Implementing IP Routing

Archived content. No warranty is made as to technical accuracy. Content may contain URLs that were valid when originally published, but now link to sites or pages that no longer exist.

By Todd Lammle,with Monica Lammle and James Chellis

Chapter 3 from MCSE: TCP/IP for NT Server 4 Study Guide, published by Sybex, Inc.

With TCP/IP basics covered and conquered in Chapters 1 and 2, our focus is going to both sharpen and shift. Our attention will now be concentrated on Microsoft-specific issues and intricacies. We will begin Chapter 3 with a crisp discussion on IP routing.

On This Page

Objectives
What Is IP Routing?
IP Routing Applied
Supernetting
Summary
Review Questions

Objectives

Our topic scope will encompass several important objectives. By the end of this chapter, you should be able to do the following:

  • Explain the difference between static and dynamic IP routing.

  • Explain the host configuration requirements to communicate with a static or dynamic IP router.

  • Configure a computer running Windows NT to function as an IP router.

  • Build a static routing table.

  • Use the TRACERT utility to isolate route or network link problems.

Tip All readers—even the wizards and gurus in the audience—shouldn't skip class today. Like I said, we're moving into Microsoft Land now, and this chapter is fundamental because it deals with IP routing as spoken in MSNT—an important Microsoft dialect.

What Is IP Routing?

IP routing is the process of sending data from a host on one network to a remote host on another network through a router, or routers. A router is either a specifically assigned computer or a workstation that's been configured to perform routing tasks. In IP terminology, routers are referred to as gateways. Gateways are basically TCP/IP hosts that have been rigged with two or more network connection interfaces. Outfitted in this manner, they're known as multihomedhosts, which we'll discuss more thoroughly later in the chapter.

The path that a router uses to deliver a packet is defined in its routing table. A routing table contains the IP addresses of router interfaces that connect to the other networks the router can communicate with. The routing table is consulted for a path to the network that is indicated by the packet's destination address. If a path isn't found, the packet is sent to the router's default gateway address—if one is configured. By default, a router can send packets to any network for which it has a configured interface. When one host attempts communication with another host on a different network, IP uses the host's default gateway address to deliver the packet to the corresponding router. When a route is found, the packet is sent to the proper network, then onward to the destination host. If a route is not found, an error message is sent to the source host.

The IP Routing Process

The IP routing process is fairly direct when a datagram's destination is located on a neighboring network. In this kind of situation, a router would follow a simple procedure, as shown in Figure 3.1.

Cc750576.implip01(en-us,TechNet.10).gif

Figure 3.1: Simple routing

First, when a workstation wants to send a packet to a destination host, in this instance 160.1.0.1 transmitting to 160.2.0.4, host 160.1.0.1 checks the destination IP address. If it determines the address isn't on the local network, it must then be routed. Next, 160.1.0.1 calls on ARP to obtain the hardware address of its default gateway. The IP address of the default gateway is configured in machine 160.1.0.1's internal configuration, but 160.1.0.1 still needs to find the hardware address of the default gateway, and sends out an ARP request to get it. IP then proceeds to address the packet with the newly obtained destination hardware address of its default router. The information utilized for addressing the packet includes:

  • Source hardware address 1

  • Source IP address 160.1.0.1

  • Destination hardware address 5

  • Destination IP address 160.2.0.4

IP, on the receiving router with the hardware address of 5, establishes that it is not the final, intended recipient by inspecting the packet's destination IP address, which indicates it must be forwarded to network 160.2. Then, IP uses ARP to determine the hardware address for 160.2.0.4. The router then puts the newly identified hardware address into it's ARP cache for easy reference the next time it's called upon to route a packet to that destination.

This accomplished, the router sends the packet out to network 160.2 with a header that includes:

  • Source hardware address 5

  • Source IP address 160.1.0.1

  • Destination hardware address 10

  • Destination IP address 160.2.0.4

As the packet travels along network 160.2, it looks for hardware address 10, with the IP address of 160.2.0.4. When an NIC card recognizes its hardware address, it grabs the packet.

It's important to note here that the source IP address is that of the host that created the packet originally, but that the hardware address is now that of the router's connection interface to network 160.1. It's also significant that although both source and destination software IP addresses remain constant, both source and destination hardware addresses necessarily change at each hop the packet makes.

