How to Setup a LAN

Check how to set a lan between two or more computers, full tricks to lan set up information, Configuration the LAN was not so much easy before that so keep continue reading to set a lan

SETTING UP YOUR OWN LAN

LAN Overview

This chapter shows how to set up a local network in your home or office (if you already have a functioning network, feel free to skip to the next chapter). In the next few pages, we’ll take a look at:


· The basics of network hardware

· Basic hardware requirements for your local network

· Installing the hardware and setting up a local peer-to-peer network

· Inspecting and changing network settings

A. NETWORK HARDWARE BASICS

Cables. We recommend Category 5 cables for new users. Officially called Ethernet 10/100BaseT,
they’re the most common type of network cable and provide a good upgrade path should you need it. Cat 5 allows either 10- or 100-megabyte communication. These terms have simple meanings, so don’t let them put you off:

· The “10” or “100” in 10/100BaseT refers to network connection speed—i.e., 10 Megabits or 100 Megabits per second. Most networks actually top out at less, though most users would never know.

· The “T” in BaseT refers to the wire type, twisted-pair, which consists of pairs of thin wires twisted around each other. It also refers to the connector, commonly called an RJ-45, which resembles a bigger and wider telephone connector.

· “Base” means that the cable is used for baseband (i.e., simple, single frequency) rather than broadband (multiplex or analog) networks. Cables can be purchased in different lengths and often different colors. They come with a male RJ-45 plug at each end. Cards and hubs have female RJ-45 jacks. Network Cards. A wide variety of network cards—officially called Network Interface Cards and nicknamed NICs—is available. Most do at least an adequate job. If you’re a novice networker, the primary things to look for are:

· Connection Jack. Be sure the NIC’s jack matches the type of cable you’re using. If you’re using 10BaseT cable, for instance, the NIC you buy should have an RJ-45 compatible connector.

· Plug and Play compatibility. This feature allows Windows 95/98 to automatically configure the card, saving you a lot of time in the process.

· Interrupt Addresses. Interrupts on any machine are at a premium, so you’ll want to determine
which ones the NIC has available. Generally, the more you pay, the more latitude you’ll have. ISA-bus cards are usually fast enough for a 10BaseT network; if you’re running 100BaseT you’ll
probably want to go with PCI-bus card for speed. If you’ve only got one interrupt left and must add two cards, use two PCI-bus network cards; part of the PCI spec is that cards can share Interrupts.

Hubs.

Ethernet is a standardized way of connecting computers together to create a network. A hub is an ethernet device used in conjunction with 10BaseT and 100BaseT cables. The cables run from the network’s computers to ports on the hub. Using a hub makes it easier to move or add computers, find and fix cable problems, and remove computers temporarily from the network (if they need to be upgraded, for instance). Hubs are available in most computer stores. It’s probably a good idea to buy one with more ports than you need, just in case your network expands. Look for:

· A connection jack compatible with your cabling.

· A cascading jack which allows you to add an additional hub later, if necessary, without replacing
the entire unit.

· Lights on the front. These can be useful when you’re trying to diagnose network connection problems.

B. HARDWARE FOR YOUR SPECIFIC LAN REQUIREMENTS

The kind of hardware you use depends on the kind of access and/or modem you’re using. If you’re using dial-up access you’ll need:

· One network card for each computer.

· One hub.

· A cable for each connection to the hub.

If you’re using cable modem, DSL modem or direct access you’ll need:

· One network card for each computer.

· One additional network card to connect to the modem (your WinProxy machine receives two cards, one for the modem and one for the local network).

· One hub.

· A cable for each hub connection.

· An additional cable for the connection from the computer to the modem. If the modem is the type that connects directly to the hub, make this last cable a cross-over cable instead and you’ll still be able to connect directly to the network card as shown. Before you rush out and buy a ransom’s worth of network hardware, however, take a few moments to draw a topography—a diagram which shows the relation between the network’s various components. Doing so lessens the chance that you’ll buy unnecessary cables or forget to buy a hub. Let’s look at a very simple topography. Assuming that you already have Internet access through an ISP, you’re probably connected to the Internet in this manner:

Now let’s look at the topography for a simple LAN. The network shown here—the number of client machines can be far greater, of course—is the standard configuration for most setups, including dialup access and cable-modem access:

As you can see, only one computer—the WinProxy computer—has a modem. The other computers are connected to each other and to the WinProxy computer by a device called a hub (more on this later). The computer using the modem and receiving the WinProxy installation must be a Windows95/98 or Windows NT machine. Other computers on the local network can be any kind—including Macs, Unix boxes, and WfWG3.11—as long as they’re capable of “speaking” TCP/IP. Once you’ve drawn your network topography, including all components, make a list of everything you need.

Notes 1. Many cable modem providers insist on installing the cable modem card themselves, and may insist upon using their own card. Before purchasing your own cables and cards, check to see what the provider’s policy is. 2. If you have only two computers, it’s possible to save the expense of a hub by connecting them back-to-back. To do so, run a cross-over cable directly from one network card to the network card on the other machine. IP addressing will still be done as described here

C. HARDWARE INSTALLATION/SETTING UP THE PEER-TOPEER

NETWORK

The best way to install an NIC is to simply follow the manufacturer’s directions. Win95/98 usually finds a new card when it starts up and then configures it for you. If it doesn’t, consult the directions that came with the card. Run a cable between each card and the hub (except for the external network card if you have a cable modem setup). Although you can probably get away with plugging/unplugging a cable from a card while the computer is running, it’s safer to do it when the computer is turned off. You can usually plug or unplug from the hub at any time.

