for Windows

User’s Guide DCE Manager User’s Guide HotWire DSLAM for 8540 and 8546 DSL Cards

User’s Guide

8000-A2-GB24

HotWire DSLAM Configuration for 8540 and 8546 DSL Cards Startup Guide

8000-A2-GB90

HotWire 8100/8200 Interworking Packet Concentrator (IPC) User’s Guide

8000-A2-GN11

HotWire Management Communications Controller (MCC) Card Installation Instructions

8540-A2-GN10

HotWire 8540 Digital Subscriber Line (DSL) Card Installation Instructions

8546-A2-GN10

HotWire 8546 Digital Subscriber Line (DSL) Card Installation Instructions

8600-A2-GN20

HotWire 8600 Digital Subscriber Line Access Multiplexer (DSLAM) Installation Guide

8800-A2-GN21

HotWire 8800 Digital Subscriber Line Access Multiplexer (DSLAM) Installation Guide

Contact your sales or service representative to order additional product documentation.

8000-A2-GB21-20

November 1997

vii

Page 10

Introduction to the HotWire DSLAM

What is the HotWire DSLAM?

The HotWire Digital Subscriber Line Access Multiplexer (DSLAM) is a multiservices DSL platform that provides high-speed Internet or Intranet access over traditional twisted-pair telephone wiring. The DSLAM chassis houses DSL cards that interoperate with multiple types of HotWire Remote Termination Units (RTU) to deliver applications at multimegabit speeds in support of packet services over a Digital Subscriber Line (DSL) link.

High-speed service traffic types from the DSL links are groomed and then concentrated for efficient forwarding to backbone routers. By enabling very high speeds using DSL technology and then concentrating Internet Protocol (IP) traffic, greater performance is realized. Backbone service nodes can be placed deeper into the network, dramatically improving the economics of service provisioning while taking advantage of the substantial speed increases of DSL.

When used in combination with a HotWire 8200 Interworking Packet Concentrator (IPC), the HotWire DSLAM provides high-speed IP service concentration over a wide array of Local Area Network (LAN) architectures as well as an Asynchronous Transfer Mode (ATM) interface to Wide Area Networks (WANs).

In addition, the HotWire DSLAM with a HotWire RTU can be multiplexed with Plain Old Telephone Service (POTS) over the same copper line providing simultaneous usage of POTS and digital applications to separate locations. That is, the optional POTS splitters allow simultaneous voice and data connections over a standard telephone line.

NOTE:

If you would like more information on DSL-based services, applications, and network deployment, refer to Paradyne’s ordered by calling 1-800-PARADYNE.

DSL Sourcebook.

The book may be

8000-A2-GB21-20

November 1997

1-1

Page 11

Introduction to the HotWire DSLAM

The following illustration shows a high-level view of a HotWire configuration:

NOTE:

The cable connection from a DSL card to a Main Distribution Frame (MDF) can either be a direct connection to the MDF or a connection through a POTS splitter to an MDF, but not both. Refer to the appropriate HotWire DSLAM Installation Guide for more information.

Central Office (CO)

Network

Service

Provider

Legend: DSL - Digital Subscriber Line RTU - Remote Termination Unit

HotWire

8200

IPC

Ethernet

DSL

CARD

DSLAM

MDF - Main Distribution Frame POTS - Plain Old Telephone Service IPC - Interworking Packet Concentrator

*CO

POTS

Splitter

Switch

MDF

POTS/DSL

* Optional

Customer Premises (CP)

Data

Interface

RTU

POTS

*CP

POTS

Splitter

Voice

Interface

97-15674-01

The HotWire DSLAM can be configured to work with multiple types of RTUs installed at the customer end of the local telephone loop. RTUs terminate the DSL line and allow users at remote locations to access Network Service Providers (NSPs) or corporate networks by means of the DSL phone line.

When using an 8540 DSL card in the DSLAM, the DSLAM can be configured to interoperate with any one of the following RADSL RTUs on each of its four DSL ports:

 5170 RTU — Operates at speeds up to 7 Mbps with a simple bridge that

supports up to 32 end-user systems.

 5171 Remote PC Network Interface Card (NIC) — Operates at speeds up to

2.5 Mbps supporting a single user’s PC.

 5216 RTU — Operates at speeds up to 1.28 Mbps supporting a single user.  5246 RTU — Operates at speeds up to 2.5 Mbps with a transparent learning

bridge that supports up to 32 end-user systems.

1-2

November 1997

8000-A2-GB21-20

Page 12

Introduction to the HotWire DSLAM

When using an 8546 DSL card in the DSLAM, the DSLAM can be configured to interoperate with up to four 5446 RTUs. The 5446 RTU operates as an IP forwarder at speeds up to 7 Mbps. This RTU supports up to 32 end-user systems with individual IP addresses or subnets.

The following illustration shows a HotWire network configuration from the 8800 DSLAM to multiple RTUs. (Stacked 8600 DSLAMs can also be used in place of an 8800 DSLAM.)

ES 1

8546

DSL

Card

8546

DSL

Card

8540

DSL

Card

POWER

8540

DSL

Card

ALARMS

MajorMinorFanBA

5446 RTU

ALMPWR

5170 RTU

ALMPWR

5246 RTU

ALMPWR

8800 DSLAM

5246 RTU

ALMPWR

5216 RTU

ALMPWR

TST

ETHERNET

DSL

TST

ETHERNET

DSL

PC With Internal 5171 PC NIC

TST

ETHERNET

DSL

TST

ETHERNET

DSL

TST

ETHERNET

DSL

10BT

Hub

Hub or Router

Hub

Hub or Router

ES 32

ES 1

ES 32

ES 1

ES 32

ES 1

ES 32

ES = End-user System RTU = Remote Termination Unit 10BT= 10BaseT NIC = Network Interface Card

8000-A2-GB21-20

November 1997

97-15675-01

1-3

Page 13

Introduction to the HotWire DSLAM

HotWire DSLAM Components

The HotWire DSLAM resides in a central office (CO) or wire center. It consists of the following components:

 HotWire DSLAM chassis  MCC card  DSL cards

In addition, optional POTS splitters can be installed at the CO. For information about a CO POTS Splitter, see the

Installation Instructions

HotWire DSLAM Chassis

There are two types of chassis:

 HotWire 8600 DSLAM chassis

The HotWire 8600 DSLAM is a low-cost alternative to the HotWire 8800 DSLAM. The 8600 DSLAM is an independent, standalone system. A stackable design provides for up to six 8600 DSLAMs to share management access through a single MCC card. In a stacked configuration, the first or base chassis is equipped with an MCC card in Slot 1, leaving Slots 2 and 3 available for up to two DSL cards or a maximum of eight DSL ports. Each additional chassis houses up to three DSL cards. This stacking capability allows you to incrementally expand your DSL access service.

HotWire POTS Splitter Central Office

1-4

Alrm

TestTXRX

INPUT

AC T5A 250V

RTN48V

AAB B

48VDC CLASS 2 OR

LIMITED PWR SOURCE

SYSTEM

DC FUSES

T4A, MIN. 48V

Alrm

46 3 2

POSITION

ETHERNET

TestTXRX

ETHERNET

TestTXRX

ETHERNET

DC PWR

FAN

. .

ALM

STACK

Col1234

DSL PORT

Col1234

DSL PORT

Col

MANAGEMENT

OUT SERIAL

MCC 1

LAN/WAN SLOT

8546

RADSL

8546

RADSL

8000

MCC

LINE

97-15350-01

For more information about the HotWire 8600 DSLAM chassis, see the

HotWire 8600 Digital Subscriber Line Access Multiplexer (DSLAM) Installation Guide

November 1997

8000-A2-GB21-20

Page 14

Introduction to the HotWire DSLAM

 HotWire 8800 DSLAM chassis

The HotWire 8800 DSLAM is a 20-slot chassis designed to house up to 18 DSL cards and one MCC card. (The remaining slot is reserved for future use.) The HotWire 8800 DSLAM chassis requires one MCC card and at least one DSL card.

ALARMS

POWER

Major MinorFanBA

MCC Card

SYSTEM

ETHERNET

DSL PORT

RADSL

Alm

Test

Coll

SLOTS 13-18

SLOTS 7-12

SLOTS 1 - 6

-48V INPUT

LINES

-48V (A)

-48V (B)

RET (A)

RET (B)

FR GND

LAN/WAN SLOT

101214

LAN/WAN SLOT

13 15

SYSTEM

Alm

Test

ETHERNET

Coll

MCC

MGT

SERIAL

ALARM

MGT

10BT

97-15280

For information about the HotWire 8800 DSLAM chassis, see the

HotWire

8800 Digital Subscriber Line Access Multiplexer (DSLAM) Installation Guide

The MCC card is a single resource in the HotWire DSLAM that provides consolidated management access for the DSL cards and the HotWire RTU from any one of the following:

 SNMP management systems, such as HP OpenView with Paradyne’s DCE

Manager (via the MCC card’s Ethernet port),

 Remote telnet sessions (via the MCC card’s Ethernet port),  Local asynchronous terminal (via the MCC card’s VT100 serial port), or  Remote asynchronous terminal connected to a modem (via the MCC card’s

serial port).

8000-A2-GB21-20

November 1997

1-5

Page 15

Introduction to the HotWire DSLAM

The MCC card performs alarm monitoring of the HotWire DSL cards, the DSLAM power and cooling systems, and interfaces to the CO alarm system. It also interfaces with external managers and servers (e.g., File Transfer Protocol servers) for system configuration and management.

DSL Cards

Each 8540 or 8546 DSL card in the HotWire DSLAM chassis contains four DSL ports with on-board IP packet forwarding functionality. The outputs of the four DSL ports are combined onto one 10BaseT interface for connecting to the Internet or Intranet by means of the Network Access Provider’s network.

For a list of the supported features of the 8540 and 8546 DSL cards, see the

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

What is an RTU?

A HotWire Remote Termination Unit (RTU) resides at the customer premises. The RTU connects to the local loop to provide high-speed connectivity to the HotWire DSLAM. In addition, the RTU and your telephone can function simultaneously over the same pair of copper wires when a POTS splitter is used at both ends of the local loop. The POTS splitter filters out the DSL signal and allows the POTS frequencies to pass through.

5170 RTU

If you have an . . .

8540 DSL Card

8546 DSL Card 5446 RTU

The following sections briefly describe these RTUs.

The HotWire 5170 RTU is a standalone unit designed for the home-office users with a LAN. The RTU communicates with any computer equipment or router using its Ethernet network interface card (NIC).

Control of the 5170 RTU is supplied by a windows-based diagnostics utility which enables users to check RTU status, network transmission status, and run diagnostic tests.

You can connect the 5170 RTU directly to your PC using an 8-pin modular Ethernet cable.

Your DSL card interoperates with a . . .

5170 RTU 5171 Remote PC NIC 5216 RTU 5246 RTU

1-6

November 1997

8000-A2-GB21-20

Page 16

Introduction to the HotWire DSLAM

The following illustration shows the HotWire 5170 RTU with its 10BaseT interface connected to multiple end-user systems (typically a PC with a LAN card).

End-user

Optional

POTS

System 1

5171 Remote PC NIC

POTS

POTS/DSL

From Network

Access Provider

NID = Network Interface Device

NID

For more information about the HotWire 5170 RTU, see the

5170 Remote Termination Unit User’s Guide

POTS

Splitter

DSL

End-user System 2

10BaseT

5170 RTU

HUB

End-user

System 32

97-15455c

HotWire

The HotWire 5171 PC Network Interface Card (NIC) is a 16-bit ISA, add-on card with a 6-pin telephone modular jack connector used for the DSL network connection. The 5171 PC NIC edge connector plugs into a 16-bit expansion slot in an IBM-compatible 80486 (or higher) system board and conforms to ISA bus standards.

The following illustration shows a PC with an internal HotWire 5171 PC NIC.

POTS/DSL

From Network

Access Provider

NID = Network Interface Device

For more information about the HotWire 5171 Remote PC NIC, see the

5171 Remote PC Network Interface Card User’s Guide

8000-A2-GB21-20

POTS

NID

November 1997

Optional

POTS

Splitter

DSL

PC with Internal

HotWire 5171

PC NIC

97-15713

HotWire

1-7

Page 17

Introduction to the HotWire DSLAM

5216 and 5246 RTUs

The HotWire 5216 and 5246 RTUs are each composed of a DSL modem and a bridge.

The HotWire 5216 RTU is designed for home office/residential applications and supports a single end user system. The 5216 RTU supports limited DSL line rates.

The following illustration shows the HotWire 5216 RTU with its 10BaseT interface connected directly to an end-user system (typically a PC or workstation with a LAN card).

Optional

POTS

POTS/DSL

From Network

Access Provider

NID = Network Interface Device

NID

POTS

Splitter

DSL

5216

5246

RTU

End-user

System

10BaseT

97-15455a

The HotWire 5246 RTU is designed for small office or home office (SOHO) applications and supports up to 32 end-user systems with a LAN. The HotWire 5246 supports full speed DSL line rates and filters local LAN traffic from traversing the DSL link by incorporating learning bridge functionality.

The following illustration shows the 5246 RTU with its 10BaseT interface connected to multiple end-user systems (typically a PC or workstation with a LAN card) via an Ethernet 10BaseT hub.

1-8

November 1997

8000-A2-GB21-20

Page 18

Optional

Introduction to the HotWire DSLAM

End-user System 1

POTS

POTS/DSL

From Network

Access Provider

NID = Network Interface Device

NID

For more information about these RTUs, see the

Termination Unit (RTU) Customer Premises Installation Instructions

POTS

Splitter

DSL

5246

10BaseT

RTU

HUB

End-user

System 32

HotWire 5216 Remote

End-user System 2

and the

97-15456

HotWire 5246 Remote Termination Unit (RTU) Customer Premises Installation Instructions

8000-A2-GB21-20

November 1997

1-9

Page 19

Introduction to the HotWire DSLAM

5446 RTU

The HotWire 5446 RTU is composed of a DSL modem supporting full speed DSL line rates and an IP forwarder that can support multiple end-user systems.

The 5446 RTU can be connected directly to an end-user system or to multiple end-user systems via an Ethernet (10BaseT) hub.

The following illustration shows the HotWire 5446 RTU with its 10BaseT interface connected directly to an end-user system (typically a PC or workstation with a LAN card) with a crossover cable:

Optional

POTS

POTS/DSL

From Network

Access Provider

NID = Network Interface Device

NID

POTS

Splitter

DSL

5446

RTU

End-user

System

10BaseT

97-15455b

A 5446 RTU supports multiple service domains and can be configured with up to four IP subnets. Each of the four IP subnets can be comprised of multiple users by appropriately sizing the respective IP subnet.

In addition, each 5446 RTU supports up to 32 end-user systems that dynamically acquire their IP addresses or have static IP addresses. The Dynamic Host Configuration Protocol (DHCP) is used for dynamic addressing. In both cases (dynamic and static), these 32 end-user systems must be in one of the configured IP subnets.

The following illustration shows a HotWire 5446 RTU with its10BaseT interface connected to multiple end-user systems (typically a PC or workstation with a LAN card) via an Ethernet (10BaseT) hub.

1-10

November 1997

8000-A2-GB21-20

Page 20

Optional

Introduction to the HotWire DSLAM

End-user System 1

POTS

Data Rates

POTS

POTS/DSL

From Network

Access Provider

NID = Network Interface Device

NID

For more information about the 5446 RTU, see the

Termination Unit (RTU) Customer Premises Installation Instructions

POTS

Splitter

DSL

5446

10BaseT

RTU

HUB

End-user

System 32

HotWire 5446 Remote

End-user System 2

97-15456

The HotWire DSL card employs Rate Adaptive Digital Subscriber Line (RADSL) devices based on Carrierless Amplitude & Phase (CAP) technology. The RADSL speed is asymmetric. This means that the downstream rate (from the DSLAM to the RTU) is faster than the upstream rate (from the RTU to the DSLAM).