Sounds simple right? Well, it is in a situation like the one we just presented. However, those of you who possess some firsthand experience with this sort of thing may now be finding yourselves just a little distracted with thoughts of a genuinely sarcastic variety. Before turning your nose up and slamming this book shut, let it be known that we, too, are fully aware that this isn't a perfect world, and that if things were that straightforward, there wouldn't be a market for books about them! On the other hand, to those readers becoming uncomfortable with the now present implications of potential chaos, we say, relax, make some tea, sit down, and read on.

Start by considering this foul and ugly possibility: What if the destination network is in the dark because it's not directly attached to a router on the delivery path for that nice little datagram? Things come a tumblin' down, that's what! Remember hearing somewhere that there's a reason for everything? Well, we're not sure about that, but the heinous, confusion-producing scenario we just posited is one most excellent reason for the existence of routing tables. With a handy-dandy routing table, the fog clears, clouds part, and destinations sing! Routers, and those dependent on them, again become happy, efficient things. Routing tables are maintained on IP routers. IP consults these to determine where the mystery network is, so that it can send its mystery packet there. Some internetworks are very complex. If this is the case, routing tables should designate all available routes to a destination network, as well as provide an estimate advising the efficiency of each potential route. Routing tables maintain entries of where networks are located, not hosts.

Dynamic vs. Static IP Routing

There are two breeds of routing tables. There are static tables, and there are dynamic tables. Static types are laboriously maintained by a network manager, while the dynamic variety is sustained automatically by a routing protocol. Here's a list spotlighting some specific routing table characteristics:

Dynamic Routing

Static Routing

Function of inter-routing protocols

Function of IP

Routers share routing information automatically

Routers do not share routing information

Routing tables are built dynamically

Routing tables are built manually

Requires a routing protocol, such as RIP or OSPF

Microsoft supports multihomed systems as routers

Microsoft supports RIP for IP and IPX/SPX

 

Windows NT Server 4.0 provides the ability to function as an IP router using both static and dynamic routing. A Windows NT-based computer can be configured with multiple network adapters and can route between them. This type of system, which is ideal for small, private internetworks, is referred to as a multihomed computer.

Windows NT Server 4.0 has the ability to function as a Routing Information Protocol (RIP) router that supports dynamic management of Internet Protocol (IP) routing tables. RIP eliminates the need to establish static IP routing tables.

Dynamic IP Routing

On large internetworks, dynamic routing is typically employed. This is because manually maintaining a static routing table would be overwhelmingly tedious, if not impossible. With dynamic routing, minimal configuration is required by a network administrator. Figure 3.2 shows an example of dynamic routing.

Cc750576.implip02(en-us,TechNet.10).gif

Figure 3.2: An example of dynamic routing

For a host to communicate with other hosts on the internetwork, its default gateway address must be configured to match the IP address of the local router's interface.

In Figure 3.2, Host 1 requires a default gateway address in order to be able to send packets to any network other than Network A. Host 1's default gateway address is configured for the router port attached to Network A. If Host 1 sends a packet that's not destined for the local network, it will be sent to the default gateway address. If no gateway is defined, the packet will be discarded. Host 2 works the same way; however, it's default gateway is the router port attached to Network C. When the router receives a packet either from Host 1 or 2, it will observe the destination's IP address and forward it according to the information in its routing table, which is built and maintained through inter-routing protocols.

Dynamic routing is a function of inter-routing, network gossip protocols such as the Routing Information Protocol (RIP) and Open Shortest Path First (OSPF). These routing protocols periodically exchange routes to known networks among dynamic routers. If a given route changes, they automatically update the router's routing table and inform other routers on the internetwork of the change.

Routing Information Protocol (RIP)

RIP is a type of protocol known as a distance vector routing protocol. RIP is used to discover the cost of a given route in terms of hops, and store that information in the routing table, which IP uses in selecting the most efficient, low-cost route to a destination. It works by watching for routing table broadcasts by other routers and updating its own routing table in the event a change occurs. RIP routing tables provide, at a minimum, the following information:

  • IP destination address

  • A metric (numbered from 1 to 15) indicative of the total cost in hops of a certain route to a destination

  • The IP address of the router a datagram would reach next on the path to its destination