You’ll need at least one protocol assigned to each card once it’s installed. Choose NetBEUI (NetBios Extended User Interface) at a minimum; you can have others as well. There isn’t any problem with having multiple protocols on your local network. You’ll need the TCP/IP protocol later in order to run WinProxy, but it’s not needed now when setting up a basic peer-to-peer network. Set up your basic network first, get it working, and we’ll add TCP/IP later on. During the card setup, you’ll be prompted for certain settings. If not already installed, be sure to add for each machine:

· Client for Microsoft Networks

· File and Printer Sharing

You can make changes to your settings at any time in the future. You must reboot the computer
for the changes to take effect.

D. INSPECTING AND CHANGING NETWORK SETTINGS

At this point, let’s double-check the computer network setup at Control Panel/Networks. In the window under the Configuration Tab, you’ll see a list of adapters and protocols. A typical setup is represented by a couple of small computer-shaped icons, one captioned Client for Microsoft Networks, and the other File and Printer Sharing. You’ll also see small green icons, similar in shape to a network card—one for each network card, and one for the Dial-Up Adapter (the Dial-Up Adapter counts as a network connection, with its own set of addresses and protocols). Finally, you’ll see a series of wire-and-node icons, each listing a different protocol-and-adapter combination, written in a form something like NetBEUI Æ NE2000 Compatible Card.

If you haven’t already added Client for Microsoft Networks, do so now:

· Highlight an adapter.

· Click through the path Add/Client/Microsoft/Client for Microsoft Networks.

To add a protocol capability to a network card:

· Highlight the network card.

· Click through the path Add/Protocol/Microsoft/Your Protocol. Click on the Identification Tab, where you’ll see three entry boxes titled:

Computer name: A name assigned by you to a computer (each computer on the network should have a unique name). Avoid punctuation marks. These names are frequently used in network configurations, and you’ll save confusion later by assigning distinctive names now. Old486 is a good name if you only have one 486 computer, but if you have several, assign them names like PapaBear, MamaBear, etc. NetBEUI uses this name to find things so it can perform its networking magic. You’ll sometimes see this computer name referred to as “the NetBios name.”

Workgroup Name: A group name you can assign to all the computers on your network (or you can use the default).

Computer Description: A caption that gives users on your local network information about an individual computer. An example: Maria’s Computer

Security Alert The designated protocol will usually be assigned exactly as you’ve requested. (In Windows 95 and 98, however, Microsoft assigns the NetBEUI protocol to all network adapters when you assign it to any single network adapter). If you don’t want that protocol in the other locations, highlight each one you don’t want and click Remove.

A Final Word on Your LAN

Congratulations! You now possess a working local network. You can see the other computers, move files between them, and print documents. To prepare for WinProxy and the Internet, you’ll need to add the TCP/IP protocol to each of the computers on your local network. You’ll learn how to do so in the next chapter. Once that’s done, it’s on to WinProxy!

ADDING TCP/IP TO YOUR NETWORK

TCP/IP Overview

The easiest way to install and configure WinProxy is to first add TCP/IP—the language spoken by WinProxy and the Internet—to your local network. This chapter covers the following topics:

1. Protocols and Addressing

2. Double-Checking Your Installed Network

3. Installing TCP protocol on your computers

4. Assigning IP addresses

5. Testing TCP/IP connectivity

A. FIRST THINGS FIRST: PROTOCOLS AND ADDRESSING Protocols. In networking terms the word “protocol” refers to the accepted standards or rules for the way data is transferred between computers and over the Internet. When everybody uses the same rules, it all works. There are many protocols in use. The three commonly used by local networks are NetBEUI, IPX/SPX, and TCP/IP.

NetBEUI is an acronym which stands for NetBios Extended User Interface. NetBEUI is a networking standard well suited for small networks and is easy to set up. It is also non-routable; since it uses computer names to find its way around, it can’t find distant computers.

IPX/SPX is Novell network’s version of IP addressing, used on Novell NetWare networks for both small and large systems. It works on Novell networks, but not between different types of networks (as TCP/IP will). TCP/IP, the language of the Internet, can be used on any size network. Data is sent over the network in chunks called packets. TCP (Transmission Control Protocol) is the protocol for packets of data sent over the wires. IP (Internet Protocol) is the addressing method used to get these packets to and from the right place. It is a routable protocol, designed to find distant computers. Some carefully-defined address groups are designated as intentionally non-routable; we’ll be using one of these to set up TCP/IP on your local network in the next chapter.