You can manually set the speed (providing the line you are using can support the specified speed) or set the mode to rate adaptive. When the mode is set to rate adaptive, the HotWire DSLAM determines the line speed during the initial handshaking session between the DSLAM and the RTU based on the local loop length, the amount of noise on the loop, and the user-configurable upper and lower speed limits.

The following are the maximum upstream and downstream data rates:

 Maximum upstream data rate: 1088 kbps (1.088 Mbps)  Maximum downstream data rate: 7168 kbps (7.168 Mbps)

Data rates and data transmission distances vary depending on existing telephone line conditions (i.e., the DSL cards measure performance during operation and can adjust the upstream or downstream rate to match changing local loop characteristics due to temperature, humidity, or electrical interference). Also, the maximum data rate will be dependent on the RTU in use.

For a complete listing of the DSL card data rates and information on how to set the line speed, see Chapter 6,

DSL Card Configuration

for 8540 and 8546 DSL Cards User’s Guide

, of the

HotWire DSLAM

8000-A2-GB21-20

November 1997

1-11

Page 21

Introduction to the HotWire DSLAM

Overview of the HotWire DSLAM Network Model

The HotWire DSLAM and the HotWire RTUs provide high-speed connectivity to the Internet, corporation, or other network service from the end-user system.

The HotWire DSLAM network model can be implemented in a number of ways. For example:

 A Small Office/Home Office (SOHO) implementation with one or more users

connected to a LAN needing high-speed connectivity to an Internet Service Provider (ISP).

 A SOHO implementation with one or more users connected to a LAN needing

high-speed connectivity to the corporate LAN or Intranet.

 A campus implementation needing internetworking between several sites,

each with a LAN.

The network model for these examples can be partitioned into the following building blocks:

 Network Service Provider (NSP)  Network Access Provider (NAP)  Service Subscriber

Network

Service

Provider

Network

Access

Provider

Service

Subscriber

97-1545

1-12

November 1997

8000-A2-GB21-20

Page 22

Introduction to the HotWire DSLAM

The following illustration shows a detailed view of the network model:

Network Service Provider Network Access Provider Service Subscriber

Internet

Corporate

ISP

Router

Intranet

Router

Access to

Point-of-

Presence

WAN

Access to

Point-of-

Presence

VLAN Switch)

WAN-C

(Router or

Wire

Center

RTU

Wire

Center

RTU

Wire Center

RTU

DSLAM

DSL

RTU

97-15499-02

 The Service Subscriber is the user (or set of users) that has contracted to

receive networking services (e.g., Internet access, remote LAN access) for the end-user system from one or more Network Service Providers (NSPs). Service Subscribers may be:

— Residential users connected to public network services (e.g., the Internet) — W ork-at-home users connected to their corporate Intranet LAN — Commercial users at corporate locations (e.g., branch offices) connected

RTUs must be installed at the customer premises to provide the Service Subscriber access by way of DSL to any of the above services.

 The Network Access Provider (NAP) is typically the network provider (e.g.,

a Regional Bell Operating Company, an Alternate Local Exchange Carrier) that has access to the copper twisted pairs over which the DSL-based service operate. The NAP provides a transit network service permitting connection of service subscribers to NSPs.

8000-A2-GB21-20

by a LAN to other corporate locations or connected to public network services

November 1997

1-13

Page 23

Introduction to the HotWire DSLAM

Typically, the NAP network is organized into three components:

— Wire center

The wire center is usually a local serving office where the wiring from the service subscribers is terminated on the HotWire DSLAM. This could be a CO.

— Wide Area Network (WAN)

The WAN concentrates and switches data traffic from multiple wire centers to one or more Regional Centers where service providers have access to the network.

— Regional center

The NSP’s Point-of-Presence (POP) is the access point to the NAP network for an NSP and is located at the regional center. The connection from the NAP to the NSP network is typically across a WAN connection to the NSP router at the regional center. This router acts as a next-hop location to the NSP’s network.

 The Network Service Providers (NSPs) can be either public access

providers to the Internet (i.e., Internet Service Providers) or private access providers to corporate LANs, providing services based on the Internet Protocol (IP). In some cases, the NSP and the NAP can be a single organization.

One or more HotWire DSLAMs are connected to a Wide Area Network Concentrator (WAN-C) via a LAN. The WAN-C concentrates data traffic from one or more DSLAMs into facilities providing access to the WAN. The WAN-C can be either a router (a layer 3 networking device) or a VLAN switch (a layer 2 networking device).

 If WAN-C is a router, the WAN must be a routed IP network (i.e., a network

of interconnected IP routers). In this case:

— The router at the wire center is required to support routing policies which

permit packets arriving from the local DSLAMs to be routed based on the service subscriber source IP address. The packets are routed to the subscribed service providers’ POP based on the source IP address.

— The routing tables in the DSLAM are configured such that the next-hop

router is the IP address of the wire center router for all authorized subscriber IP source addresses. (See the discussion on source-based routing in Chapter 6,

IP Routing

In addition, the router at the regional center may need to participate in an exterior gateway protocol, such as the Border Gateway Protocol, to exchange routing information between the NSP and NAP routing networks.

— Packets flowing from the NSP network to the end-user systems are

routed within the NAP network based on the packet destination IP address.

1-14

November 1997

8000-A2-GB21-20

Page 24

Introduction to the HotWire DSLAM

 If WAN-C is a VLAN switch, the WAN must be a layer 2 switching network

supporting a Virtual LAN overlay. In this case:

— Each NSP would be a member of a different Virtual LAN. — The VLAN switch at the wire center would support either port-based

VLAN switching (i.e., switching all MAC frames received on a specific port to a specific NSP VLAN on the WAN) or port-based VLAN switching with MAC-based attributes (i.e., switching frames received on a specific port to a specific NSP VLAN on the WAN based on the destination MAC address) for packets sent from the DSLAMs.

— The router at the NSP premises would either be front ended by a VLAN

switch or have an integrated VLAN card that supports protocols consistent with the wire center VLAN switch (e.g., ATM Forum LAN Emulation Protocol).

— The routing tables in the DSLAM are configured such that the next-hop

address field points to the IP address of the NSP premises router for all authorized subscriber IP source addresses.(See the discussion on source-based routing in Chapter 6,

IP Routing

— A different next-hop router is specified for each NSP address domain in

contrast to the routed network case where a single next-hop router was specified for all NSP domains. If the DSLAM does not know the MAC address of the NSP premises router, it uses ARP to obtain the MAC address from the NSP premises router prior to forwarding the packet (i.e., the wire center VLAN switch forwards an ARP request over the WAN to the NSP router).

— Packets flowing from the NSP network to the subscribers are routed to

the subscriber based on the destination IP address of the subscriber as is most common for IP-routed networks. In this case, the LAN on which the DSLAM resides appears to be part of a local subnet connected directly to the NSP premises router. If the NSP router does not know the MAC address of the subscriber, it uses ARP to obtain the MAC address from the DSLAM that acts as a proxy for the subscriber. (See the discussion on proxy ARP in Chapter 2,

Service Domain Features

8000-A2-GB21-20

November 1997

1-15

Page 25

Introduction to the HotWire DSLAM

Understanding the Domain Types

Functionally, the HotWire DSLAM network model can be divided into:

 Features supporting customers

Features integral to supporting customers are the DSL cards and HotWire RTUs.

 Features supporting overall system management

The central point of access for overall system management is the MCC card. However, the features integral to supporting overall system management are also distributed throughout the HotWire DSLAM and the HotWire RTUs.

To monitor and control the operation of the overall system, the IP addresses of the HotWire DSLAM and the HotWire RTU must be partitioned into two distinct domains.

 Service domain(s)

The service domain (also known as the NSP domain) resides in a mutually exclusive domain from that of the management domain. (There should be one service domain for each NSP served by the HotWire DSLAM.) One service domain encompasses an NSP and all of the end-user systems that subscribe to that NSP.

For more information about the service domain, its features and components, see Chapter 2,

Network Model

Service Domain Features

, and Chapter 4,

Components of the

 Management domain

The management domain resides in a mutually-exclusive domain from that of the service domains. The NAP provisions IP addresses for the management domain.

For more information about the management domain, its features and components, see Chapter 3,

Components of the Network Model

For more information about assigning IP addresses, see Chapter 5,

Allocation

Management Domain Features

, and Chapter 4,

IP Address

1-16

November 1997

8000-A2-GB21-20

Page 26

Service Domain Features

Overview

This chapter describes the following features that are supported in the service domain:

 Protocols  Address Resolution Protocol (ARP) with Proxy ARP  Filtering

Protocols

The HotWire DSLAM and HotWire RTU forward IP packets between the end-user system and the Network Service Provider using the following protocols:

 Point-to-Point Protocol/High-level Data Link Control (PPP/HDLC)

Packets transmitted over DSL links on an 8546 DSL card are encapsulated in PPP/HDLC frames. PPP/HDLC is not supported on the 8540 DSL card.

 MAC

Packets transmitted over LAN ports are encapsulated in Ethernet II MAC frames.

 IP

IP packets arriving over the DSL interface are forwarded to the LAN interface. IP packets arriving over the LAN interface are forwarded to the appropriate DSL interface.

NOTE:

Directed broadcasts

1s (ones) in the host field) are forwarded upstream, but are not forwarded downstream.

Also, multicasting is not supported.

(also referred to as

subnet broadcasts —

all

8000-A2-GB21-20

November 1997

2-1

Page 27

Service Domain Features

 Internet Control Management Protocol (ICMP)

In general, ICMP is supported. However, the options field is not reflected back if the HotWire DSLAM is the destination address (i.e., the HotWire DSLAM receives the data and then returns the packet without the options field). The HotWire DSLAM does, however, pass the packet with the options field to the next hop if the DSLAM is not specified as the destination address.

 Dynamic Host Configuration Protocol (DHCP)

DHCP is the protocol used for automatic IP address assignment. A DHCP discover or request message from an end-user system is transmitted over DSL ports and forwarded to the designated DHCP server, which is typically maintained and operated by the NSP for its address domain. The DHCP server assigns an IP address to the end-user system. The HotWire RTU routing tables and the DSLAM routing tables are automatically updated with the IP address information by the DSLAM.

Proxy ARP (Theory of Operation)

An Address Resolution Protocol (ARP) request is used to dynamically bind an IP address to a MAC address. Proxy ARP is a technique by which a router answers ARP requests intended for another machine by supplying its own MAC address (also referred to as the physical address). By answering for another device, the router accepts responsibility for forwarding packets to that device.

ARP is supported by the HotWire DSLAM and the HotWire RTU. Proxy ARP allows the end users to appear to be directly connected to the router or VLAN switch providing access to the NSP network. This is an advantage because routers connected to a device running proxy ARP require less configuration. The following scenarios show why this is an advantage.

Scenario 1: Without Proxy ARP

In this scenario, Router B does not have proxy ARP software and the networks of the default router (Router A) for workstation 1 (175.1.2.3) and workstation 2 (135.1.3.45) are different.

WS1

Workstation 1 (WS1) needs to send a packet to workstation 2 (WS2). For the packet to arrive successfully at WS2:

Router A

175.1.2.3

LAN A LAN B

Router B

175.1.2.6

135.1.3.9

WS2

135.1.3.45

97-15458-02

 There is a static route on Router A for WS2. The next hop is Router B and

the destination is WS2.

2-2

November 1997

8000-A2-GB21-20

Page 28

 WS1 sends a packet to Router A.  Router A consults its routing table to determine the next hop address (i.e.,

router IP address) for WS2 because WS2 is on another network (135.1.0.0).

Now that it knows the next hop address to Router B, Router A then ARPs for Router B. Router B receives the ARP request for its IP address and does an ARP reply with its MAC address. After Router A receives the ARP reply, it sends the packet to the Router B which, in turn, forwards it to WS2.

Scenario 2: With Proxy ARP

In this scenario, Router B is running the proxy ARP software, and WS2 and Router A for WS1 are on the same network (135.1.0.0).

WS1

Router A

135.1.2.3/

255.255.0.0

Service Domain Features

LAN A LAN B

Router B

135.1.2.6/

255.255.255.0

135.1.3.9/

255.255.255.0

WS2

135.1.3.45/

255.255.255.0

97-15459-02

WS1 again needs to send a packet to WS2. This time, however, Router B is running proxy ARP and knows that WS2 lies on LAN B on the same logical subnetwork as Router A (135.1.0.0). Router B uses proxy ARP to maintain the illusion that only one physical network exists. Router B keeps the location of WS2 hidden from Router A, allowing Router A to communicate as if directly connected to WS2.

NOTE:

Router A does not need a static route entry for the WS2 route because the two LANs appear to be one.

Therefore, when WS1 needs to send a packet to WS2, this is the sequence of events:

 WS1 sends a packet to Router A.  Router A invokes ARP to map the WS2’s IP address into a MAC address,

because WS2 appears to Router A to be on the same 135.1 subnet.

 Router B running proxy ARP software receives the broadcast ARP request

from Router A, knows that WS2 is on LAN B, and responds to Router A’s ARP request with its own MAC address.

 Router A receives the ARP reply, then sends the packet to the MAC address

of Router B.

 Router B then forwards the packet destined for WS2 on LAN B.

8000-A2-GB21-20

November 1997

2-3

Page 29

Service Domain Features

Filtering

NOTE:

The proxy ARP capability is card- or system-dependent and detailed examples for the MCC card, DSL card, and HotWire RTU are given in Chapter 4,

By default, filtering is disabled on the HotWire DSLAM system, but you can enable filtering to selectively filter source or destination packets being routed through the MCC or DSL cards. Filtering provides security advantages on LANs by restricting traffic on the network and hosts based on the IP source and/or destination address.

Components of the Network Model

For more information about filtering, see Chapter 7, information about dynamic IP addressing and the dynamic access control option, see Chapter 5,

8546 DSL Cards User’s Guide

IP Address Allocation

, and the

IP Filtering

HotWire DSLAM for 8540 and

. For more

2-4

November 1997

8000-A2-GB21-20

Page 30

Management Domain Features

Overview

This chapter describes the following features that are supported in the management domain:

 Network Management Systems (NMSs)  Applications for Diagnostics

Network Management Systems — SNMP and DCE Manager

You may want to use an SNMP NMS to simplify the operation and management of very large networks. In a UNIX environment you may use HP OpenView (UNIX) as your NMS with Paradyne’s DCE Manager, or in a Windows environment, HP OpenView (MS-Windows) with Paradyne’s DCE Manager. The HotWire DSLAM and HotWire RTUs provide features for the DCE Manager to allow you to monitor and manage your network from a central point.

The following lists some of the features of DCE Manager:

 Graphical User Interface (GUI) showing physical representation of the

HotWire DSLAM and each active card

 Multiple integrated functions to provide on-demand health and status

information

 Color-coded graphic representations to provide instant visual status  Loopback and pattern tests via telnet to help isolate problems quickly  Integrated management optimizes network performance and availability  Direct telnet support

8000-A2-GB21-20

November 1997

3-1

Page 31

Management Domain Features

These SNMP capabilities provided by Paradyne’s DCE Manager provide access to MIB II, Entity MIB, and private-enterprise MIB extensions to monitor information:

 From the HotWire DSLAM, and  To the MCC card, DSL cards, and HotWire RTUs.

The DSLAM uses a processor card called the MCC card in conjunction with DCE Manager. The MCC card provides the single management interface to the HotWire DSLAM cards and RTUs. The MCC card gathers operational status for each of the HotWire DSL cards in the DSLAM and RTUs, and reports events and alarms to the DCE Manager. For more information, see the

OpenView for Windows User’s Guide

Applications for Management Domain

The HotWire DSLAM user interface provides the following management applications:

or the

DCE Manager for HP

DCE Manager User’s Guide

Ping

 Ping  TraceRoute  Trivial File Transfer Protocol (TFTP) client  Telnet

The ping program is an IP-based application used to test reachability to a specific IP address by sending an ICMP echo request and waiting for a reply. It is supported from both the DSL and MCC cards. As a diagnostic tool, the ping program from the MCC card can be used to verify reachability in the management domain to the DSL card, the HotWire RTU, and the DCE manager. Similarly, invoking the ping program from the DSL card can test the service and management domains by verifying reachability downstream to the HotWire RTU and the end-user system (ES), and to verify reachability upstream to the NSP.