  • A marker signaling recent changes to a route

  • Timers

Some drawbacks to RIP include a problem known as "counting to infinity," as illustrated in Figure 3.3. In certain internetwork configurations, an endless loop between routers can occur if one of the networks becomes unavailable. RIP keeps counting hops each time the broadcast reaches a router, in hopes of finding a new route to the formerly available network. To prevent this, a hop-limit count between 1 and 15 is configured to represent infinity, which necessarily imposes size restrictions on networks. RIP can't be utilized on a network with an area consisting of more than 15 hops. In Figure 3.3, Network 6's location was lost between Routers B and D. Router B then looks for a new route to Network 6. Router B already knows that Router C can get to Network 6 with four hops because Router C advertised this information in a broadcast, and all routers save this broadcasted information in their routing tables. Since Router B is looking for a new route to Network 6, Router B references its routing table and finds that Router C can reach Network 6 in four hops. Router B determines it can reach Network 6 in five hops because Router C can make it in four hops. This is because Router B must add an extra hop for itself—four from Router C plus one for Router B. Router B then broadcasts the new route information back out onto the network. Router C receives this information, and enters into its route table that Network 6 is now six hops away—five from Router B, plus one for itself. This process continues until the 15-hop limit is reached. At this point, the route to Network 6 is finally dubbed an unreachable destination, and all related route information regarding it is removed from both Router B's and Router C's routing tables.

Cc750576.implip03(en-us,TechNet.10).gif

Figure 3.3: "Counting to infinity"

Another problem with large internetworks centers around the fact that RIP routers broadcast routing table advertisements every 30 seconds. On today's gargantuan networks, populated with an abundance of routers, momentous amounts of bandwidth can get gobbled up simply accommodating all the RIP response packet noise.

Open Shortest Path First (OSPF)

Because of these potentially network-hostile characteristics, OSPF is quickly gaining popularity within the Internet community. OSPF is based on link-state algorithms, and is therefore known as a link-state routing protocol. It's deployed within an autonomous system, which is a group of routers that share a certain routing protocol. When that protocol happens to be OSPF, each router retains its own database describing the topology of the autonomous system on which it's located. This kind of system is much more flexible, and has the following advantages as well:

  • Network administrators can assign costs to a particular link.

  • The total cost for a given path doesn't necessarily have to have a limit.

  • Its upper metric limit being 65,535, it has the ability to accommodate vast networks.

  • Each node creates a link-state database tree representing the network and places itself as that tree's root, where it can choose the most direct path to a given destination.

  • Related to the above benefit, in the event that more than one route of equal cost exists, OSPF routers can balance the load of network traffic between all available and equally cost-effective routes.

  • Link-state routing advertisements are broadcasted much less often—only when a change is detected, thereby reducing network overhead.

  • Link-state routing update packets can efficiently carry information for more than one router.

  • This type of packet is only sent to adjacencies, or neighboring routers selected to swap routing information—a "tell a friend" arrangement that again contributes to network efficiency.

Static IP Routing

Static routing is a function of IP. Static routers require that routing tables be built and updated manually. If a route changes, static routers are secretive and do not share this information to inform each other of the event. They're also very cliquey, and do not exchange routes with dynamic routers.

Windows NT can function as an IP router using static routing. NT network administrators must maintain their tables or acquire a commercial router. A Windows NT-based computer can be configured with multiple network adapters and routes between them. This type of system, which is ideal for small, private internetworks, is referred to as a multihomed computer.

Routing tables inventory known networks and the IP addresses used to access them. Windows NT static routing tables are maintained by a route utility, and are comprised of five columns of data, reading left to right. In the list below, the first entry represents the left-most column, the second represents the next one, and so on.

Network Address: A roster of addresses for known networks. Included here is an entry both for the local network (0.0.0.0) and for broadcasts (255.255.255.255).

Netmask: This column lists all subnet masks in use for each network.

Gateway Address: This is a list of the IP addresses employed as the primary datagram receivers for each network.

Interface: Each network card installed in a computer is assigned an interface number.

Metric: This is a list providing an estimate of the number of hops the route would cost. A hop is each pass a datagram makes through a router.

Here are the key things to remember about static routers:

  • A static router can only communicate with networks with which it has a configured interface.

  • A Windows NT computer can be configured as a multihomed computer.

  • A static route can be configured as either a default gateway address or an entry in a routing table.

  • A static router, such as a Windows NT multihomed computer, can only communicate with networks to which it has a configured interface. This limits communications to local networks.

Figure 3.4 illustrates static routing.

Cc750576.implip04(en-us,TechNet.10).gif

Figure 3.4: Static routers

As shown in Figure 3.4, Host 1 has local connections to Networks A and B. This means that hosts on Network A can communicate with hosts on Network B, and vice versa, because Host 1 knows about both networks and will pass packets destined for either one. Hosts on Network A will not be able to communicate with hosts on Network C.

Host 2 has local connections to Networks B and C and will be able to pass packets destined for either network. However, Network C will not be able to send packets to Network A.