Network Addresses. These addresses may be assigned manually by the user, or automatically by another computer. They’re called static (i.e., fixed) assignments when assigned by the user, because they stay the same over time. When assigned automatically by computer, they’re known as dynamic (i.e., changing) assignments. If you connect via a dial-up connection, you’ll probably have a dynamic IP assignment to your Dial-Up Adapter. Your ISP assigns a different IP address to your Dial-Up Adapter each time you connect. If connecting with a cable modem, you’ll most likely have to make a static IP assignment for your Internet connection. Once this assignment is made, the IP address will not change. In addition, you’ll also have the choice of static or dynamic addresses on most of your networked computers. You can either set static IP addressing information yourself or have WinProxy make dynamic IP assignments for you. Addresses are not assigned to the computer itself, though people often speak that way as a convenient shorthand. The addresses are actually assigned to each network connection. The computer on which WinProxy will be installed, for instance, will have two network connections: an internal connection to the rest of your computers, and an external connection to your Internet Service Provider, or ISP.

In “Internet speak,” any machine with a network address is called a Host. For most simple TCP/IP systems, each host is a computer, and each computer is a host. The IP address is a 32-bit address, subdivided into four fields. Although it’s a binary number, it’s usually written in decimal form—222.5.83.47, for example. Each field can have a value from 0 through 255. However, since the end values are used for special purposes, the actual range available is from 1 to 254. What this boils down to for you, the user, is this: when entering an IP address, use only numbers between 1 and 254 in that last field.

Associated with the IP address is the subnet mask. This mask tells the computer which part of the address is unique to that machine, and which part is the general network address. Subnet masks allow you to accomplish many esoteric addressing capabilities; however, for most simple networks the subnet mask of 255.255.255.0 is the best and easiest choice. When you use this mask, the numbers in the final field of the IP address are unique to each computer, and the preceding three fields define the network address. To learn more about the intricacies of subnet masks. Some specific IP address ranges are reserved for special uses. We’ll discuss these later when setting up IP addressing on your local network. Network addresses reserved for testing or for local networks are 10.x.x.x, 90.x.x.x, 172.16-31.x.x and 192.168.x.x. These addresses all share a crucial distinction: routing computers on the Internet will not route these numbers. Since they are perfectly good numbers on a local network, but cannot be routed across the Internet, using them adds security to your local network.

Parts of a TCP packet are fields that specify the source and destination ports. These are 16-bit fields, and can thus specify more than 65 thousand ports. You’ll see many references to ports when interfacing your local net to the Internet. Ports 1 through 1024 are set aside for specific uses. Each Internet protocol has a standard port assigned to its use (e.g., Port 25 to send mail, Port 119 for news groups). In many cases, things can be easier to follow if you consider a port designation to be part of the address; some software even allows you to specify an IP address and port combination in the same statement.

B. DOUBLE-CHECKING YOUR INSTALLED NETWORK

Before going further, let’s double-check to be sure you have a basic network installed. At this point, your network should look like this:

· Your computers are connected via a working Ethernet network.

· One of the computers has an Internet connection, and is using Windows 95/98 or NT. That computer gets the WinProxy installation and will be known as the WinProxy computer.

· You already have some network protocols installed, including NetBEUI, and your computers already have NetBios names. The NetBios name of each computer can be found at Control Panel/Network/Identification/Computer Name. If your network doesn’t match these specifications, please bring it into line, using Chapter 3 to guide you, before attempting to install TCP Protocol and WinProxy. On the other hand, if you do have a basic network, read on!

C. INSTALLING TCP/IP PROTOCOL ON YOUR COMPUTERS

All communication between the client applications and WinProxy, and between WinProxy and the Internet, use TCP/IP protocols. Thus, the first thing you must do is add the TCP protocol and IP addresses to the network’s computers. As you proceed, pay attention to the dictates of the following three connection types:

1. The external WinProxy connection to the Internet. The type of IP address used—dynamic (commonly used for standard modems) or static (commonly used for cable modems)—is dictated by the ISP to which you connect and the type of service it provides.

2. The internal WinProxy connection. This connection must be a static IP assignment, and it must be assigned by you. Two reasons exist for a static assignment. First, some client applications must have a single, known address for the proxy server; second, the static assignment is used by WinProxy as a starting place for its DHCP assignments when providing tcp/ip assignments to your other computers.

3. The client computer network connections. These connections can be either dynamic or static. If they’re dynamic, WinProxy automatically makes all IP assignments and settings—the preferred method when using the WinProxy 3.0 Install Wizard. If they’re static, you must enter IP settings for each client computer. We recommend dynamic assignments for new users. Several protocols can co-exist on a local network, and you’ll usually need to have more than one. One protocol is sufficient on the connection to the Internet, and for security reasons you should have only TCP/IP. Let’s proceed. To install TCP.

1. On the machine receiving the WinProxy installation, click Control Panel/Network/Configuration. You’ll see a list of installed new components, and there should be listings for a Dial-Up adapter and a LAN adapter (exact wording varies). Look under LAN adapter to see if you have TCP installed—if it is, the listing will read something like TCP/IP ® LAN Adapter. Again, the exact wording varies.

2. If TCP/IP isn’t listed, click through this path: LAN Adapter/Add/Protocol/ Microsoft/TCPIP/
OK. That’s it! You’ll be prompted to restart, finishing the installation. Do so if you like, or you can wait until completing the next step before restarting.