NOTE:

Record route and other ICMP options facilitating trace route are also supported. However, the options field is not reflected back if the HotWire DSLAM is the destination address (i.e., the HotWire DSLAM receives the data and then returns the packet without the options field). The HotWire DSLAM does, however, pass the packet with the options field to the next hop if the DSLAM is not specified as the destination address.

For more information, see Chapter 8,

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

3-2

November 1997

Diagnostics and Troubleshooting

8000-A2-GB21-20

, of the

Page 32

TraceRoute

Management Domain Features

The TraceRoute program is a TCP/IP diagnostic tool that allows you to learn the path a packet takes from its local host to its remote host. If you are unable to ping a device in a HotWire network configuration, you may want to run TraceRoute to identify the links (destinations up to 64 hops) between the DSL card and an RTU as well as which device is not forwarding the ping message.

TFTP Client

Telnet

For more information, see Chapter 8,

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

The MCC card and DSL cards in the DSLAM provide client TFTP applications that work with the firmware download and configuration upload or download features. TFTP sessions are established between the MCC card or the DSL card to a TFTP server accessible through the LAN interface.

A recommended use for configuration transfers is to upload a DSL card configuration to save (archive) the configuration set. Then, if necessary, you can recover the configuration by downloading (restoring) the saved configuration.

For more information, see Chapter 5,

Card Configuration

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

the

The HotWire DSLAM system provides support for telnet, which is a simple remote terminal protocol that is part of the TCP/IP protocol suite. With telnet, a network administrator can establish a virtual access connection to the HotWire DSLAM from a remote client to configure or monitor the HotWire DSLAM. The user interface presented for a telnet session is the same as that used with the DSLAM’s local serial port.

, and Appendix C,

Diagnostics and Troubleshooting

MCC Card Configuration

, Chapter 6,

Download Code and Apply Download

, of the

DSL

, of

A telnet connection from the HotWire DSLAM to another HotWire DSLAM or remote server is also supported. This feature is supported from the Ethernet (10BaseT) interface on the MCC card.

For more information, see Chapter 8,

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

8000-A2-GB21-20

November 1997

Diagnostics and Troubleshooting

, of the

3-3

Page 33

Components of the Network

Model

Overview

The service and management domains logically comprise the network model. This chapter describes the components that comprise these domains.

Service Domain Components

The primary purpose of the service domain network is to provide IP routing of customer data between the Network Service Provider (NSP) and the end-user system (ES).

The basic service domain configuration consists of the following components:

 An end user (PC or workstation) or multiple users on an Ethernet LAN

connected to the HotWire RTU, which in turn, is connected to one of the DSL card ports of the HotWire DSLAM

 The 10BaseT port of the HotWire DSLAM DSL card connected to a router or

switch that may also reside in the Central Office (CO) or wire center

 The router or switch is then connected to the NSP typically over a Wide Area

Network (WAN)

 The NSP may also be directly connected to the same LAN as the DSL card

End-user

System

97-15461-0

NSP

WAN

Router

or VLAN

Switch

10BaseT

DSL

Card

DSL/POTS

HotWire

RTU

10BaseT

8000-A2-GB21-20

November 1997

4-1

Page 34

Components of the Network Model

The following illustration shows another internetworking configuration. This configuration has multiple end users connected to the HotWire RTU using a hub. The number of supported end-user systems depends on whether you use a host or structured subnetting. For more information, see Chapter 5,

Allocation

NOTE:

This illustration does not apply to the 5171 PC NIC and 5216 RTU. The 5171 PC NIC and 5216 RTU are for single end-user system configurations only. They do not support multiple end-user system configurations.

NSP

WAN

Router

or VLAN

Switch

10BaseT

DSL

Card

DSL/POTS

10BaseT

HotWire

RTU*

IP Address

End-user

System 1

LAN

End-user System 2

End-user

System 32

* This RTU can be a 5170 RTU,

5246 RTU or 5446 RTU.

97-15462-0

When multiple end users are connected, they may opt to access different NSPs, as illustrated on page 4-3. When all 18 DSL cards are used, the HotWire DSLAM can support simultaneous access up to 288 different NSPs or private intranets by the end users (16 NSPs or private intranets per DSL card).

4-2

November 1997

8000-A2-GB21-20

Page 35

Components of the Network Model

A maximally configured HotWire DSLAM system will have 18 DSL cards with each DSL card having its four ports connected to a HotWire RTU for a total of 72 modem ports. Each modem can connect via a hub to 32 active end-user systems to support a total of 2304 users.

NOTE:

The following illustration does not apply to the 5171 PC NIC and 5216 RTU. The 5171 PC NIC and 5216 RTU are for single end-user system configurations only .

WAN

10BaseT

NSP1

NSP2

NSP16

* This RTU can be a 5170 RTU,

5246 RTU or 5446 RTU

Router

or VLAN

Switch

DSL

Card

DSL/POTS

10BaseT

LAN

HotWire

RTU*

End-user

System 1

End-user

System 2

End-user

System 32

97-15463-01

Additionally, by setting up structured subnets behind each HotWire RTU, hundreds of active end-user systems can be supported by each 5446 RTU instead of 32. Careful network traffic analysis must be performed to determine if very large networks will have acceptable response times. For information on how to set up structured subnets, see Chapter 5,

IP Address Allocation

NOTE:

Usually a user is active only in one service domain at a time. However, if the end user’s system can be multihomed, it may be possible to be active in more than one NSP domain at a time. A multihomed system is a system with connections to two or more logical networks, which may be assigned to one or more physical networks.

8000-A2-GB21-20

November 1997

4-3

Page 36

Components of the Network Model

Proxy ARP

Proxy ARP is supported by the DSL cards and the HotWire 5446 RTU. It allows the end users to appear to be directly connected to the router providing access to the NSP network. This is an advantage because routers connected to a device running proxy ARP require less configuration. The following scenarios show why this is an advantage.

DSL Card Proxy ARP

When an ARP request is sent by an NSP connected to the DSL card 10BaseT interface for a downstream end-user system (one on the same IP network), the DSL card will proxy ARP for the ES. The following figure shows the packet flow when the NSP wants to send a packet to the ES.

DSL Card

155.1.3.2/

255.255.255.0

RTU

135.1.3.3

155.1.3.4

97-15470-0

NSP

155.1.2.2/

255.255.0.0

NSP: Sends packet to

Local Router

Local Router: ARP Request for ES

DSL: Proxy ARP (for ES)

Local Router: Sends packet to DSL

card, which forwards it to ES

Local Router

155.1.2.1/

255.255.255.0

155.1.3.1/

255.255.255.0

In this illustration:

 The local router receives the IP packet and does an ARP request for the ES.  The DSL card receives the broadcast ARP request. The DSL card does an

ARP reply for the ES by replying with its own MAC address. Addresses for which the DSL card will proxy ARP must be configured as part of static route configuration. See the

for more information.

Guide

HotWire DSLAM for 8540 and 8546 DSL Cards User’s

 When the local router receives the ARP reply, it sends the packet to the DSL

card, and the DSL card forwards it to the ES.

NOTE:

In certain network configurations, the use of proxy ARP on the DSL cards will cause HP OpenView to log a major event. This will happen since HP OpenView received the same IP address from two different MAC addresses.

By default, the HP OpenView system logs and displays all events. However, you may elect to filter specific unwanted events. Instructions on how to filter out these events are dependent on the release of HP OpenView/Netview that you are running. For detailed instructions, see the appropriate HP OpenView user documentation.

4-4

November 1997

8000-A2-GB21-20

Page 37

5446 RTU Proxy ARP

The HotWire 5446 RTU utilizes proxy ARP to enable connectivity between end systems that are attached to separate RTUs, but reside on the same subnetwork. The HotWire 5446 RTU will proxy ARP for the end-user system that is physically connected to another HotWire 5446 RTU where the destination end-user system is logically connected to the same subnetwork as the sender end-user system.

Management Domain Components

The following illustration shows the components of the network management domain. Note that the router between the MCC card’s 10BaseT interface and the DCE Manager is optional. The MCC card, as previously noted, provides consolidated management for the DSL cards and HotWire RTUs from remote network management workstations by means of SNMP, telnet, or by local access through its VT100-serial interface.

Router

135.1.3.254/255.255.255.0

135.1.2.1/255.255.255.0

DCE Manager

Components of the Network Model

Ordinarily, when DCE Manager

is on a separate subnetwork, it will not be in the 135.1.3 or

135.1.2 subnetwork.

RTU1

RTU2

RTU3

RTU4

10BaseT

135.1.2.2/255.255.255.0

MCC Card

135.1.3.1/255.255.255.0

Backplane

135.1.3.3/255.255.255.0

8540 DSL Card

DSL Port 1

DSL Port 2

DSL Port 3

DSL Port 4

DSLAM

135.1.3.2/255.255.255.0

8546 DSL Card

DSL Port 1

DSL Port 2

DSL Port 3

DSL Port 4

RTU1

135.1.3.4

RTU2

135.1.3.5

RTU3

135.1.3.6

RTU4

135.1.3.7

97-15464-0

8000-A2-GB21-20

November 1997

4-5

Page 38

Components of the Network Model

To facilitate management of the DSL cards and HotWire RTUs through the MCC card:

 Assign IP addresses from the management domain to the internal backplane

interfaces of each DSL card and 5446 RTU interface in the same subnet as the MCC card’s backplane interface (as shown in the previous illustration). This is a separate subnetwork from the MCC card’s 10BaseT port.

These IP addresses are stored in the Entity MIB on the MCC card where they can be accessed by the NMS.

NOTE:

Management functions of RTUs associated with an 8540 DSL card are performed by an internal agent on the 8540 DSL card.

 Provide IP addresses on the router’s interface attached to the MCC card for

both subnetworks, so that the router appears to be directly connected to the MCC card’s Ethernet interface as well as the HotWire DSLAM system backplane.

In other words, the router’s interface to the MCC can be multihomed to support proxy ARP.

Discovering Devices on the Network (Discovery)

In the illustration on page 4-5, the IP addresses assigned for the router’s interface to the MCC card are 135.1.2.1 and 135.1.3.254. The second IP address (135.1.3.254) is on the same subnetwork (135.1.3.0) as the internal addresses of the DSL cards and the HotWire RTUs. The MCC card will not forward broadcasts on the management network (135.1.2.n) across the HotWire DSLAM system backplane because it is a separate subnetwork, as the DSL cards do not need to

discovered

How does an NMS learn the address of a device beyond the MCC card?

 DCE Manager gets the IP address of a DSL card from the Entity MIB on the

MCC card. After the DCE Manager has learned the IP address of a DSL card through

the Entity MIB, it addresses management traffic directly to that card.

 DCE Manager gets the IP address of the 5446 RTU from the Entity MIB on

the DSL card. After the DCE Manager has learned the IP address of the RTU through the

Entity MIB, it addresses management traffic directly to that RTU.

When the HotWire DSLAM and HotWire RTU systems networks are configured as described above, the DCE Manager provides a view of the entire network from information contained in the MCC card’s entity MIB.

by the management system.

4-6

November 1997

8000-A2-GB21-20

Page 39

MCC Card

Components of the Network Model

MCC Card Proxy ARP

Proxy ARP is also supported by the MCC card. In a HotWire DSLAM network configuration, when an ARP request is sent by a device (such as a router) to the MCC card’s 10BaseT interface to resolve either the DSL card or HotWire 5446 RTU MAC address, the MCC card will proxy ARP for those devices so long as their IP addresses are on the same network (135.1.3.n) as the backplane interface of the MCC card. The MCC card responds to these ARP requests with its own MAC address (proxy ARP). Incoming packets are then forwarded to that appropriate DSL card across the HotWire DSLAM system backplane.

DSL Card 1

RTU RTU RTU RTU RTU RTU

DSL Card 2

97-1549

NOTE:

It is not recommended that the DCE Manager access a DSL card via its Ethernet port because the Entity MIB on the DSL card does not reflect a view of the entire HotWire DSLAM system. It reflects only the view of the DSL card

discovered

discovered and appear on your network map. If you want to manage DSL devices across the NSP network, use telnet. For

more information on telnet see Chapter 8, of the

. Also, in a fully configured DSLAM, 18 additional devices will be

Diagnostics and Troubleshooting

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

8000-A2-GB21-20

NOTE:

Management functions of RTUs associated with an 8540 DSL card are performed by an internal agent on the 8540 DSL card.

November 1997

4-7

Page 40

Components of the Network Model

The following illustration shows the packet flow when the DCE Manager wants to send a packet to the HotWire 5446 RTU using proxy ARP.

Local Router

DCE

Manager

135.1.1.1/

255.255.0.0

DCE Manager:

Sends packet to

Local Router

MCC: Proxy ARP (for RTU)

Local Router: Sends packet to

RTU via the MCC and DSL cards

135.1.1.2/

255.255.0.0

255.255.255.0

ARP Request for RTU

135.1.2.1/

135.1.3.254

Local Router:

MCC Card

e1a:135.1.2.2/

255.255.255.0 s1b:135.1.3.1/

255.255.255.0

8546 DSL

Card

s1b:135.1.3.2/

255.255.255.0

5446

RTU

135.1.3.3

97-15465-02

In this illustration:

 The local router does an ARP request to acquire the HotWire 5446 RTU MAC

address.

 The MCC card is in the same network (135.1.3.1) and sees the ARP request.

The MCC card contains a route to the HotWire RTU and knows the RTU is downstream (generally a host route). The MCC card does an ARP reply for the HotWire 5446 RTU by responding with its own MAC address.

 When the local router receives the ARP reply, it forwards the packet to the

MCC card. The MCC card forwards it to the DSL card which forwards it to the HotWire 5446 RTU.

For security reasons, a separate management domain is recommended. However, the management and service domains can share the same subnet. Separation can be maintained by extending the subnet mask down to the fourth octet (255.255.255.255).

For example, one management subnet and three service domain subnets could use the combined subnet mask: 135.1.2.0/255.255.255.0. The management subnet could be 135.1.2.192/255.255.255.192 and service domain subnets could be 135.1.2.0/255.255.255.192, 135.1.2.64/255.255.255.192, and

135.1.2.128/255.255.255.192. With these subnet masks, management addresses use the top quarter of the range (135.1.2.192 through 135.2.2.254) and service addresses use the lower three-quarters (135.1.2.1 through 135.1.2.191).

4-8

November 1997

8000-A2-GB21-20

Page 41

IP Address Allocation

Overview

IP addresses are assigned throughout the network model for components comprising both the service and management domains. This chapter describes the IP address allocation schemes for the components that make up the HotWire DSLAM network model. It also describes the naming convention used for the HotWire DSLAM system interfaces.

Port Naming Convention

The following is the naming convention used for the HotWire DSLAM interfaces:

NOTE:

Interfaces are sometimes referred to as ports. The term usually is reserved for referring to the physical layer attributes of an interface.

 e1a — Interface name of the DSLAM system 10BaseT interface on the MCC

and each DSL card.

 s1b — Interface name of each card’s interface to the DSLAM system

backplane bus.

 s1c, s1d, s1e, and s1f — Interface names of the four DSL ports on a DSL

card.

NOTE:

These names are used throughout the remainder of this guide to reference the HotWire DSLAM interfaces. These are also the names used in the HotWire DSLAM software when configuring the HotWire DSLAM system.

ports

, however,

8000-A2-GB21-20

November 1997

5-1

Page 42

IP Address Allocation

The following illustrates the logical interface naming convention.

