Configuring Dynamic Addressing

The commands to configure dynamic addressing depend on the method used. Although DHCP
is one option (used as an example in the following text), there are other methods, including proprietary
ones.

Before we discuss incorporating a dynamic addressing solution, it is best to acknowledge the
option of manually addressing the client. In Windows 95/98, this is accomplished by using the Dial-
Up Connection Properties menu to access the TCP/IP Settings dialog box. This dialog box is shown
in Figure 24.2. Note that you must select Specify an IP Address to manually enter a selection.
Windows will provide a warning if you attempt to use Network Control Panel to configure
the dial-up adapter, as shown in Figure 24.3. As shown, configuration parameters in properties
will overwrite any custom parameters on the individual dial-up connection. Many users might
connect to different locations, with each location requiring a different set of parameters. As
such, the warning is well heeded and administrators will likely choose to configure all settings
per connection.
If you are configuring a router to provide the dial-up connection between the client and
remote access server, you should use the standard Ethernet configuration commands. These
entries, shown in Figure 24.4, include the IP address, subnet mask, default gateway, and
name servers. Please note that although a static configuration is shown, the administrator
could use DHCP.

Manual IP address configuration in Windows 95/98


Configuring a dial-up adapter from Windows Control Panel


Ethernet-based manual IP address configuration in Windows 95/98


NOTE:The configuration dialog box is accessed through Control Panel 
Network 
TCP/IP 
Adapter.

On the router, the configuration is straightforward, but it depends on the role of the router
and the type of dynamic assignment desired. The
async dynamic address
command enables
the client to provide its address, but the
peer default ip address [ip-address | dhcp |
pool
poolname
]
command is used more often. This command enables the administrator to
select manual, DHCP, or pool-based address selection.
When selecting the DHCP option, the administrator must also configure the router for one
of three choices:
 IP helper address
 IP DHCP server
 DHCP server on router
The IP helper address option is often found in router configurations, but without additional
configuration, this option will forward broadcast traffic to the helper address. The helper
address is the address of the server or group of servers that provide the required service—
DHCP, in this example.
A newer command is ip dhcp-server, which the administrator can use to specify the
address of the DHCP server.
In addition, some routers might also provide DHCP server functionality. This should be
considered for smaller installations only; routers are best suited to provide routing. However,
this feature might be ideal for small office/home office installations.


NOTE:It is important to remember that certain IP broadcast traffic will be forwarded to
the helper address by default. This can be blocked to include only DHCP datagrams
by using the no ip forward-protocol udp udp-port-number command.
The following UDP (User Datagram Protocol) ports are enabled by default: 69
(TFTP), 53 (DNS), 37 (Time), 42 (name server), 49 (TACACS), 67 (BOOTP Client),
68 (BOOTP Server), 137 (NetBIOS WINS), and 138 (NetBIOS datagram). BOOTP
(Bootstrap Protocol) was the predecessor to DHCP and shares the same UDP
port numbers.To configure DHCP services on the router, the administrator must first decide if they
wish to use a DHCP database agent to help manage the lease process. Cisco calls this feature
conflict logging.
If conflict logging is desired, the administrator must also configure an FTP or TFTP server,
which is defined with the ip dhcp database command. If the administrator does not wish
to implement conflict logging, the command no ip dhcp conflict logging must be used
instead. Note that in some instances the administrator must exclude an address from the
DHCP pool. To do this, they must use the ip dhcp excluded-address low-address
{high-address} command.
An entire configuration file for DHCP services is shown here:
service dhcp
ip dhcp database ftp://dhcp:cisco@10.11.1.10/dhcp
ip dhcp pool 0
network 10.10.1.0 /24
default-router 10.10.1.1
domain-name foo.com
dns-server 10.2.20.51
netbios-name-server 10.2.20.51
The preceding configuration example uses an FTP server at 10.11.1.10 to capture information
regarding the DHCP leases. The pool is for 10.10.1.0/24 and a default gateway of 10.10.1.1. The
domain is foo.com, and DNS and WINS services are provided by 10.2.20.51. The service dhcp
command used here is optional; the service is available by default. The FTP server username is
dhcp with a password of cisco in the preceding output; however, this is not a very secure option.
Although this chapter focuses on Windows 95/98 configuration, readers should note that
Windows NT and 2000 differ little in most regards. Figure 24.5 shows Windows 2000’s Dial-Up
Connection Properties dialog box.