3. Return to the initial screen. Look under Dial-Up Adapter to see if you have TCP/IP installed. If not, click through this path: DialUp Adapter/Add/ Protocol/Microsoft/TCP-IP/OK. When prompted to restart the computer, do so.

4. Add the TCP/IP protocol to each client machine (unless it’s already installed). The process is the same: in Control Panel/Networks look for a TCP/IP ® LAN Adapter line, adding the TCP/IP
protocol to the LAN adapter if it isn’t already installed.

For Client Machines Only After completing Step 4, take a quick look at any dial-up adapters. If any are installed and have the TCP protocol assigned, look under Properties to ensure that the dial-up adapter does not have the option Assign a specific IP address selected. It should be set to Obtain an IP address automatically. This will save you trouble down the road.

D. ASSIGNING IP ADDRESSES TO ALL NETWORK COMPUTERS

Each computer must be assigned a unique IP address. Strictly speaking, an IP address is assigned to each network connection, but it’s convenient to speak of a “machine address.” If you set your client computers to Obtain an IP address automatically (see the boxed note immediately above), WinProxy takes care of all of these settings for you. We recommend using the 90.0.0.x series of addresses on your local network. You’ll reap three major benefits by doing so:

· Your setup will match the numbers used for diagrams and instructions on the WinProxy website.

· You’ll find it much easier to follow explanations and trouble-shoot your network problems should the need arise.

· You’ll add to the security of your local computers by using this non-routing series on your local
network.

Now let’s proceed to assigning IP addresses.

1. First, let’s assign an IP address to the WinProxy machine. To do so, follow this path:

Control Panel/Network Configuration/TCP/IP/LANAdapter/Properties. Bring the IP Address Tab to the front. Click Specify an IP Address and enter an IP address and subnet mask. We recommend 90.0.0.1 and 255.255.255.0, as shown in the screen. You shouldn’t need to make any changes on other tabs for this basic installation.

2. Use the method shown above to install an IP address on each client machine. It’s easiest to use a sequential series such as 90.0.0.2, 90.0.0.3 and so on. Each computer gets a subnet mask of 255.255.255.0. Each IP address on your local network must be unique, and you can only vary the number in the final group—in other words, don’t change the 90.0.0 portion of the address.

3. If you’ll be using a dial-up connection to an Internet provider, the dial-up adapter does not get a specific IP assignment. Set it to Obtain an IP Address Automatically. The IP address for this connection will be dynamically assigned by the ISP each time you connect. These addresses come from a pool, and will probably (but not necessarily) be different each time you connect.

4. If, instead, you’ll be using a direct connection to your Internet provider (as many cable modems do), the network card connected to the modem should be assigned the IP address and subnet mask specified by your ISP for your individual Internet connection. Remember: you must have two network cards on this machine—one for the direct external connection to your provider and one for the internal connection to the rest of the computers on your local network.
The network card connecting to the rest of your local network retains the IP assignment it received in Step 1, above. At the conclusion of your installation, click through to WinProxy/Advanced Properties/General/ Multiple IP. While there, check to see that the IP number assigned to your Internet connection is defined as an external connection, and the IP number assigned to your local network is defined as an internal connection.

E. TESTING TCP/IP CONNECTIVITY

Now that you’ve added TCP/IP to all your computers, let’s run a test to determine if Network Neighborhood is up and running properly. If it is, you’ll know that the hub and cables are working correctly. We’ll use Ping for our test. It’s a simple tool included in Windows 95/98 and NT that allows easy checking of TCP/IP connectivity. First, open a DOS box (Start/Program/MS-DOS Prompt in Windows 95/98, and Start/Program/Command Prompt in Windows NT) and type the word ping. You’ll see a list of. commands and command syntax. If you’re on, say, client machine 90.0.0.2, you can check your connectivity to the WinProxy machine by typing in its IP address (90.0.0.1) after you type the word ping. If TCP/IP is properly set up on both machines you’ll get several lines that say Reply from 90.0.0.1…, as shown in the screen below. If you get no reply, something is wrong with the protocol installation of the IP address on one (or both) machines.

This series of three tests, run on each machine with a communications problem, will probably help isolate the problem:

1. Ping 127.0.0.1 to ensure that your tcp/ip software is working.

2. Ping yourself to ensure that the card is working.

3. Test to see that you can communicate with another machine.

· To run the first test (pinging the loopback address), type ping 127.0.0.1 at the DOS prompt. This verifies that the software TCP/IP stack on that machine is working and that the TCP protocol has been assigned (bound) to the card. The loopback address is specifically designated for such tests and doesn’t generate any actual network traffic. A failure at this point would implicate the software. If that’s the case, consider re-installing Winsock from your Windows CD-ROM, or download and install the latest Winsock from Microsoft.

· Now ping the IP address of the WinProxy computer, verifying that the card is working and IP addressing is correctly configured on that machine. If you discover a problem at this point, check to see that your network card is working properly. In Windows 95/98, go to Control Panel\System\Device Manager to see if there is a yellow exclamation point or question mark on your network card. If there is, click Drivers, and then choose View Resources to determine if Windows reports a conflict—e.g., an interrupt conflict. If so, you may be able to resolve the conflict by assigning an unused interrupt. If not, try reinstalling the card.