DSLAM System

10BaseT

Interface to

Management

Network

e1a

MCC Card

s1b

System

Backplane

Bus

s1b

e1a

s1b

e1a

DSL

Card 1

DSL

Card 2

s1c s1d s1e

s1f

s1c s1d s1e

s1f

DSL Port 1 DSL Port 2 DSL Port 3 DSL Port 4

10BaseT Interface to NSP

DSL Port 1 DSL Port 2 DSL Port 3 DSL Port 4

10BaseT Interface to NSP

97-15467-01

Assigning IP Addresses

In the HotWire DSLAM network model, there are two distinct domains:

 a management domain  a service domain

Within the management domain, there are two subnets as described in

Management Domain Components

of two IP address allocation schemes can be followed: host addressing or structured subnet addressing. The following sections describe these schemes.

5-2

November 1997

in Chapter 4. Within the service domain, one

8000-A2-GB21-20

Page 43

Host Addressing

IP Address Allocation

Host addresses within the service domain are assigned to end-user systems. Because they are host addresses, they have a subnet mask of 255.255.255.255 and can be geographically dispersed. (When structured subnet addressing is discussed in the next section, you will see how IP addresses are allocated to certain areas.) This conserves address space, but may not scale well to large numbers of end-user systems. Manual configuration is required for every host address and routing performance may be decreased.

The following illustration is an example of host addressing.

200.200.200.n /

255.255.255.0

= Any valid IP address

Structured Subnet Addressing

As an alternative to using host routes for end-user systems, structured subnetting can be used. It scales better and performs better, but it does not allow geographically dispersed subnets.

DSL Card

DSL Port 1

LAN Port

DSL Port 2

DSL Port 3

DSL Port 4

RTU1

RTU2

RTU3

RTU4

200.200.200.1 /

255.255.255.255

200.200.200.2 /

255.255.255.255

200.200.200.3 /

255.255.255.255

200.200.200.4 /

255.255.255.255

ES1

ES2

ES3

ES4

97-15501-0

NOTE:

Structured subnetting is supported on the 8546 DSL card and the 5446 RTU. It is not supported, however, on the 8540 DSL card and its corresponding RTUs on the DSL ports.

Structured subnet addressing uses the following method:

 Within the service domain, the NSP would provision a subnet of its domain to

a DSL card and all devices behind it.

 The NSP would further subdivide that subnet into four additional subnets

(one behind each DSL port).

8000-A2-GB21-20

November 1997

5-3

Page 44

IP Address Allocation

The following illustration is an example of structured subnet addressing.

DSL Card

200.200.200.240 /

200.200.200.n /

255.255.255.0

DSL Port 1

LAN Port

DSL Port 2

DSL Port 3

DSL Port 4

RTU1

RTU2

RTU3

RTU4

255.255.255.240

200.200.200.224 /

255.255.255.240

200.200.200.208 /

255.255.255.240

200.200.200.192 /

255.255.255.240

= Any valid IP address, but not within the other subnets

97-15466-0

To understand why this subnetting scheme works, you may want to consider the IP addresses and subnet masks in hexadecimal:

Dotted Decimal

200.200.200.00 / 255.255.255.0 C8.C8.C8.00 / FF.FF.FF.00

200.200.200.240 / 255.255.255.240 C8.C8.C8.F0 / FF.FF.FF.F0

200.200.200.224 / 255.255.255.240 C8.C8.C8.E0 / FF.FF.FF.F0

200.200.200.208 / 255.255.255.240 C8.C8.C8.D0 / FF.FF.FF.F0

200.200.200.192 / 255.255.255.240 C8.C8.C8.C0 / FF.FF.FF.F0

Dotted Hexadecimal

In the previous illustration:

 Each of the four DSL ports is on a different subnetwork of size 16, and the

subnet mask for the four ports is 255.255.255.240.

 The LAN port (10BaseT port) IP address is 200.200.200.

(where n can be any valid IP address, but cannot be an IP address within the other subnets), and its subnet mask is 255.255.255.0.

The illustration on page 5-5 shows one NSP connected to one DSL card, while the illustration on page 5-6 shows 16 NSPs connected to one DSL card. On page 5-6, the NSP router is multihomed to support all 16 NSPs. Also, each RTU has 32 end-user systems (ES).

In summary, if 32 end-user systems are connected to the DSL card’s port 1 (s1c) and all are using host addressing, then 32 host routes must be configured on the RTU. If they are using structured subnet addressing, then only one route is configured on the 5446 RTU. Remember that structured subnet addressing applies only to the 5446 RTU.

5-4

November 1997

8000-A2-GB21-20

Page 45

IP Address Allocation

ES1: 155.1.3.4/255.255.255.0

ES2: 155.1.3.5/255.255.255.0

10BaseT

RTU*

135.1.3.3/

255.255.255.255

...

ES3: 155.1.3.6/255.255.255.0

ES4: 155.1.3.7/255.255.255.0

155.1.3.3/

255.255.255.0

DSLAM

ES32: 155.1.3.35/255.255.255.0

ES1: 155.1.3.37/255.255.255.0

ES2: 155.1.3.38/255.255.255.0

ES3: 155.1.3.39/255.255.255.0

10BaseT

RTU*

155.1.3.36/

255.255.255.0

135.1.3.4/

255.255.255.255

ES1

...

ES4: 155.1.3.40/255.255.255.0

ES32: 155.1.3.68/255.255.255.0

s1c

DSL

s1d

DSL

s1e

DSL

ES2

DSL

s1f

ES3

ES1

ES2

...

ES4

10BaseT

RTU

ES3

ES32

...

ES4

10BaseT

97-15475a

ES32

RTU

DCE Manager

Server

10BaseT

Router

135.1.3.254/255.255.255.0

135.1.2.1/255.255.255.0

MCC Card

e1a: 135.1.2.2/

s1b: 135.1.3.1/

255.255.255.0

System Backplane

255.255.255.0

WAN

NSP1

s1b: 135.1.3.2/

NSP Router

155.1.2.2/

255.255.0.0

DSL Card*

255.255.255.0

155.1.3.1/

155.1.2.1/

255.255.255.0

IP Interface

155.1.3.2/

255.255.255.0

e1a:

associated RTUs will not have an IP address.

*If DSL card is an 8540 DSL card,

8000-A2-GB21-20

November 1997

5-5

Page 46

IP Address Allocation

10BaseT

RTU

135.1.3.3/

255.255.255.255

ES1: 155.1.3.4/255.255.255.0

ES2: 156.1.3.4/255.255.255.0

155.1.3.3/

255.255.255.0

...

ES3: 157.1.3.4/255.255.255.0

ES4: 158.1.3.4/255.255.255.0

ES32: 158.1.3.11/255.255.255.0

156.1.3.3/

157.1.3.3/

158.1.3.3/

255.255.255.0

DSLAM

ES1: 159.1.3.4/255.255.255.0

ES2: 160.1.3.4/255.255.255.0

ES3: 161.1.3.4/255.255.255.0

ES4: 162.1.3.4/255.255.255.0

10BaseT

159.1.3.3/

RTU

135.1.3.4/

255.255.255.255

160.1.3.3/

255.255.255.0

...

ES32: 162.1.3.11/255.255.255.0

161.1.3.3/

162.1.3.3/

255.255.255.0

s1c

DSL

s1d

DSL

ES1

ES2

255.255.255.0

DSL

s1e

s1f

ES3

ES1

ES2

...

ES4

10BaseT

RTU

ES3

ES32

...

ES4

10BaseT

RTU

ES32

97-15475-02

DCE Manager

Server

10BaseT

Router

135.1.3.254/255.255.255.0

135.1.2.1/255.255.255.0

e1a: 135.1.2.2/

WAN

NSP1

155.1.2.2/

MCC Card

s1b: 135.1.3.1/

255.255.255.0

255.255.0.0

255.255.255.0

NSP2

156.1.1.1/

255.255.0.0

System Backplane

NSP Router

NSP3

157.1.1.1/

255.255.0.0

s1b: 135.1.3.2/

255.255.255.0

155.1.3.1/

155.1.2.1/

255.255.255.0

NSP4

158.1.1.1/

255.255.0.0

DSL Card

IP Interface

155.1.3.2/

e1a:

...

170.1.3.1/

255.255.255.0

NSP5

255.255.255.0

159.1.1.1/

255.255.0.0

...

255.255.255.0

156.1.3.2/

255.255.255.0

NSP6

NSP7

160.1.1.1/

255.255.0.0

161.1.1.1/

170.1.3.2/

255.255.255.0

NSP8

162.1.1.1/

255.255.0.0

...

255.255.0.0

NSP16

170.1.1.1/

255.255.0.0

5-6

November 1997

8000-A2-GB21-20

Page 47

Management IP Address Allocation

The primary functionality of the management domain is monitoring and configuring the network. To provide this capability, IP addresses must be allocated for the components that are monitored and configured by an NMS and MCC card.

IP Address Allocation

Component

MCC Card The MCC card must have two IP addresses:

DSL Card Each DSL card must have one management IP address in

HotWire 5446 RTU Each 5446 RTU must have one management IP address

IP Address Requirement

 One IP address for connectivity to the NMS or Router

(connecting to the NMS). This address is also known as the Router ID.

 One IP address to communicate to the DSL cards (over

the s1b system backplane interface) in the HotWire DSLAM chassis.

These two IP addresses must be on separate subnetworks of the NMS domain. That is, they can be on:

 Completely separate networks (e.g., 135.1.0.0/16 and

143.1.0.0/24).

 Completely separate subnets (e.g., 135.1.1.0/24 and

135.1.2.0/24), or

 Subnets of the management domain (e.g., 135.1.0.0/16

and 135.1.2.0/24).

the same subnetwork as the MCC card’s system backplane IP address.

NOTE: The backplane subnet cannot be set to the

same e1a subnet on that DSL card.

in the same subnetwork as the MCC card’s system backplane IP address.

NOTE: Since there could be four HotWire 5446 RTUs

per DSL card and 18 DSL cards per HotWire DSLAM, a maximally-configured system would have 72 HotWire 5446 RTU management IP addresses. These must be in the same subnetwork as the MCC card’s system backplane interface and the 18 DSL cards’ management IP addresses (total of 91 addresses).

NOTE:

Management functions of RTUs associated with an 8540 DSL card are performed by an internal agent on the 8540 DSL card.

To configure the MCC card, the DSL card management IP addresses, and the HotWire 5446 RTU management IP addresses, use the HotWire DSLAM user interface. For step-by-step instructions, see Chapter 4,

DSLAM

8000-A2-GB21-20

, of the

Configuring the HotWire

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

November 1997

5-7

Page 48

IP Address Allocation

Peer IP Addresses

The s1b backplane ports are configured with peer IP addresses. Peer IP addresses are used to indicate directly-connected systems.

 For the MCC card’s s1b (backplane) interface, the peer IP address should be

set to indicate the subnet encompassing the DSL cards and RTUs. The following illustration shows a HotWire DSLAM system configured with

one MCC card and four DSL cards.

DSLAM System

DSL Card 1

DSL Card 2

DSL Card 3

DSL Card 4

97-15468-01

MCC Card

s1b base address:

135.1.3.1/255.255.255.0 peer IP address:

135.1.3.0 net

System Backplane Bus

s1b: 135.1.3.2/255.255.255.0

s1b: 135.1.3.7/255.255.255.0

s1b: 135.1.3.12/255.255.255.0

s1b: 135.1.3.17/255.255.255.0

— The IP address of the MCC card’s s1b interface is 135.1.3.1. — The IP addresses of the DSL card’s s1b interfaces are all in the same

subnet (135.1.3).

— Therefore, the directly connected peer subnet is its peer IP address,

135.1.3.0.

NOTE:

For structured subnetting on the backplane, the peer IP address must be the first in the subnet. For example, if s1b has an IP address of 135.1.3.65 and a subnet mask of 255.255.255.192, then the peer IP address must be

135.1.3.64.

 For the DSL card’s s1c through s1f interfaces, the peer IP address should be

set to indicate the management IP address of the directly connected 5446 RTU. (Peer IP addresses need to be set for 5446 RTUs only. They do not apply to any other RTU type.)

The peer address for the DSL card is a host route because the peer address identifies a specific 5446 RTU. Specifically, the peer address of each DSL card’s DSL port is the HotWire 5446 RTU’s management IP address. The peer address is assigned to the 5446 RTU through Internet Protocol Control Protocol (IPCP) negotiation.

5-8

November 1997

8000-A2-GB21-20

Page 49

IP Address Allocation

The following illustration shows the DSL card with four 5446 RTUs connected to its DSL ports. The peer address for the four DSL card ports are:

— s1c = 135.1.3.3 — s1d = 135.1.3.4 — s1e = 135.1.3.5 — s1f = 135.1.3.6

DSL Card

DSL Port 1 (s1c)

DSL Port 2 (s1d)

DSL Port 3 (s1e)

DSL Port 4 (s1f)

RTU1

135.1.3.3

RTU2

135.1.3.4

RTU3

135.1.3.5

RTU4

135.1.3.6

97-15469-0

8000-A2-GB21-20

November 1997

5-9

Page 50

IP Address Allocation

Service IP Address Allocation

Each NSP allocates IP addresses for the components in each service network as described below. How the IP addresses are allocated is also noted.

Component

Service Domain Router The router that routes NSP traffic to the HotWire DSLAM

DSL Card Each DSL card can support 16 NSP domains (four for

HotWire 5446 RTU Each HotWire 5446 RTU can support four NSP domains.

IP Address Requirement

DSL cards must have one IP address in each service domain. The router should be multihomed on its LAN port connection to the HotWire DSLAM.

Since 16 service domains are supported per DSL card and there can be 18 DSL cards per HotWire DSLAM, up to 288 NSP IP addresses may be required on the router’s interface to support a maximally configured HotWire DSLAM system. However, typically you would organize your domains in such a way that fewer IP addresses would be needed.

each HotWire RTU). For each different NSP supported by the DSL card, there must be an IP address in the same domain for the DSL card 10BaseT interface (e1a). Therefore, the total number of DSL card IP addresses required is determined by the number of NSPs supported by the HotWire RTUs.

Each HotWire 5446 RTU with an end-user system in the domain of an NSP must have one service domain IP address in the same subnetwork.

There could be:

 Four service domain IP addresses per HotWire 5446

RTU,

 Four HotWire 5446 RTU per 8546 DSL card, and  18 DSL cards per HotWire DSLAM.

This means that a maximally-configured HotWire DSLAM system with 72 HotWire 5446 RTUs could have 288 service domain IP addresses.

To configure the HotWire 5446 RTU service domain IP addresses, use an SNMP application, such as Paradyne’s DCE Manager. They are also configurable using Paradyne’s IP Injection Tool or from any MIB browser.

End-User System (ES) Each end-user system must have an IP address.

5-10

The IP address is assigned by the NSP either statically or dynamically . For information about dynamic IP addressing, see the following section.

November 1997

8000-A2-GB21-20

Page 51

Dynamic IP Addressing

The HotWire DSLAM system allows the use of Dynamic Host Configuration Protocol (DHCP) to facilitate the automatic assignment of end-user system IP addresses. With the dynamic IP addressing feature, NSPs can administer IP addresses to the end users dynamically (automatically) rather than statically (manually). An IP address can be reused once the end user no longer requires the address (i.e., the end-user system no longer requires access to the NSP) or the lease time has expired. This feature allows NSPs to maintain a pool of IP addresses that services many end users rather than one fixed IP address per end-user system.

In addition, an authentication feature can also be configured to confirm an enduser system’s access location.