· Ping the IP address of another machine on your network. To work properly, the configuration must be correct on both machines. A problem at this stage usually indicates an IP addressing error. You’ve probably violated one of the basic IP rules, perhaps assigning the same number to two different machines, assigning a number outside the allowed range, or simply mis-typing an address. Check and double-check the assigned addresses. If you get a response such as request timed out, it means that ping did not reach (or return from) the other machine. Look for misconfigured IP addresses or unplugged hubs. If your response is something like destination unreachable, then ping didn’t know how to follow through on your request. You might get this response if, for example, you pinged an address with a different set of network fields. Look for misnamed nets or misconfigured subnet masks.

USER’S CHECKPOINT: If everything works except the last test (pinging another computer) an old proxy installation may be interfering. Proxy software that requires installation of software components on client machines as well as on the proxy server can cause tcp/ip communication problems. This software must be removed from each machine for proper tcp/ip communication.

If there seems to be a problem with a network card, go to Control Panel/ System/Device Manager/View Devices by Type. Look under Network Adapters. If you see a yellow exclamation point or question mark over the adapter, the system is having a problem with that adapter. Use the Win95/98 wizards to help track down problems. If you upgraded from Windows 95 to Windows 98, your network card drivers are probably out of date. Download new drivers made specifically for Windows 98 from the manufacturer’s web site.

Dial UP Technology

Dial-up Technology

Dialup is simply the application of the Public Switched Telephone Network (PSTN) to carry data on behalf of the end user. It involves a customer premises equipment (CPE) device sending the telephone switch a phone number to direct a connection to. The AS3600, AS5200, AS5300, and AS5800 are all examples of routers that have the capability to run a PRI along with banks of digital modems. The AS2511, on the other hand, is an example of a router that communicates with external modems. Since the time of Internetworking Technologies Handbook, 2nd edition, the carrier market has continued to grow, and there have been demands for higher modem densities. The answer to this need was a higher degree of interoperation with the telco equipment and the refinement of the digital modem: a modem capable of direct digital access to the PSTN. This has allowed the development of faster CPE modems that take advantage of the clarity of signal that the digital modems enjoy. The fact that the digital modems connecting into the PSTN through a PRI or a BRI can transmit data at more than 53 K using the V.90 communication standard attests to the success of the idea.

A Short Dialup Technology Background

Dialup technology traces its origins back to the days of the telegraph. Simple signals being sent across an extended circuit were created manually by tapping contacts together to turn the circuit either on or off. In an effort to improve the service, Alexander Graham Bell invented the telephone in 1875 and changed communication forever. Having the capability to send a voice across the line made the technology more accessible and attractive to consumers. By 1915, the Bell system stretched from New York to San Francisco. Demand for the service drove technological innovations, which led to the first transatlantic phone service in 1927 via radio signal. Other innovations along the way included. microwave stations that started connecting American cities in 1948, integrated digital networks to improve the quality of service, and communication satellites, which went into service in 1962 with the launch of Telstar 1. By 1970, more than 90 percent of American homes had telephone service. In 1979 the modulator-demodulator (modem) was introduced, and dialup networking was born. The early modems were slower and subject to proprietary communication schemes. Early uses of modems were for intermittent point-to-point WAN connections. Often, the call would come into a regular phone at a data center. An operator would hear modem tones and place the handset onto a special cradle that was the modem. In the late 1980s, the ITU-T began setting up V-series recommendations to standardize communications between both data communications equipment (DCE) and data terminal equipment (DTE). Early standards included these:

• V.8—Standardized the method that modems use to initially determine the V-series modulation at which they will communicate. Note that this standard applies only to the communication session between the two DCE devices. This was later updated with V.8bis, which also specified some of the communication standards between the DTE devices going over the DCE’s connection.
• V.21, V.23, V.27ter, V.29—Defined 300, 600/1200, 2400/4800, and 9600 baud communications, respectively.
• V.25, V.25bis, V.25ter—Served as a series of standards for automated dialing, answering, and control. Modems increased greatly in sophistication in the late 1980s. This was due in part to the breakup of the Bell system in 1984. With the client premises equipment in the hands of free enterprise, competition spurred on the development of speedier connections. More recent standards include these:
• V.32bis, V.34, V.90—Standardized 14400, 33600, and up to 56000 baud communication speeds.
• V.110—Allowed an asynchronous DTE device to use an ISDN DCE (terminal adapter). The first access servers were the AS2509 and the AS2511. The AS2509 could support 8 incoming connections using external modems, while the AS2511 could support 16. The AS5200 was introduced with 2 PRIs and could support 48 users using digital modems—this represented a major leap forward in technology. Modem densities have increased steadily, with the AS5300 supporting four and then eight PRIs. The AS5800 was later introduced to fill the needs of carrier class installations needing to handle dozens of incoming T1s and hundreds of user connections A couple of outdated technologies bear mentioning in a historical discussion of dialer technology. 56
Kflex is an older (pre-V.90) 56 K modem standard that was proposed by Rockwell. Cisco supports version 1.1 of the 56 Kflex standard on its internal modems, but it recommends migrating the CPE modems to V.90 as soon as possible. Another outdated technology is the AS5100. The AS5100 was a joint venture between Cisco and a modem manufacturer. The AS5100 was created as a way to increase modem density through the use of quad modem cards. It involved a group of AS2511s built as cards that were inserted into a backplane shared by quad modem cards, and a dual T1 card. Today dialup is still used as an economical alternative (depending on the connection requirements) to dedicated connectivity. It has important uses as backup connectivity, in case the primary lines go down. Dialup also offers the flexibility to create dynamic connections as needed.