IP Address Allocation

For more information about dynamic routes, see Chapter 6,

Recording Your Configuration Settings

It is recommended that you keep a record of your configuration settings when assigning IP addresses to the devices on your network. Appendix A,

Configuration Worksheets

settings. Store the worksheets for reference, as needed. You may also save your configuration settings (download your code) on the TFTP

server. For more information, see Chapter 5, Chapter 6,

Download

DSL Card Configuration

, of the

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

, contains the worksheets to help you record those

MCC Card Configuration

, and Appendix C,

IP Routing

Download Code and Apply

Network

8000-A2-GB21-20

November 1997

5-11

Page 52

IP Routing

Overview

Routing Table

This chapter presents information regarding the theory behind the configuration of routes (static and dynamic) on the HotWire DSLAM, as well as examples. Both standard destination-based routes and source-based routes are described.

The routing table stores information about possible destinations for packets that are routed through the HotWire DSLAM. It also identifies the next hop address to which to send the packet. The MCC, DSL cards, and RTUs maintain their own routing tables. There are two types of routes: static and dynamic. A static route is a permanent entry into the routing table that is manually entered. A dynamic route is an automatically forced (assigned) entry into the routing table; it does not need to be manually entered. Static routes can be destination based or source based. However, dynamic routes can be only destination based.

Although the HotWire DSLAM routing table supports both destination-based routing and source-based routing, this section discusses destination-based routing only. (Source-based routing is discussed later in this chapter.)

The routing table is comprised of:

 Configured routes (static and/or dynamic)  Routes learned by implication of directly connected hosts/networks  Routes learned by the MCC card from the DSL about its directly connected

hosts (RTUs)

8000-A2-GB21-20

November 1997

6-1

Page 53

IP Routing

With destination-based routing, the destination address of the packet being sent is compared to the destination address entries in the routing table. The destination address could possibly match one or more of three types of addresses in the routing table. It could match a:

 Host route address (that is, a specific destination IP address) e.g., 135.1.3.5,

 Subnet route, e.g., 135.1.3.0, or  Network route, e.g., 135.1.0.0

If a match is found for more than one destination address, the order of precedence is:

1. Host route

2. Subnet route

3. Network route

4. Default route

Therefore, the packet is sent to the next-hop address specified for that destination which matches and has the highest precedence.

A packet routed through the HotWire DSLAM that has a destination address not matching any entry in the routing table is dropped unless a default route is specified. If a default route is specified using the conventional address 0.0.0.0 as the destination IP address, the packet is sent to the associated next-hop address.

Static Routes for Static IP Addressing

If you plan to use static addressing, then you will need to create static routes to route to the end-user systems. Use the following routing table form:

Host/Net, Subnet Mask, Next Hop, Pref, S/D, PA

Where:

 The

Host/Net

— A host address (for example, the specific IP address of an RTU or

end-user system), or — A subnet or network portion of a destination or source IP address, or — The default route, which is defined to be 0.0.0.0.

Subnet Mask

default routes.

is one of the following:

for host, subnet, or network. This is not applicable to

 The

forwarded. For example, the IP address of the router connected to the LAN or the HotWire RTU.

6-2

Next Hop

is the IP address to which the given datagram should be

November 1997

8000-A2-GB21-20

Page 54



Pref

indicates the measurement of preference of one route to another, if you

have two routes going to the same destination. (The lower the number the more preferable.) This route is compared to others for the same address.

S/D



indicates if the address in the

Host/Net

field is a source address or a

destination address.

(proxy ARP) indicates whether or not the router answers ARP requests



intended for another machine.

IP Routing

For more information about the routing table, see the

and 8546 DSL Cards User’s Guide

MCC Card Static Route Example

The following illustration shows an example of the MCC card routing table.

DCE

Manager

135.1.1.1

MCC Routing Table

Host/Net Subnet Mask

1) 135.1.3.4*

2) 0.0.0.0

This entry is automatically generated and does not need to be statically configured. The entry also

automatically activates proxy ARP.

Router

135.1.1.2

135.1.2.1

135.1.3.254

255.255.255.255

0.0.0.0

HotWire DSLAM for 8540

MCC Card

e1a:135.1.2.2

Next-Hop Address S/D (Source/Destination)

135.1.3.1

135.1.2.1

DSL Card

s1b:135.1.3.1

dst (destination) dst (destination)

Unnumbered Interface

RTU

135.1.3.4

97-15478-0

In this example, the IP address of the MCC card’s management e1a IP address is

135.1.2.2.  A packet being routed from the RTU to the NMS is routed using route #2

because no routes for the packet (i.e., destination 135.1.1.1) are specified. Therefore, the default route is used as the next hop address.

 A packet sent by NMS to the RTU is routed using route #1 because the

destination IP address of the packet matches the route’s Host/Net/Subnet entry (135.1.3.4). Therefore, the next-hop address would be the DSL card (135.1.3.1).

Note also that the router is multihomed so that both the MCC card’s and the DSL card’s (management domain) subnetworks appear local (i.e., 135.1.2 and 135.1.3).

8000-A2-GB21-20

November 1997

6-3

Page 55

IP Routing

DSL Card Static Route Example

The following illustration shows an example of how static routes configured on a DSL card are used in its routing table:

NMS

Ethernet Interface

NSP

155.1.2.2/

255.255.0.0

*If DSL card is an 8540 DSL card,

associated RTU will not hav e an IP address.

DSL Routing Table

Host/Net

1) 155.1.3.4

2) 135.1.2.0

3) 0.0.0.0

Subnet Mask

255.255.255.255

255.255.255.0

0.0.0.0

Router

155.1.2.1/

255.255.0.0

155.1.3.1/

255.255.255.0

8540 DSL Card NextHop Address

s1c

135.1.3.1

155.1.3.1

In this example:

135.1.2.2/

255.255.0.0

MCC Card

135.1.3.1/

255.255.0.0

DSL Card*

155.1.3.2/

255.255.0.0 s1c

8546 DSL Card NextHop Address

135.1.3.3

135.1.3.1

155.1.3.1

Unnumbered DSL Interface

135.1.3.3

S/D (Source/ Destination)

dst (destination) dst (destination) dst (destination)

RTU*

155.1.3.4

PA (Proxy ARP)

Y (yes) N (no) N (no)

97-15471-0

 The DSL card’s Ethernet port is connected to the router’s port, which has an

IP address of 155.1.3.1. Packets being routed in the upstream direction (to an NSP) would use the

third routing table entry; i.e.,

Next Hop

They would use this route because no other destination would match.

 The management domain IP address of the RTU is 135.1.3.3 and the IP

address of the ES is 155.1.3.4. Packets being routed downstream use the first routing table entry, i.e., address of 135.1.3.3. Note that this is a host route.

 The second routing table entry is for upstream routing to the NMS via the

MCC card. Note that this is a subnet route.

6-4

Host/Net

address of 155.1.3.1.

Host/Net

November 1997

IP address 0.0.0.0 (by definition) and a

IP address of 155.1.3.4 and a

Next Hop

8000-A2-GB21-20

Page 56

Dynamic Routes for Dynamic IP Addressing

Alternatively, NSPs can administer IP addresses to the end users dynamically (automatically) rather than statically (manually).

The dynamic IP addressing feature consists of the following components:

 DHCP relay agent

The DSL card in the DSLAM acts as a DHCP relay agent. The DHCP relay agent is an intermediary function between the end-user system and the DHCP server. Its functions are to:

— Detect and forward a DHCP request message from an end-user system

to the appropriate DHCP server.

— If you configure the system for optional authentication:

1. Hold the DHCP request message.

2. Send an authentication request to the authentication server in RADIUS or XTACACS format.

3. Receive the authentication response. If negative, drop the held DHCP message. If positive, relay the held DHCP message to the DHCP server.

IP Routing

— Track the end-user system dynamically allocated IP address and lease

time from the DHCP acknowledgement by updating the routing table automatically.

 Local host (DSLAM) route injection

The DSL card’s routing table is used to determine the DSL port on which to forward incoming packets. This is achieved by examining the packet’s destination IP address and comparing it to the list of IP addresses in the routing table. The subnet masks are set to 255.255.255.255 for host routes.

The IP address and subnet mask are then used to determine the end-user system destination port. The DHCP relay agent adds (injects) the end-user system IP address and subnet mask into the routing table automatically.

 Remote host (RTU) route injection (when using an 8546 DSL card and a

5446 RTU only)

The DSLAM also injects the end-user system’s IP addresses into the 5446 RTU. The routing table in the 5446 RTU is used to determine if traffic on its 10BaseT (Ethernet) port is local or if it should be sent over the DSL. It determines this by checking if any of the addresses match the addresses in its local host routing table. This routing table is automatically updated by the 8546 DSL card after the DHCP relay agent has intercepted the end-user system’s IP address in the DHCP reply message.

8000-A2-GB21-20

November 1997

6-5

Page 57

IP Routing

 Automatic dynamic access control

The DSL card supports IP filters to validate user access to the NSP network. If the automatic dynamic access control feature is enabled, filters are configured automatically. The IP filters examine the IP source address of the upstream traffic to validate the end-user system’s IP address. This feature enhances security by preventing an end user from spoofing the IP address of another user on a different DSL port. The DSLAM checks the end-user’s IP address. If it does not match any valid IP addresses in the routing table, then the packet is dropped. Use the DHCP Relay Servers screen to enable this feature.

NOTE:

The DHCP server is typically maintained and operated by the NSP for its address domain. The HotWire RTU routing tables and the DSLAM routing tables are automatically updated by the DSLAM.

Also, an RTU will not be able to obtain its address dynamically if the DHCP server assigns an address for which there is a static route (destination) already configured on the card.

How Does Dynamic IP Addressing Work?

The following illustration shows an example of a basic IP address request and assignment. This illustration assumes there are no problems associated with the request or assignment of the IP address.

Authentication

Server

Authentication Response

DHCP Server

Original DHCP Request

Yes

DHCP ACK

DSLAM

DHCP Request

DHCP ACK

DHCP Release

End-user

System

97-15721

6-6

November 1997

8000-A2-GB21-20

Page 58

IP Routing

1. The end-user system requests an IP address by broadcasting a DHCP request message to the DHCP server.

2. The DSLAM performs a DHCP relay by acting as a DHCP relay agent. The DHCP relay function of the DSLAM acts as an intermediary between the end-user system and the DHCP server, and works with DHCP servers that support structured subnetting. At this point, the following events occur:

A. The DHCP relay within the DSLAM intercepts the end user’s DHCP

request for an address.

B. If a domain name is detected, the DHCP relay determines if the

domain has been configured to the DHCP server. C.It determines if authentication is required. If authentication is not required, it injects a gateway address into the

message and forwards it to the DHCP server. If authentication is required, the DHCP relay agent determines the

authentication type (Radius or XTACACS). Then, it holds the original message and creates an authentication request message and passes the message to the appropriate authentication server.

3. An authentication response is received by the DSLAM. If the authentication is confirmed, the DHCP relay agent inserts the gateway address (i.e., the e1a IP address associated with the domain name) into the original DHCP request message. Then, it forwards the message to the DHCP server within the configured service domain.

4. The original DHCP request (with the gateway address) is relayed to the DHCP server.

5. The DHCP relay function of the DSLAM intercepts the DHCP ACK (acknowledge) message. At this point, the following events occur:

— The DHCP relay agent extracts the IP address and lease time

information from the DHCP ACK message.

— The IP address is injected to the RTU (if the DHCP relay agent is an

8546 DSL card and the RTU is a 5446 RTU).

— The DHCP relay agent injects the IP address, subnet mask of

255.255.255.255, lease time, and port number into the routing table. The routing tables are updated automatically.

6. After successful completion of these events, the DHCP ACK message is forwarded to the end user.

7. The IP addresses are automatically deleted from the DSLAM routing tables when the end user releases the IP address (by sending a DHCP release message) or the lease time expires without a renewal. Once the DHCP relay has deleted the configuration information, the end user will no longer be able to access the NSP.

To regain access to the NSP, the end user must initiate a DHCP discover or request again to the DHCP server, and a new IP address will be assigned.

8000-A2-GB21-20

November 1997

6-7

Page 59

IP Routing

General DHCP Relay Agent Configuration

To configure a DHCP relay agent, you must do the following:

1. Make sure that the gateway address used in relaying DHCP requests is configured as an e1a address on the IP Network screen

Configuration→Interfaces→IP Network).

(

2. Assign domain names to the e1a addresses that will be used as DHCP gateway addresses. Assign these domain names on the Domain Names screen (

3. Configure the first eight NSP domain names on the Servers 1-8 screen (

Configuration→DHCP Relay Servers→Servers 1-8

domain names on the Servers 9-16 screen (

Servers

On the appropriate DHCP Relay Servers screen, you will need to enter the DHCP Server IP address for each domain. You will also need to determine whether or not you want to use the authentication feature. There are several fields that must be completed if you plan to use the authentication feature. In addition, you must also give the administrator of the authentication server some necessary information. See the following section,

Authentication Server Administrator

Configuration→DHCP Relay→Domain Names

→

Servers 9-16

, for more information.

) and the last eight NSP

Configuration→DHCP Relay

Notes to the

For detailed information about the various DHCP relay screens, see the

DSLAM for 8540 and 8546 DSL Cards User’s Guide.

Configuration Worksheets

record your network configurations for dynamic IP addressing.

, in this guide provides worksheets to help you plan and

Notes to the Authentication Server Administrator

If the authentication process is to be invoked as part of dynamic addressing, the authentication request from the DSLAM must be in either RADIUS or XTACACS format. The authentication server will receive an authentication request from the HotWire DSLAM before the end-user’s request for an address is relayed to the DHCP server.

NOTE:

The IP source address for these requests will be the e1a interface IP address associated with the domain.

The following sections describe the contents of the authentication request message for a RADIUS authentication server and an XTACACS authentication server.

Also Appendix A,

HotWire

Network

6-8

November 1997

8000-A2-GB21-20

Page 60

RADIUS Authentication

If the authentication server is a RADIUS server, an Access-Request message will have the following format:

 The user_name will be the end-user’s user ID as received by the DSLAM in

 The password will always be hotwire.

 The NAS-IP will be the DSL card’s e1a address (gateway address)

 The NAS-PORT will be the port number that received the end-user’s request.  The service type will be Authentication-Only.  The RADIUS Secret value used for encryption is configured on the DHCP

IP Routing

the type 0 client ID field of the DHCP request. If the end-user request does not contain a user ID, the corresponding domain

name is used as the user_name.

The passwords configured at the authentication server should not be set with an expiration time.

associated with this domain.

Relay Server screen.

The authentication request is sent to UDP port 1812 (as specified in RFC 2138). If an Access-Accept message is returned, the DHCP request is relayed to the

DHCP server.

XTACACS Authentication

If the authentication server is an XTACACS server, a Login message will have the following format:

 The user_name will be the end-user’s user ID as received by the DSLAM in

the type 0 client ID field of the DHCP request. If the end-user request does not contain a user ID, the corresponding domain

name is used as the user_name.

 The password will be the e1a IP address (gateway address) associated

with this domain in ASCII dotted decimal format. The passwords configured at the authentication server should not be set with

an expiration time.

 The local_line will be the port number that received the end-user’s

request.

The authentication request is sent to UDP port 49 (as specified in RFC 1492). If the authentication request is successful, the DSLAM sends a LOGOUT message

to the XTACACS server and the DHCP request is relayed to the DHCP server.

8000-A2-GB21-20

November 1997

6-9

Page 61

IP Routing

Source-Based Routing

In addition to destination-based routing, the HotWire DSLAM system also supports source-based routing. Source-based routing is a security feature for preventing ES-to-ES routing when they are attached to different RTUs that are attached to the same DSL card. That is, sourced-based routing can ensure that all upstream traffic within a service domain is sent to the NSP.

Without Source-Based Routing

The following illustration shows that with destination routing ES1 can send packets to ES2 based on the static route table. That is, when ES1 sends a packet to ES2, the destination route is 155.1.3.5 and the next hop address for this destination is 135.1.3.4 (RTU 2).

DSL Card*

s1c

Router

155.1.3.1

s1d

*If DSL card is an 8540 DSL card,

associated RTU will not hav e an IP address.

DSL Routing Table

8540 DSL Card Next-Hop

Host/Net

1) 155.1.3.4

2) 155.1.3.5

3) 0.0.0.0

Subnet Mask

255.255.255.255

0.0.0.0

Address s1c

s1d

155.1.3.1

RTU 1*

135.1.3.3

RTU 2*

135.1.3.4

8546 DSL Card Next-Hop Address

135.1.3.3

135.1.3.4

155.1.3.1

ES1

155.1.3.4

Packet Flow

ES2

155.1.3.5

S/D (Source/ Destination)

dst (destination) dst (destination) dst (destination)

97-15472-0

With Source-Based Routing

With source-based routing, the source address of upstream packets sent from an ES are compared to the source address listed in the static route table. If a match is found, the packet is sent to the next-hop address specified for that source address.

The following illustration shows the packet flow when ES1 sends to ES2, and when source-based routes are defined for ES1 and ES2 (indicated by the S/D flag).

6-10

November 1997

8000-A2-GB21-20

Page 62

IP Routing

Router

155.1.3.1

155.1.2.1

155.1.2.2

NSP

Partial DSL Routing Table

Host/Net

1) 155.1.3.4

2) 155.1.3.5

Subnet Mask

255.255.255.255

DSL Card*

s1c

e1a

s1d

*If DSL card is an 8540 DSL card,

associated RTU will not hav e an IP address.

Next-Hop Address

155.1.3.1

RTU 1*

135.1.3.3 Packet Flow

RTU 2*

135.1.3.4

S/D (Source/Destination) src (source)

src (source)

ES1

155.1.3.4

ES2

155.1.3.5

97-15473-0

Upstream packets from ES1 (and ES2) are sent to 155.1.3.1, where in turn the router would forward them to the NSP. Downstream packets from the NSP are sent to ES2.

For upstream packets only (i.e., packets arriving over the DSL ports), the order of routing precedence is:

1. Source host route

2. Source subnet route

3. Source network route

4. Destination host route

5. Destination subnet route

6. Destination network route

7. Default route

NOTE:

When using source routing, do not use the default route.

8000-A2-GB21-20

November 1997

6-11

Page 63

IP Routing

The following illustration shows the packet flow when ES1 sends to ES3, ES1 and ES3 are in different service domains, and source-based routes are defined for ES1 and ES2 (indicated by the S/D flag).

Network

NSP1

155.1.2.2

155.1.2.1

Router

155.1.3.1

159.1.3.1

159.1.2.1

159.1.2.2

NSP2

Partial DSL Routing Table

Host/Net

1) 155.1.3.4

2) 155.1.3.5

3) 159.1.3.4

*If DSL card is an 8540 DSL card,

Subnet Mask

255.255.255.255

DSL Card*

s1c

e1a

s1d

associated RTU will not hav e an IP address.

Next-Hop Address

155.1.3.1

159.1.3.1

RTU 1*

135.1.3.3

RTU 2*

135.1.3.4

S/D (Source/Destination) src (source)

src (source) src (source)

Packet Flow

ES1

155.1.3.4

ES2

155.1.3.5

ES3

159.1.3.4

97-15560-01

6-12

November 1997

8000-A2-GB21-20

Page 64

IP Filtering

Overview

A filter is a useful mechanism and can be used to:

 Secure a network by implementing security rules (policies)  Prevent unauthorized network access without making authorized access

difficult.

By default, filtering is not active on the HotWire DSLAM system. However, you can enable filtering to selectively filter source or destination packets being routed through the MCC or DSL cards. Appendix B,

Worksheets

configurations.

, provides worksheets to help you plan and record your filter

IP Filtering Configuration

What is a Filter?

This chapter provides an overview of packet filters and describes why you may want to set filters on your network.

An IP filter is a rule (or set of rules) that is applied to a specific interface to indicate whether a packet can be forwarded or discarded.

A filter works by successively applying its rules to the information obtained from the packet header until a match is found. (Host rules have precedence over network rules.) The filter then performs the action specified by the rule on that packet, which can be either to forward or discard. If the packet header information does not match any of the rules, then the user-specified default filter action is performed. The filter does not change any state or context, and the decision is made based only on the packet contents.

8000-A2-GB21-20

November 1997

7-1

Page 65

IP Filtering

NOTE:

If your system is set up for dynamic IP addressing and you have enabled the dynamic access control feature, you do not need to configure filters because this is done automatically. However, you will need to bind the filters to the appropriate interface. The dynamic access control feature is configurable on the DHCP Relay Servers screen. See Chapter 6 of the

8540 and 8546 DSL Cards User’s Guide

You can create the following filter types:

 An input filter to prevent packets entering the DSL card through a specified

interface from being forwarded. You may want to set up filtering on input to protect against address spoofing. Use the IP Network screen

Configuration→Interfaces→IP Network

( to a particular interface.

 An output filter to prevent packets from going out of the DSL card through a

specified interface. Use the IP Network screen (

IP Network

) to specify binding of an output filter to a particular interface.

for more information.

) to specify binding of an input filter

Configuration→Interfaces

HotWire DSLAM for

→

For each filter type, you must set up one or more of the following rule types on the IP Filter Configuration screen (

 A network address rule type to discard or forward packets/traffic from a

specified network or a segment of the network. This rule type can also be used to enhance security by allowing access only to certain networks. The IP address and subnet mask specified in the Destination address and

Destination address mask fields, or the Source address and Source address mask fields of the IP Filter Configuration screen are

compared to the destination/source address contained in the IP header of the packet.

 A host address rule type to discard or forward packets/traffic from a

specified host. This rule type can also be used to enhance security by allowing access only to certain hosts. The IP address and subnet mask specified in the Destination address and Destination address mask fields, or the Source address and Source address mask fields of the IP Filter Configuration screen are compared to the destination/source address contained in the IP header of the packet.

Configuration→IP Router→IP Router Filters

NOTE:

Host address rules have precedence over network address rules. All host address rules will be invoked sequentially before the first network address rule is invoked.

 A socket address rule type to limit certain applications. This rule type is

used primarily when filtering TCP or UDP packets, and may be used in conjunction with a network address rule type or a host address rule type. The destination (socket) port number specified in the Destination Port No. field and source (socket) port number specified in the Source Port No. field of the IP Filter Configuration screen are compared to the destination and source port numbers in the TCP or UDP header of the packet.

7-2

November 1997

8000-A2-GB21-20

Page 66

NOTE:

If both the source and destination port numbers are 0s (zeros), the system filters ICMP packets in addition to the packet types defined in the rule.

In this release, you can configure up to two filters on the MCC card and up to eight filters on each DSL card. Also, up to 33 rules can be configured for each filter. Keep in mind that for each filter, you will need to configure the default filter action (either to forward or discard packets).

For detailed information on the IP Filter Configuration screen and the IP Network screen, see Chapters 5 and 6 of the

Cards User’s Guide

Security Advantages

Filtering provides security advantages on LANs as described in the following subsections.

NOTE:

All upstream traffic from an ES is forwarded by a HotWire 5246 or 5446 RTU to the DSL card unless it is addressed to another ES (in the same subnet) on the same LAN.

IP Filtering

HotWire DSLAM for 8540 and 8546 DSL

Management Traffic Leakage

Filtering can be used to prevent unwanted traffic from leaking into the management domain. That is, filtering prevents NSP packets with management IP destinations from being accepted for local delivery or routing.

For example, if the NSP network is 155.1.00.00 and the management network is

135.1.00.00, filters can be defined that would prevent any traffic entering from the 10BaseT port from being forwarded to the 135.1.00.00 network through the DSL card.

135.1.00.00

NSP

NOTE:

Filters reduce packet throughput.

Router

155.1.00.00

10BaseT

MCC Card

DSL Card

97-15460-01

8000-A2-GB21-20

November 1997

7-3

Page 67

IP Filtering

Service Security

For instructions on how to set filters to prevent unwanted traffic from leaking into the management domain, see Chapter 5 of the

8546 DSL Cards User’s Guide

HotWire DSLAM for 8540 and

Filtering on the upstream DSL ports can be used to ensure that only end-user systems with valid IP addresses are able to route traffic to the service domain. That is, filtering would block traffic from being routed upstream by another end-user system that spoofs (attempts to gain access to another system by posing as an authorized user) an IP address of an end-user system connected to a different HotWire RTU.

The following illustration is an example of this type of filtering:

End-user

System 1

155.1.3.4

DSL Card

155.1.3.4

RTU

End-user

System 2

97-1549

For information on how to set filters on the upstream DSL ports, see Chapters 5 and 6 of the

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

7-4

November 1997

8000-A2-GB21-20

Page 68

Service Security Filtering Scenario

The following is an example of filtering to ensure service security:

IP Filtering

NSP1

155.1.2.2

*If DSL card is an 8540 DSL card,

associated RTU will not hav e an IP address.

DSL Routing Table

Host/Net Subnet Mask

1) 155.1.3.4

2) 155.1.3.4

3) 155.1.3.6

4) 155.1.3.6

Router

155.1.2.1

155.1.3.1

255.255.255.255

DSL Card*

s1c

155.1.3.2

s1d

8540 DSL Card Next-Hop Address

s1c s1d

RTU 1*

135.1.3.3

RTU 2*

135.1.3.5

8546 DSL Card Next-Hop Address

155.1.3.1

135.1.3.3

155.1.3.1

135.1.3.5

ES1

155.1.3.4

ES2

155.1.3.6

S/D (Source/ Destination)

src (source) dst (destination) src (source) dst (destination)

97-15476-0

The RTU forwards upstream any traffic on its LAN interface for which it does not know the host.

8000-A2-GB21-20

November 1997

7-5

Page 69

IP Filtering

In the following illustration, ES2 spoofs ES1’s IP address (that is, ES2 assumes ES1’s IP address of 155.1.3.4):

NSP1

155.1.2.2

*If DSL card is an 8540 DSL card,

associated RTU will not hav e an IP address.

DSL Routing Table

Host/Net Subnet Mask

1) 155.1.3.4

2) 155.1.3.4

3) 155.1.3.6

4) 155.1.3.6

Router

155.1.2.1

155.1.3.1

255.255.255.255

155.1.3.2

8540 DSL Card Next-Hop Address

155.1.3.1 s1c

155.1.3.1 s1d

DSL Card*

s1c

s1d

RTU 1*

135.1.3.3

RTU 2*

135.1.3.5

8546 DSL Card Next-Hop Address

155.1.3.1

135.1.3.3

155.1.3.1

135.1.3.5

ES1

155.1.3.4

ES2

155.1.3.4

ES2 spoofing

ES1’s address

S/D (Source/ Destination)

src (source) dst (destination) src (source) dst (destination)

97-15477-0

With no input filtering on the DSL ports, ES2 can successfully send traffic to the NSP identifying itself as ES1 (155.1.3.4).

Now, consider that the following filter rules are applied to s1d:

IP Address

155.1.3.6 255.255.255.255 Source Forward Default — — Discard

Subnet Mask Source/Destination Action

With these filter rules active on s1d, when ES2 tries to send packets to ISP1, the filter on the DSL card blocks the packets from being forwarded, because only packets with a source IP address of 155.1.3.6 are forwarded.

7-6

November 1997

8000-A2-GB21-20

Page 70

SNMP Agent

Overview

The Simple Network Management Protocol (SNMP) is an application-level protocol used in network management. A Network Management System (NMS), such as Paradyne’s DCE Manager, communicates to an SNMP agent via SNMP in order to obtain (get) specific parameters or variables within control of the SNMP agent.

When DCE Manager is configured properly, it can communicate with the HotWire DSLAM SNMP agent. Almost all communications between the DCE Manager and the HotWire DSLAM SNMP agent originate with a request message from the DCE Manager to the HotWire DSLAM. When the DSLAM receives the request, the HotWire DSLAM SNMP agent processes the request message and transmits a response (positive or negative) message back to the DCE Manager. When certain significant events occur within the SNMP agent, this can result in transmission of unprompted SNMP trap messages to the DCE Manager. (Note that the HotWire DSLAM SNMP agent is SNMP Version 1 (V1) compliant with community-based management.)

This chapter describes what you need to know to configure the SNMP agent within the HotWire DSLAM. This chapter does not, however, describe the procedures on how to configure the SNMP agent. For those procedures, see the

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

MIB Compliance

Various pieces of configuration, status, and statistical data within the HotWire DSLAM SNMP agent form a database of information that is accessible from the DCE Manager. This collection of information is called a Management Information Base (MIB). The basic definitions of the content of an SNMP agent’s MIB are defined within various Internet Request for Comments (RFC) documents.

An HP OpenView MIB browser requires the operator to load the appropriate MIB files into its database before it can manage the HotWire DSLAM network. For more information about DCE Manager, see the

for Windows User’s Guide

8000-A2-GB21-20

or the

November 1997

DCE Manager User’s Guide

DCE Manager for HP OpenView

8-1

Page 71

SNMP Agent

The HotWire DSLAM supports the following MIBs:

 MIB II — System Group (described in RFC 1213)  MIB II — ICMP Group (described in RFC 1213)  MIB II — UDP Group (described in RFC 1213)  MIB II — Transmission Group (described in RFC 1213)  MIB II — SNMP Group (described in RFC 1213)  MIB II — Definitions of Managed Objects for the Ethernet-like Interface Types

(described in RFC 1398)

 MIB II — Definitions of Managed Objects for the Link Control Protocol of the

Point-to-Point Protocol (described in RFC 1471)

 MIB II — Definitions of Managed Objects for the IP Network Control Protocol

of the Point-to-Point Protocol (described in RFC 1473)

 MIB II — Evolution of Interfaces Group (described in RFC 1573)  MIB II — Ethernet Interface MIB (described in RFC 1643)  Entity MIB (described in RFC 2037)  Paradyne DSL Enterprise MIBs:

— HotWire System MIB (hot_sys.mib) — HotWire xDSL MIB (hot_xdsl.mib) — Security MIB (devSecurity.mib) — Device Health and Status MIB (devHealthAndStatus.mib) — DHCP Relay Agent MIB (hot_dhcp.mib) — HotWire 5446 RTU Traps MIB (trapdefs.mib)

8-2

November 1997

8000-A2-GB21-20

Page 72

Supported Traps

SNMP defines six basic or standard traps. These messages are identified with a value of 0 through 5 within the generic-trap field of the trap message. (Note that the HotWire DSLAM SNMP agent does not support trap messages with a value of 5.) The specific-trap field of standard trap messages is set to 0 (zero). The specific-trap field of enterprise-specific messages defines the trap.

The HotWire DSLAM SNMP agent supports generation of the following standard trap messages (specific-trap=0):

SNMP Agent

 coldStart(0). The sending SNMP agent reinitializes itself such that the

agent’s configuration may be altered.

 warmstart(1). The sending SNMP agent is reinitialized without altering the

agent’s configuration.

 linkDown(2). A link on the sending SNMP agent is no longer operational.  linkUp(3). A link on the sending SNMP agent has become operational.  authenticationFailure(4). The sending SNMP agent has received an SNMP

message specifying a community name which it does not recognize, or requesting an action not permitted for the specified community.

There are additional HotWire Enterprise supported traps, which can be found in the Paradyne DSL Enterprise MIBs. See the MIBs for a complete list of traps. MIBs can be accessed through the Paradyne Power Pages (

The generation of SNMP trap messages can be selectively enabled per configured community. Additionally, the authenticationFailure trap can be selectively enabled for all configured communities that have traps enabled. If any communities have the generation of trap messages enabled, then the generation of authenticationFailure traps is determined by the state of the global authenticationFailure switch.

General SNMP Agent Configuration

Depending on your specific network configuration, various aspects of the HotWire DSLAM SNMP agent may need to be configured. For example, you may want to set up your system to send SNMP traps to a specific SNMP NMS manager. The HotWire DSLAM system provides four default community names (two read/write community names and two read-only community names) per MCC or DSL card. These community names are similar to passwords.