Dialup Connectivity Technology

This section provides information from various dialup options. Also included are advanced options for dialup connectivity and various dialup methods.

Plain Old Telephone Service

The regular phone lines used in voice calls are referred to as Plain old telephone service (POTS). They are ubiquitous, familiar, and easy to obtain; local calls are normally free of charge. This is the kind of service that the phone network was built on. Sounds carried over this service are sampled at a rate of 8000 times per second (using 8 bits per sample) in their conversion to digital signals so that sound can be carried on a 64 kbps channel at acceptable levels.The encoding and decoding of voice is done by a piece of telco gear called a CODEC. The CODEC was needed to allow backward-compatibility with the old analog phones that were already in widespread use when the digital network was introduced. Thus, most phones found in the home are simple analog devices. Dialup connectivity across POTS lines has historically been limited to about 33,600 bps via modem—often referred to as V.34 speeds. Recent improvements have increased the speed at which data can be sent from a digital source to a modem on a POTS line, but using POTS lines on both ends of the connection still results in V.34 connectivity in both directions.

Basic Rate Interface

Intended for home use, this application of ISDN uses the same copper as a POTS line, but it offers direct digital connectivity to the telephone network. A special piece of equipment known as a terminal adapter is required (although, depending on the country, it may be integrated into the router or DCE device). Always make sure to check—the plug used to connect to the wall socket looks the same whether it’s the S/T or U demarcation point. Normally, a Basic rate interface (BRI) interface has two B (bearer) channels to carry data, and one D (delta) channel to carry control and signaling information. Local telephone carriers may have different plans to suit local needs. Each B channel is a 64 K line. The individual 64 K channels of the telephone network are commonly referred to as digital service 0 (DS0). This is a common denominator regardless of the types of services offered, as will be shown later in this chapter. The BRI interface is a dedicated connection to the switch and will remain up even if no calls are placed. The T1/E1 line is designed for use in businesses. T1 boasts 24 TDM channels run across a cable with 2 copper pairs. E1 offers 32 channels, although 1 is dedicated to frame synchronization. As is the case with the BRI, the T1/E1 connection goes directly into the telco switch. The connection is dedicated, so like a BRI, the T1/E1 remains connected and communicating to the switch all the time—even if there are no active calls. Each of the channels in the T1/E1 is just a B channel, which is to say that it’s a 64-K DS0. The T1/E1 is also referred to as digital service 1 (DS1). The North American T1 uses frames to define the timing between individual channels. For T1s, each frame has 24 9-bit channels (8 bits of data, 1 bit for framing). That adds up to 193 bits per frame. So, at
8000 of those per second, the T1 is carrying 1.544 Mbps between the switch and the customer premises equipment (CPE). The E1 similarly uses frames for timing, but the E1 uses 32 8-bit channels for a 256-bit frame. Again at the 8000 Hz rate, the channel yields 2.048 Mbps of traffic between the switch and the CPE. Most of the world uses the E1. Depending on the region, various line code and framing schemes will have to be used for the CPE and the switch to understand each other. For example, in North America, the encoding scheme most often seen is called binary 8 zero substitution (B8ZS), and the most common framing done is extended super frame (ESF). The telco through which the T1/E1 service is purchased must indicate which line code and framing should be used. For dialup purposes, there are two types of T1/E1: Primary Rate Interface (PRI) and channel associated signaling (CAS). PRI and CAS T1/E1s are normally seen in central locations that receive calls from remote sites or customers.

Primary Rate Interface

T1 Primary rate interface (PRI) service offers 23 B channels at 64 kbps at the cost of one D-channel (the 24th channel) for call signaling. Using NFAS to allow multiple PRIs to use a single D channel can minimize this disadvantage. E1 PRI service allows 30 channels, but it uses the 16th channel for ISDN signaling. The PRI service is an ISDN connection. It allows either voice-grade (modem) or true ISDN calls to be made and received through the T1/E1. This is the type of service most often seen in access servers because it fosters higher connection speeds.

Channel Associated Signaling

T1 Channel associated signaling (CAS) lines have 24 56K channels—part of each channel is borrowed for call signaling. This type of service is also called robbed-bit signaling. The E1 CAS still uses only the 16th channel for call signaling, but it uses the R2 international standard for analog call signals. CAS is not an ISDN interface; it allows only analog calls to come into the access server. This is often done to allow an access server to work with a channel bank, and this scenario is seen more commonly in South America, Europe, and Asia,sends a call into a channel that isn’t expecting it, the switch will get back a message indicating that the channel isn’t available. An access server must maintain state information on its lines and be prepared to coordinate inward and outward calls with the switch.

Modems

From a terminology standpoint, a modem is considered data communication equipment (DCE), and the device using the modem is called data terminal equipment (DTE). As indicated earlier, modems must adhere to a number of communication standards to work with other modems: Bell103, Bell212A, V.21, V.22, V.22bis, V.23, V.32, V.32bis, V.FC, and V.34, to name a few. These standards reflect a dual analog conversion model,Notice that the signal goes through only one analog conversion. Because the conversion is done on the client’s side, traffic generated by the client modem is limited to V.34 speeds. The traffic coming from the access server is not subject to the noise problems that an analog conversion would introduce, so it can be sent at much higher speeds. Thus, the client can receive data at v.90 speeds but can send data at only V.34 speeds.

PPP -

PPP bears mentioning because it is so vital to the operation of dialup technologies. Until PPP came along in 1989 (RFC 1134—currently up to RFC 1661), dialup protocols were specific to the protocol being used. To use multiple protocols, it was necessary to encapsulate any other protocols within packets of whatever protocol the dialup link was running. Many of the proprietary link methods (such as SLIP) didn’t even have the capability to negotiate addressing. Fortunately, PPP does this and many more things with flexibility and extensibility. PPP connection establishment happens in three phases: Link Control Protocol (LCP), authentication, and Network Control Protocol (NCP).

LCP-

LCP is the lowest layer of PPP. Because PPP does not follow a client/server model, both ends of the point-to-point connection must agree on the negotiated protocols. When negotiation begins, each of the peers wanting to establish a PPP connection must send a configure request (CONFREQ). Included in the CONFREQ are any options that are not the link default. These often include maximum receive unit, async control character map, authentication protocol, and the magic number. At this stage, the peers negotiate their authentication method and indicate whether they will support PPP multilink. In the general flow of LCP negotiations, there are three possible responses to any CONFREQ:

1. A configure-acknowledge (CONFACK) must be issued if the peer recognizes the options and agrees to the values seen in the CONFREQ.
2. A configure-reject (CONFREJ) must be sent if any of the options in the CONFREQ are not recognized (such as some vendor-specific options) or if the values for any of the options have been explicitly disallowed in the configuration of the peer.
3. A configure-negative-acknowledge (CONFNAK) must be sent if all the options in the CONFREQ are recognized, but the values are not acceptable to the peer. The two peers continue to exchange CONFREQs, CONFREJs, and CONFNAKs until each sends a CONFACK, until the dial connection is broken, or until one or both of the peers indicates that the negotiation cannot be completed.

Authentication

Authentication is an optional phase, but it is highly recommended on all dial connections. In some
instances, it is a requirement for proper operation—dialer profiles, being a case in point. The two principal types of authentication in PPP are the Password Authentication Protocol (PAP) and the Challenge Handshake Authentication Protocol (CHAP), defined by RFC 1334 and updated by RFC 1994. When discussing authentication, it is helpful to use the terms requester and authenticator to distinguish the roles played by the devices at either end of the connection, although either peer can act in either role. Requester describes the device that requests network access and supplies authentication information; the authenticator verifies the validity of the authentication information and either allows or disallows the connection. It is common for both peers to act in both roles when a DDR connection is being made between routers. PAP is fairly simple. After successful completion of the LCP negotiation, the requester repeatedly sends its username/password combination across the link until the authenticator responds with an acknowledgment or until the link is broken. The authenticator may disconnect the link if it determines that the username/password combination is not valid. CHAP is somewhat more complicated. The authenticator sends a challenge to the requester, which then responds with a value. This value is calculated by using a “one-way hash” function to hash the challenge and the CHAP password together. The resulting value is sent to the authenticator along with the requester’s CHAP host name (which may be different from its actual host name) in a response message. The authenticator reads the host name in the response message, looks up the expected password for that host name, and then calculates the value that it expects the requester to send in its response by performing the same hash function the requester performed. If the resulting values match, the authentication is successful. Failure should lead to a disconnection. By RFC standards, the authenticator can request another authentication at any time during the connection.

NCP-

NCP negotiation is conducted in much the same manner as LCP negotiation with CONFREQs, CONFREJs, CONFNAKs, and CONFACKs. However, in this phase of negotiation, the elements being negotiated have to do with higher-layer protocols—IP, IPX, bridging, CDP, and so on. One or more of these protocols may be negotiated. Refer to the following RFCs for more detail on their associated protocols:

• RFC 1332 “IP Control Protocol”
• RFC 1552 “IPX Control Protocol”
• RFC 1378 “AppleTalk Control Protocol”
• RFC 1638 “Bridging Control Protocol”
• RFC 1762 “DECnet Control Protocol”
• RFC 1763 “VINES Control Protocol”

A Couple of Advanced Considerations

The Multilink Point-to-Point Protocol (MLP, RFC 1990) feature provides a load-balanced method for splitting and recombining packets to a single end system across a logical pipe (also called a bundle) formed by multiple links. Multilink PPP provides bandwidth on demand and reduces transmission latency across WAN connections. At the same time, it provides multivendor interoperability, packet fragmentation with proper sequencing, and load calculation on both inbound and outbound traffic. The Cisco implementation of multilink PPP supports the fragmentation and packet sequencing specifications in RFC1717. Multilink PPP works over the following interface types (single or multiple):

• Asynchronous serial interfaces
• BRIs
• PRIs

Multichassis multilink PPP (MMP), on the other hand, provides the additional capability for links to terminate at multiple routers with different remote addresses. MMP can also handle both analog and digital traffic. This functionality is intended for situations in which there is a large pool of dial-in users, and a single access server cannot provide enough dial-in ports. MMP allows companies to provide a single dialup number to their users and to apply the same solution to analog and digital calls. This feature allows Internet service providers, for example, to allocate a single ISDN rotary number to several ISDN PRIs and not have to worry about whether a user’s second link is on the same router. MMP does not require reconfiguration of telephone company switches.