Make sure that the SNMP NMS manager that will receive SNMP trap messages knows and uses the correct community name, as specified on the HotWire DSLAM. You can change the default community names to match the name of the SNMP NMS manager. Without the correct community name, the NMS manager will not be able to communicate with the DSLAM.

www.paradyne.com

8000-A2-GB21-20

November 1997

8-3

Page 73

SNMP Agent

As a minimum configuration, you must do the following on the SNMP Communities/Traps screen in order for an NMS to receive SNMP traps:

 Assign an SNMP NMS manager to a R/W (Read-Write) or R/O (Read-only)

community by specifying the SNMP NMS manager’s IP address. Y ou can specify up to three SNMP NMS managers for each community name.

 Configure the generation of trap messages by specifying E (for Enable) on

the SNMP Communities/Traps screen.

 Enable/Disable the generation of authenticationFailure trap messages.

To enable the set capability, the NMS manager needs the correct Read/Write (R/W) community name. If security is enabled, the NMS manager’s IP address must be specified with R/W privileges on the SNMP Security screen. This applies only to MCC or DSL card SNMP security.

Additionally, you can configure logical entities. Logical entities can be used to provide various management populations with different levels of management access to the HotWire DSLAM SNMP agent. These configurations provide a system-wide view of the SNMP agent. Use the Configure SNMP Logical Entity Table screens to configure access to logical entities. On an MCC card, logical entities are all of the active DSL cards in the DSLAM. On DSL cards, the logical entities are all of the active RTUs.

For detailed information about the various SNMP Agent screens mentioned in this chapter, see Chapters 5 and 6 of the

Cards User’s Guide

. Also Appendix C,

HotWire DSLAM for 8540 and 8546 DSL

SNMP Configuration Worksheets,

in this guide provides worksheets to help you plan and record your SNMP configurations.

NOTE:

The HotWire RTUs that operate with the 8540 DSL card do not have their own SNMP agent. Therefore, limited SNMP support is provided by the 8540 DSL card in the DSLAM (limited support including remote system object ID, remote system description, and remote system services).

To configure RTU information for an 8540 DSL card, use the HotWire DSLAM user interface (RTU Config screen). On this screen, you can enter the RTU type, system name, contact, and location. For detailed information, see the

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

8-4

November 1997

8000-A2-GB21-20

Page 74

Packet Walk-Throughs

Overview

This chapter provides examples of how data packets are routed through the service and management domains.

Packet Walk-Through Using an 8540 DSL Card

Service Domain Packet Walk-Through

To examine how data packets flow through the service domain, an example of ES1 issuing a ping to NSP1 will be used. The following assumptions are made:

 A source domain IP entry exists for ES1  A static route exists between the DSL card and ES1  Filtering is disabled

The following illustration shows how data packets flow through the service domain. In this illustration ES1 is connected to the same LAN as the HotWire RTU.

8000-A2-GB21-20

November 1997

9-1

Page 75

Packet Walk-Throughs

NSP1

155.1.2.2

1 2 3 4 5 6 7 8

NSP1 issues reply to ping * The RTU can be a 5170, 5171, 5216, or 5246 RTU.

Partial DSL Routing Table

Host/Net

1) 155.1.3.4

2) 155.1.3.4

Router

155.1.2.1

155.1.3.1

Subnet Mask

255.255.255.255

8540 DSL

Card

155.1.3.2

Next-Hop Address

155.1.3.1 s1c

Unnumbered DSL Interface

s1c

When ES1 pings NSP1:

ES1 pings NSP1

RTU*

ES1

155.1.3.4

97-15474-0

S/D (Source/Destination) src (source)

dst (destination)

123

1. ES originates a packet addressed to 155.1.2.2. Because they are both on the

155.1 network, ES1 ARPs to map NSP1’s IP address into a MAC address.

2. The RTU forwards the ARP to the 8540 DSL card over its DSL interface (e.g., s1c).

3. The 8540 DSL card replies to the ARP request with its own MAC address (proxy ARP).

4. After ES1 receives the ARP reply, it sends the packet to the MAC address of the 8540 DSL card.

5. Upon receiving this packet, the RTU forwards it to the 8540 DSL card over its DSL interface.

6. When the 8540 DSL card receives this packet, the 8540 DSL card consults its routing table to determine how to route the packet. Since a source route is defined for ES1 (route #1), the 8540 DSL card forwards the packet to the router (151.1.3.1), which is the next hop.

7. The router then forwards the packet to NSP1.

NSP1 then issues a reply to the ping.

1. The NSP sends the ping reply packet addressed to 155.1.3.4.

2. By normal means, the packet arrives at the router.

3. Because the router has an interface with an address 155.1.3.1 (on 155.1.3 subnet), it ARPs for 155.1.3.4.

4. Because the 8540 DSL card has a host route (marked PA=y) for 155.1.3.4, it responds to the ARP request with its own MAC address (proxy ARP).

5. Then, the ping reply is sent directly to the 8540 DSL card.

9-2

November 1997

8000-A2-GB21-20

Page 76

6. The 8540 DSL card then consults its routing table to identify the next hop to forward the packet. Since a host route is defined for ES1 (route #2), the DSL interface is used as the next hop.

7. The 8540 DSL card then forwards the packet over the DSL port to that RTU.

8. Upon receiving the packet, the RTU forwards the packet to its 10BaseT port.

Management Domain Packet Walk-Through

For an 8540 DSL card and its associated RTUs, all management functions are performed by an agent on the DSL card.

Packet Walk-Through Using an 8546 DSL Card

Service Domain Packet Walk-Through

To examine how data packets flow through the service domain, an example of ES1 issuing a ping to NSP1 will be used. The following assumptions are made:

Packet Walk-Throughs

 A host route entry has been configured in the HotWire RTU for ES1  A source domain IP entry exists for the HotWire RTU  A static route exists between the 8546 DSL card and the HotWire RTU  Filtering is disabled

The following illustration shows how data packets flow through the service domain. In this illustration ES1 is connected to the same LAN as the HotWire RTU.

ES1 pings NSP1

NSP1

155.1.2.2

NSP1 issues reply to ping

Partial DSL Routing Table

Host/Net

1) 155.1.3.4

2) 155.1.3.4

Router

155.1.2.1

155.1.3.1

2 3

Subnet Mask

255.255.255.255

155.1.3.2

4 5

8546 DSL

Card

Next-Hop Address

155.1.3.1

135.1.3.3

Unnumbered DSL Interface

RTU

135.1.3.3

S/D (Source/Destination) src (source)

dst (destination)

155.1.3.4

1234

ES1

97-15474a

8000-A2-GB21-20

November 1997

9-3

Page 77

Packet Walk-Throughs

When ES1 pings NSP1:

1. ES originates a packet addressed to 155.1.2.2. Because they are both on the

155.1 network, ES1 ARPs to map NSP1’s IP address into a MAC address.

2. The RTU receives the broadcast ARP request from ES1.

3. The RTU replies to the ARP request with its own MAC address (proxy ARP).

4. After ES1 receives the ARP reply, it sends the packet to the MAC address of the RTU.

5. Upon receiving this packet, the RTU forwards it to the 8546 DSL card over its DSL interface.

6. When the 8546 DSL card receives this packet, the 8546 DSL card consults its routing table to determine how to route the packet. Since a source route is defined for ES1 (route #1), the 8546 DSL card forwards the packet to the router (151.1.3.1), which is the next hop.

7. The router then forwards the packet to NSP1.

NSP1 then issues a reply to the ping.

1. The NSP sends the ping reply packet addressed to 155.1.3.4.

2. By normal means, the packet arrives at the router.

3. Because the router has an interface with an address 155.1.3.1 (on 155.1.3 subnet), it ARPs for 155.1.3.4.

4. Because the 8546 DSL card has a host route (marked PA=y) for 155.1.3.4, it responds to the ARP request with its own MAC address (proxy ARP).

5. Then, the ping reply is sent directly to the 8546 DSL card.

6. The 8546 DSL card then consults its routing table to identify the next hop to forward the packet. Since a host route is defined for ES1 (route #2), the RTU

135.1.3.3 is used as the next hop.

7. The 8546 DSL card then forwards the packet over the DSL port to that RTU.

8. Upon receiving the packet, the RTU forwards the packet to its 10BaseT port because it has a host route for ES1.

Management Domain Packet Walk-Through

To examine how data packets flow through the management domain, an example of the DCE Manager workstation 1 (WS1) performing a ping to the HotWire RTU is used. The following is assumed:

 A host route to the RTU (135.1.3.4) exists on the MCC card. (This is

generated automatically .)

 A static route to WS1 (135.1.1.1) is configured on the 8546 DSL card.

9-4

November 1997

8000-A2-GB21-20

Page 78

Packet Walk-Throughs

In the following illustration, WS1 is connected to the same LAN as the NMS.

RTU issues reply to ping

12345

DCE

Manager

WS1

135.1.1.1

WS1 pings the RTU

MCC Routing Table

Host/Net Subnet Mask

1) 135.1.3.4

2) 0.0.0.0

Partial DSL Routing Table

Host/Net Subnet Mask

1) 135.1.2.0

2) 135.1.1.0

3) 135.1.3.4

Router

135.1.1.2

135.1.2.1

135.1.3.254

1 2 3 4 5 6 7 8

255.255.255.255

0.0.0.0

255.255.255.0

MCC Card

e1a:135.1.2.2 s1b:135.1.3.1

Next-Hop Address S/D (Source/Destination)

135.1.3.2

135.1.2.1

Next-Hop Address S/D (Source/Destination)

135.1.3.1

135.1.3.4

8546 DSL

Card

s1b:135.1.3.2

Unnumbered Interface

dst (destination) dst (destination)

dst (destination) dst (destination) dst (destination)

RTU

135.1.3.4

97-15479-0

When WS1 pings a HotWire RTU:

1. The packet addressed to 135.1.3.4 is routed to the router by normal means.

2. The router then does an ARP request for the RTU because the router’s IP address of 135.1.3.254 is on the same subnetwork as the RTU (with an IP address of 135.1.3.4).

Note that the router’s interface to the MCC is multihomed (i.e., it has two IP addresses (135.1.2.1 and 135.1.3.254) assigned to the one interface).

3. The MCC does an ARP reply with its own MAC address (proxy ARP).

4. The router then forwards the ping packet to the MCC card.

5. Upon receiving the ping, the MCC card consults its routing table to identify to which 8546 DSL card to forward the ping.

In this case, route #1 contains a host route for 135.1.3.4 with a next hop of DSL 135.1.3.2.

6. The ping request is then forwarded to the 8546 DSL card from the MCC

s1b

card’s

interface to the 8546 DSL card’s

s1b

interface (which is over the

DSLAM system backplane).

8000-A2-GB21-20

November 1997

9-5

Page 79

Packet Walk-Throughs

7. From the routing table, the 8546 DSL card determines that 135.1.3.4 is directly connected over s1c (one of the 8546 DSL card’s DSL ports).

8. The 8546 DSL card then forwards the ping to the RTU over s1c.

The HotWire RTU then issues a ping reply to IP address 135.1.1.1.

1. The RTU forwards the ping reply to the 8546 DSL card.

2. The 8546 DSL card consults its routing table to identify how to forward the reply. Route #2 is used because the destination address (135.1.1.1) is the

135.1.1 subnet. Therefore, the next-hop address is the MCC card’s

s1b

interface (135.1.3.1).

3. Similarly, upon receiving the packet, the MCC card consults its routing table to identify how to forward the packet. Since the destination IP address of the ping is WS1 (135.1.1.1) and this does not match any entry in the route table, the next-hop IP address (135.1.2.1) of the default route is used.

4. The MCC card then forwards the packet to its 10BaseT interface to the router.

5. The router forwards the packet toward WS1 by normal means.

9-6

November 1997

8000-A2-GB21-20

Page 80

Network Configuration Worksheets

Overview

This appendix summarizes the minimum configuration steps and provides worksheets to assist you in preparing for the configuration of your HotWire DSLAM network. Use the worksheets to record configuration settings such as IP addresses and subnet masks for the MCC card, DSL cards, and RTUs. After the worksheets are completed, you can then configure your network with the assigned settings.

These worksheets are based on the network model and theories described in this guide. They map the network theories to the HotWire user interface screens. For an explanation of the network model and theories, review the chapters in this guide. For specific information about the user interface screens and fields, see

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

the

Summarizing the Network Configuration

In summary, to configure the network:

 The management domain and service domain IP addresses and static routes

are assigned to the HotWire DSLAM system using the HotWire user interface.

 If using a HotWire 5446 RTU, the RTU’s management IP address is also

assigned from the HotWire user interface. In addition, the service domain IP addresses and host routes on the HotWire 5446 RTU are assigned by using:

— Paradyne’s IP Injection Tool, — An SNMP application, such as Paradyne’s DCE Manager, or — A MIB browser.

 The IP addresses of the end-user systems are assigned by the NSP.

8000-A2-GB21-20

November 1997

A-1

Page 81

Network Configuration Worksheets

Management Domain Configuration Worksheets

For the management domain, configure the MCC card, DSL cards, and HotWire 5446 RTUs as follows:

Perform this task . . .

1. Assign an IP address to

the MCC card.

(See page A-3.)

2. Clear NVRAM if the Who

Am I screen does not appear in Task 1.

(See page A-5.)

3. Assign an IP address to

the backplane (s1b) on

the MCC card. (See page A-6.)

4. Assign IP addresses to

the DSL cards.

(See page A-7.)

5. Create a default route to

the upstream router in the management domain.

(See page A-9.)

6. Reset the MCC card.

(See page A-11.)

On this screen . . . To access the screen . . .

Who Am I Power on the HotWire

DSLAM system. The system displays the Who

Am I screen.

(HotWire – MCC) NVRAM Clear

(HotWire – MCC) IP Network

(HotWire – MCC) Configure DSL IP Addr

(HotWire – MCC) Static Routes

(HotWire – MCC) Card Reset

From the HotWire – MCC menu, select:

Configuration→Card Status

→

NVRAM Clear

From the HotWire – MCC menu, select:

Configuration→Interfaces IP Network

From the HotWire – MCC menu, select:

Configuration→DSL Cards

→

Set IP Address

From the HotWire – MCC menu, select:

Configuration→IP Router

→

Static Routes

From the HotWire – MCC menu, select:

Configuration→Card Status

→

Card Reset

→

A-2

7. (When Using an 8546

DSL Card) Assign an IP

address within the management

subnetwork for each HotWire 5446 RTU connected to an 8546 DSL card.

(See page A-12.)

8. Configure a static route

to an NMS (on each DSL

card). (See page A-14.)

(HotWire – DSL) IP Network

(HotWire – DSL) Static Routes

From the HotWire – DSL menu, select:

Configuration→Interfaces IP Network

From the HotWire – DSL menu, select:

Configuration→IP Router

→

Static Routes

→

Use the worksheets in the following sections to record your network configuration settings. Photocopy the worksheets as needed.

8000-A2-GB21-20November 1997

Page 82

TASK 1: Assign an IP Address to the MCC Card

On the Who Am I screen, assign an IP address to the MCC card.

Network Configuration Worksheets

Access the . . .

Who Am I screen Powering on the HotWire DSLAM system.

By . . .

8000-A2-GB21-20

November 1997

A-3

Page 83

Network Configuration Worksheets

Who Am I Screen

Prompt Your Configuration Setting

1. Enter the IP address to the MCC card

(e1a) at the (nnn.nnn.nnn.nnn): prompt.

2. Enter the subnet mask at the

(nnn.nnn.nnn.nnn): prompt. Note that the system automatically

calculates the subnet mask. Press Return to accept the default value or enter a new value at the prompt.

3. Reboot the system by typing yes at the

yes/no: prompt, when the system highlights OK to restart?.

IP Address =

Subnet Mask =

NOTE:

To continue configuring the management domain, you must select the MCC card.

After the system reboots, press Return to display the HotWire Chassis menu.

— From the HotWire Chassis menu, select Card Selection.

The Card Selection screen appears.

— At the Goto Card (M for MCC or slot# for DSL): prompt,

enter M and press Return. The HotWire – MCC menu appears.

A-4

8000-A2-GB21-20November 1997

Page 84

TASK 2: Clear NVRAM

On the Clear NVRAM screen, clear the non-volatile RAM if the Who Am I screen does not appear after power up (in Task 1) by entering yes at the Initialize NVRAM: yes/no prompt.

Network Configuration Worksheets

NOTE:

An answer of yes causes the loss of all static configuration information. Any changed parameters will return to default values, including user accounts, filtering information, interface configurations, and port configurations.

Access the . . .

Clear NVRAM screen

By . . .

Selecting

Clear

Configuration→Card Status→NVRAM

from the HotWire – MCC menu.

8000-A2-GB21-20

November 1997

A-5

Page 85

Network Configuration Worksheets

TASK 3: Assign an IP Address to the Backplane (s1b)

On the IP Network screen, assign an IP address to the backplane (s1b).

NOTE:

You will need to create a separate and distinct network or subnetwork for the 8546 DSL cards and 5446 RTUs, or for 8540 DSL cards. However, the RTUs associated with the 8540 DSL cards do not need to be included in the network.

Access the . . .

IP Network screen

By . . .

Selecting from the HotWire – MCC menu.

Configuration→Interfaces→IP Network

IP Network Screen A-C-B

Prompt Your Configuration Setting

1. Enter the interface name at the Input

Interface Name: prompt.

2. Enter the base IP address at the

(nnn.nnn.nnn.nnn): prompt.

3. Enter the base subnet mask at the

(nnn.nnn.nnn.nnn): prompt.

4. Enter the peer IP address at the

(nnn.nnn.nnn.nnn)or address-pool: prompt.

5. Enter route type NET (for network) at

the Route to peer (host/net): prompt.

A-6

IP Interface = s1b

Base IP Addr =

Base Subnet Mask =

Peer IP Address =

Route to Peer= NET

8000-A2-GB21-20November 1997

Page 86

TASK 4: Assign IP Addresses to the DSL Cards

On the Configure DSL IP Addr screen, assign an IP address to each DSL card in the system.

Network Configuration Worksheets

Access the . . .

Configure DSL IP Addr screen

By . . .

Selecting

Address

Configuration→DSL Cards→Set IP

from the HotWire – MCC menu.

8000-A2-GB21-20

November 1997

A-7

Page 87

Network Configuration Worksheets

Configure DSL IP Addr Screen A-G-A

Prompt Your Configuration Setting

1. Enter the DSL card subnet mask at

the (nnn.nnn.nnn.nnn): prompt.

This is the subnet mask for the backplane (s1b) management subnet.

2. Enter the IP address for each DSL

card in the system. Select the appropriate slot number by using the arrow keys to move from one field to another.

Once the slot number is selected, enter the IP address for that DSL card at the (nnn.nnn.nnn.nnn: prompt.

DSL Card Subnet Mask =

Slot 1 IP Address = Slot 2 IP Address = Slot 3 IP Address = Slot 4 IP Address = Slot 5 IP Address = Slot 6 IP Address = Slot 7 IP Address = Slot 8 IP Address = Slot 9 IP Address = Slot 10 IP Address = Slot 11 IP Address = Slot 12 IP Address = Slot 13 IP Address = Slot 14 IP Address = Slot 15 IP Address = Slot 16 IP Address = Slot 17 IP Address = Slot 18 IP Address =

A-8

8000-A2-GB21-20November 1997

Page 88

TASK 5: Create a Default Route

On the Static Routes screen, create a default route to the management domain next hop router. This default route will be used when no other routes in the routing table apply.

Network Configuration Worksheets

Access the . . .

Static Routes screen

By . . .

Selecting

Routes

Configuration→IP Router→Static

from the HotWire – MCC menu.

8000-A2-GB21-20

Static Routes Screen A-E-A

Prompt Your Configuration Setting

1. Enter 0 or press Return at the Item

Number (0 to add new record): prompt to add a new

record.

2. Enter 0.0.0.0 at the Destination

(or space to delete route):

prompt.

3. Enter 0.0.0.0 or press Return at the

Subnet:(nnn.nnn.nnn.nnn):

prompt.

4. Enter the management domain

next-hop router’s IP address at the Next Hop IP Address (nnn.nnn.nnn.nnn): prompt.

November 1997

Host/Net = 0.0.0.0

Subnet Mask = 0.0.0.0

Next Hop =

A-9

Page 89

Network Configuration Worksheets

Static Routes Screen A-E-A

Prompt Your Configuration Setting

5. Enter 1 at the Input Number:

prompt to specify the preference for this route.

1 has the highest preference. The greater the number the lower the preference.

6. Enter dst or press Return at the

Source (Src)/ Destination(dst): prompt.

7. Enter no or press Return at the

yes/no: prompt to keep the NO value under the PA (proxy ARP) column.

8. When the system highlights Save

Changes?, enter yes at the yes/no: prompt.

Pref= 1

S/D= dst

PA= no

A-10

8000-A2-GB21-20November 1997

Page 90

TASK 6: Reset the MCC Card

After configuring the MCC card for the management domain, reset the card to install the configuration setting. On the Card Reset screen (

Status

prompt.

→

Card Reset)

Network Configuration Worksheets

Configuration→Card

, reset the MCC card by entering yes at the yes/no:

NOTE:

After resetting the MCC card, select a DSL card to continue with the management domain configuration. To select a DSL card:

— Press Return to display the top-level menu (HotWire Chassis menu). — Select Card Selection from the HotWire Chassis menu.

The Card Selection screen appears.

— Verify that the DSL card you want to configure appears on the Card

Status screen.

— At the Goto Card (M for MCC or slot# for DSL): prompt,

enter the number of the slot. Then, press Return. For example, if you want to configure the DSL card in slot 4, enter 4.

The HotWire – DSL menu appears.

8000-A2-GB21-20

November 1997

A-11

Page 91

Network Configuration Worksheets

TASK 7: (When Using an 8546 DSL Card) Configure the HotWire 5446 RTU Management Domain IP Addresses

On the IP Network screen, configure the HotWire 5446 RTU IP addresses on each 8546 DSL card, which are the RTU’s management domain IP addresses.

Access the . . .

IP Network screen

By . . .

Selecting from the HotWire – DSL menu.

Configuration→Interfaces→IP Network

A-12

IP Network Screen A-C-B

Prompt Your Configuration Setting

For DSL port 1 (s1c):

1. Enter the interface name at the Input

Interface Name: prompt.

2. Enter the port 1 RTU’s IP address at

the (nnn.nnn.nnn.nnn)or address-pool: prompt.

3. Enter route type HOST at the Route

to peer (host/net): prompt.

IP Interface = s1c

Peer IP Address =

Route to Peer= HOST

8000-A2-GB21-20November 1997

Page 92

Network Configuration Worksheets

IP Network Screen A-C-B

Prompt Your Configuration Setting

For DSL port 2 (s1d):

1. Enter the interface name at the Input

Interface Name: prompt.

2. Enter the port 2 RTU’s IP address at

the (nnn.nnn.nnn.nnn)or address-pool: prompt.

3. Enter route type HOST at the Route

to peer (host/net): prompt.

For DSL port 3 (s1e):

1. Enter the interface name at the Input

Interface Name: prompt.

2. Enter the port 3 RTU’s IP address at

the (nnn.nnn.nnn.nnn)or address-pool: prompt.

3. Enter route type HOST at the Route

to peer (host/net): prompt.

For DSL port 4 (s1f):

1. Enter the interface name at the Input

Interface Name: prompt.

2. Enter the port 4 RTU’s IP address at

the (nnn.nnn.nnn.nnn)or address-pool: prompt.

IP Interface = s1d

Peer IP Address =

Route to Peer= HOST

IP Interface = s1e

Peer IP Address =

Route to Peer= HOST

IP Interface = s1f

Peer IP Address =

3. Enter route type HOST at the Route

to peer (host/net): prompt.

Route to Peer= HOST

8000-A2-GB21-20

November 1997

A-13

Page 93

Network Configuration Worksheets

TASK 8: Create a Static Route to an NMS

On the Static Routes screen, create a static route to the NMS (on each DSL card). Use this screen to enable the management traffic from the 8540 DSL cards, or the 8546 DSL cards and their downstream 5446 RTUs to be routed back through the MCC card.

Access the . . .

Static Routes screen

By . . .

Selecting

Routes

Configuration→IP Router→Static

from the HotWire – DSL menu.

A-14

8000-A2-GB21-20November 1997

Page 94

Network Configuration Worksheets

Static Routes Screen A-E-A

Prompt Your Configuration Setting

1. Enter 0 or press Return at the Item

Number (0 to add new record): prompt to add a new

record.

2. Enter the IP address of the NMS at the

Destination (or space to delete route): prompt.

3. Do

one

Subnet:(nnn.nnn.nnn.nnn):

prompt:

of the following at the

– Enter 255.255.255.255 if you want

to create a host route to the IP address specified in Step 2, or

– Enter the appropriate subnet mask if

you want to enter a network or subnet route.

1) Host/Net =

2) Host/Net =

3) Host/Net =

4) Host/Net =

5) Host/Net =

6) Host/Net =

7) Host/Net =

8) Host/Net =

9) Host/Net =

10) Host/Net =

11) Host/Net =

12) Host/Net =

1) Subnet Mask =

2) Subnet Mask =

3) Subnet Mask =

4) Subnet Mask =

5) Subnet Mask =

6) Subnet Mask =

7) Subnet Mask =

8) Subnet Mask =

9) Subnet Mask =

10) Subnet Mask =

11) Subnet Mask =

12) Subnet Mask =

4. Enter the backplane IP address of the

MCC card (s1b) at the Next Hop IP

Address (nnn.nnn.nnn.nnn):

prompt.

5. Enter 1 at the Input Number:

prompt to specify the preference for this route.

1 has the highest preference. The greater the number the lower the preference.

Up to 12 Network Management Systems (NMSs) can be specified per DSL card.

8000-A2-GB21-20

November 1997

Next Hop =

Pref= 1

A-15

Page 95

Network Configuration Worksheets

Service Domain Configuration Worksheets

For the service domain, select the DSL card you want to configure, and then configure the following for each of the DSL cards in the HotWire DSLAM:

Perform this task . . .

1. Assign IP addresses to

the DSL card LAN

interface (e1a). (See page A-17.)

2. Reset the DSL card.

(See page A-19.)

On this screen . . . To access the screen . . .

(HotWire – DSL) IP Network

(HotWire – DSL) Card Reset

From the HotWire – DSL menu, select:

Configuration→Interfaces

→

IP Network

From the HotWire – DSL menu, select:

Configuration→Card Status

→

Perform the following tasks only if assigning addresses statically

3. Create static routes to

end-user systems on each DSL card.

(See page A-20.)

4. Create default route or a

source route.

(See page A-22.)

(HotWire – DSL) Static Routes

From the HotWire – DSL menu, select:

Configuration→IP Router

From the HotWire – DSL menu, select:

Configuration→IP Router

→

Perform the following task only if assigning addresses dynamically

5. Define DHCP relay

features to enable dynamic IP address configuration.

(See page A-24.)

(HotWire – DSL) DHCP Relay

From the HotWire – DSL menu, select:

Configuration→ DHCP Relay

Card Reset

Static Routes

A-16

Use the worksheets in the following sections to record your network configuration settings. Photocopy the worksheets as needed.

8000-A2-GB21-20November 1997

Page 96

Network Configuration Worksheets

TASK 1: Assign IP Addresses to the DSL Card LAN Interface (e1a)

On the IP Network screen, assign IP addresses to the DSL card LAN interface (e1a). Up to 16 ISP domains can be supported per DSL card.

Access the . . .

IP Network screen

By . . .

Selecting from the HotWire – DSL menu.

Configuration→Interfaces→IP Network

8000-A2-GB21-20

November 1997

A-17

Page 97

Network Configuration Worksheets

IP Network Screen A-C-B

Prompt Your Configuration Setting

1. Enter the interface name at the Input

Interface Name: prompt.

2. Enter the IP address at the

(nnn.nnn.nnn.nnn): prompt. This address must be different than the

management domain IP address.

3. Enter the subnet mask at the

(nnn.nnn.nnn.nnn): prompt.

IP Interface = e1a

1) IP Addr =

2) IP Addr =

3) IP Addr =

4) IP Addr =

5) IP Addr =

6) IP Addr =

7) IP Addr =

8) IP Addr =

9) IP Addr =

10) IP Addr =

11) IP Addr =

12) IP Addr =

13) IP Addr =

14) IP Addr =

15) IP Addr =

16) IP Addr =

1) Subnet Mask =

2) Subnet Mask =

3) Subnet Mask =

4) Subnet Mask =

5) Subnet Mask =

6) Subnet Mask =

7) Subnet Mask =

8) Subnet Mask =

9) Subnet Mask =

10) Subnet Mask =

11) Subnet Mask =

12) Subnet Mask =

13) Subnet Mask =

14) Subnet Mask =

15) Subnet Mask =

16) Subnet Mask =

Up to 16 IP addresses and subnet masks can be entered. Enter the IP addresses and subnet masks for each ISP domain supported by the specified DSL card.

A-18

8000-A2-GB21-20November 1997

Page 98

TASK 2: Reset the DSL Card

After configuring the e1a interface, reset the card. On the Card Reset screen

Configuration→Card Status→Card Reset)

( at the yes/no: prompt.

Network Configuration Worksheets

, reset the DSL card by entering yes

8000-A2-GB21-20

November 1997

A-19

Page 99

Network Configuration Worksheets

TASK 3: Create Static Routes to End-User Systems

NOTE:

Perform this task only if you are assigning IP addresses to the end-user systems statically. If you are planning to assign addresses dynamically, skip

Task 5: Define DHCP Relay Features to Enable Dynamic IP Address

Configuration

On the Static Routes screen, create a static route to end-user systems on each DSL card. For host addressing, fill out one worksheet for each end-user system. For structured subnet addressing, complete up to 16 worksheets (up to four worksheets for each of the DSL ports corresponding to the four domains supported on each port).

NOTE:

Each time you create a static route for an end-user system behind an RTU, you should also create a corresponding source-based input filter rule. See Chapter 7,

Worksheets

IP Filtering

, for introductory information about the filtering screens. See the

HotWire DSLAM for 8540 and 8546 DSL Cards User’s Guide

information.

, and Appendix B,

IP Filtering Configuration

for detailed

Access the . . .

Static Routes screen

By . . .

Selecting

Routes

Configuration→IP Router→Static

from the HotWire – DSL menu.

A-20

8000-A2-GB21-20November 1997

Page 100

Network Configuration Worksheets

Static Routes Screen A-E-A

Prompt Your Configuration Setting

1. Enter 0 or press Return at the Item

Number (0 to add new record): prompt to add a new

record.

2. Enter the IP address of the end-user

system at the Destination (or

space to delete route):

prompt.

3. Do

one

Subnet:(nnn.nnn.nnn.nnn):

prompt:

4. Do one of the following at the Next

Hop IP Address (nnn.nnn.nnn.nnn): prompt:

5. Enter dst or src at the Source

(Src)/ Destination(dst):

prompt.

of the following at the

– Enter 255.255.255.255 if you want

to create a host route to the IP address specified in Step 2, or

– Enter the appropriate subnet mask if

you want to enter a network or subnet route. (For an 8540 DSL card, the subnet mask must be

255.255.255.255.)

– If using an 8540 DSL card, enter the

DSL port number of the associated RTU.

– If using an 8546 DSL card, enter the

IP address of the associated 5446 RTU management IP address.

Host/Net =

Subnet Mask =

Next Hop =

S/D=

6. Enter yes at the yes/no: prompt. PA= yes

7. When the system highlights Save

Changes?, enter yes at the yes/no: prompt.

8000-A2-GB21-20

November 1997

A-21