AAA

Another technology that should be mentioned because of its importance is Authentication, Authorization, and Accounting (AAA). The protocols used in AAA can be either TACACS or RADIUS. These two protocols were developed in support of a centralized method to keep track of users and accesses made on a network. AAA is employed by setting up a server (or group of servers) to centrally administer the user database. Information such as the user’s password, what address should be assigned to the user, and what protocols the user is allowed to run can be controlled and monitored from a single. workstation. AAA also has powerful auditing capabilities that can be used to follow administratively important trends such as connection speeds and disconnect reasons. Any medium or large dialup installation should be using AAA, and it’s not a bad idea for small shops, either.

Dialup Methods

Most routers support automated methods for dynamic links to be connected when traffic that needs to get to the other end arrives. Cisco’s implementation is called dial-on-demand routing (DDR). It provides WAN connectivity on an economical, as-needed basis, either as a primary link or as backup for a nondial serial link. At its heart, DDR is just an extension of routing. Interesting packets are routed to a dialer interface that triggers a dial attempt. Each of the concept’s dialer interface and interesting traffic bear explanation.

What’s a Dialer?

The term dialer has a few meanings, depending on the specifics of the configuration, but in general, it refers to the interface where the routing is actually happening. This is the interface that knows the address and phone number where the traffic is supposed to go. When looking at the routing table, the dialer interface should be the interface referenced for the next hop to reach the network on the other side. The dialer interface does not have to be the physical interface that is doing the dialing, but it can be made so by placing the configuration command dialer in-band in a physical interface. Thereafter, the interface becomes a dialer. For example, an async interface is not a dialer by default, but placing the configuration command dialer in-band in the async interface causes dialer behavior on that interface. For example, calls received by that async interface after applying the command will have an idle timeout applied to the connection from then on. An example of a physical interface that is also a dialer by default would be the BRI interface. Beyond making physical interfaces into dialers, there are interfaces called dialer interfaces. These are logical interfaces that call upon real interfaces to place calls. The advantage of using a dialer interface is flexibility. A group of potential DDR links can share a handful of BRI interfaces. Dialer interface configuration comes in two flavors: dialer map-based (sometimes referred to as legacy DDR) and dialer profiles. Which method you use depends on the circumstances under which you need dial connectivity. Dialer map-based DDR was first introduced in IOS Version 9.0; dialer profiles were introduced in IOS Version 11.2.

Interesting Traffic

The term interesting is used to describe packets or traffic that will either trigger a dial attempt or, if a dial link is already active, reset the idle timer on the dialer interface. For a packet to be considered interesting, it must have these characteristics:

• The packet must meet the “permit” criteria defined by an access list.
• The access list must be referenced by the dialer–list, or the packet must be of a protocol that is
universally permitted by the dialer–list.
• The dialer-list must be associated with a dialer interface by use of a dialer group.

Packets are never automatically considered to be interesting (by default). Interesting packet definitions must be explicitly declared in a router or access server configuration.

Benefits and Drawbacks

The benefits of dialup are flexibility and cost savings. First, let’s look at why flexibility is important. Intermittent connectivity is most often needed in mobile situations. A mobile workforce needs to be capable of connecting from wherever they are. Phone lines are normally available from wherever business is transacted, so a modem connection is the only reasonable choice for mobile users. In long-distance situations, a user often dials into a local ISP and uses an IPSec-encrypted tunnel going back to a home gateway system that allows access to the rest of the corporate network. In this example, the phone call itself costs nothing, and an account with the local ISP could be significantly less expensive than the long-distance charges that would otherwise be incurred. As another example, a BRI attached at a central office located in an area that offers inexpensive rates on ISDN could have database servers configured to call out to other sites and exchange data periodically. Each site needs only one BRI line, which is significantly less expensive than dedicated links to each of the remote locations. Finally, in the case of a backup link, the savings are seen when the primary link goes down but business continues, albeit slower than normal. Cost savings is a two-edged sword where dialup is concerned, however. The downside of a dialup line is that connection costs for a heavily used line are higher than the price of dedicated connectivity. Going over long distance raises the price even higher. There’s also speed to consider. Dialup connectivity has a strong high-end bandwidth, particularly with the capability to tie channels together using PPP multilink, but dedicated connectivity through a serial port can outperform dialup connections. Another consideration is security. Certainly, any PPP connection should be authenticated, but this presents anyone with the dialup number an opportunity to break into the system. A significant part of any dialup system’s configuration concerns the capability to keep out unwanted guests. The good news is that it can be done, and AAA goes a long way toward dealing with this problem. However, it is a disadvantage to have potential intruders coming in through dialup lines.