Note: After taking all this in, you may have been left with the impression that dynamic routing is the method of choice for everyone's routing needs. While that's certainly true when the network in question is large and complex, providing a multiplicity of paths to destinations and/or growing rapidly, static routing is wonderfully suited for small to moderately sized networks that rarely change. An important consideration is, as is so often the case, cost. All that fabulous intelligence and flexibility, and all those bells and whistles that dynamic routers offer cost a lot—up to around $50k or more apiece! They're one of the most expensive pieces of equipment one can hook up to a network! Windows NT comes out of the box equipped with static routing built right in—in other words...it's free! It's also free of charge in terms of overhead costs on your network, and it creates the environment for a much closer, more involved relationship between you and your beloved network.

IP Routing Applied

Now that you have a clear picture of exactly what IP routing is, and what it involves, you're ready to learn how it's done. In this section, we'll give you the skinny on configuration and integration issues, and the procedures required for implementation.

Using the Default Gateway Address on a Static Router

Gateways are most often dedicated computers, or routers. The default gateway is like a network mediator with connections. It's the node on the local network that knows the network IDs of other networks linked to the greater internetwork. Since it has access to this privileged information, when a given workstation sends out some data that reaches the default gateway, it can forward it along to other gateways as required to reach its proper destination.

One method of configuring a static route without manually adding routes to a routing table is to configure each multihomed computer's default gateway address as the local interface to the other multihomed computer on the common network. It's a type of circular reasoning for computers.

A multihomed computer (a computer with more than one NIC card) can send IP packets to destinations other than those they are locally attached to by setting the internal configuration of the default gateway to the other multihomed computer's network interface. For example, in Figure 3.4, Host 2 would set its default gateway to the network interface on Host 1. Network C then would be able to pass packets to Network A. Host 1 would set its default gateway to the network interface on Host 2, enabling Network A to communicate with Network C.

Whenever Host 1 receives a packet destined for a host on Network C, it'll check its local routing table. If it doesn't find a route to Network C, it forwards the packet to its default gateway, which is a local interface on Host 2. Host 2 will then route the packet to the appropriate interface for delivery on Network C.

Note: Using the default gateway address as a static route only works well with two routers. If more than two are used, you must manually add an entry in the routing table.

Using Additional Default Gateways

Although more than one default gateway can be configured, only the first one will be used for routing purposes. The others will be used only as backups should the primary one become unavailable for some reason. This means you can't use multiple gateways to nab more network bandwidth. However, better fault tolerance is still a definite plus. As I'm sure you are well aware, this backup stuff is by no means unimportant.

Let's say Router A in Figure 3.5 goes on the blink and is out of commission when the client boots up. The client wants to connect to the server, but the default gateway that the client is defined for is currently unavailable, so it's out of luck. The client has no other default gateway defined, so the client will not be redirected to Router C to connect to the server. However, if the client was to define a second default gateway for Router C, the client would then be directed on towards the server.

Cc750576.implip05(en-us,TechNet.10).gif

Figure 3.5: A hypothetical routing dilemma

The client is going to have to sit on the other side of the abyss dreaming of multiple default gateways, wishing he or she had a more thorough network administrator who took these kind of precautions. Would that be you? If so, Exercise 3.2 shows you how to create additional default gateways using the Advanced TCP/IP configuration.

When one configures one's Windows NT system in this way, retransmission problems at the TCP layer will cause the IP routing software to try the routers listed in the Additional Gateways value. This is again a great backup plan. Why? Well…what if Client and Server were in the middle of an established session, and that troublesome Router A went down again? TCP would send a message to IP to try one of the additional routers in the registry. IP would then try Router C, use a double hop route, and continue exchanging data at no cost to the session. When Router A again breathes a breath of life, the inter-routing protocol will force Router C to redirect the session's traffic back through the more optimal, one-hop route provided by the now living Router A.

Note: In order for any host to be able to communicate with other hosts located somewhere out there on the internetwork, its default gateway address positively must be configured to match the IP address of the router's interface on the local segment.

Default Routing Table Entries

Windows NT 4.0, NT 3.51, and Windows for Workgroups 3.11 routing tables maintain the default entries shown in Table 3.1.

Table 3.1 Default routing tables for Windows NT 4.0, 3.51, and Windows for Workgroups 3.11

Address Examples

Description

0.0.0.0

The address used as a default route for any network not specified in the route table

Subnet broadcast

The address used for broadcasting on the local subnet

Network broadcast

The address used for broadcasting throughout the internetwork

Local loopback 127.0.0.1

The address used for testing IP configurations and connections

Local network

The address used to direct packets to hosts on the local network

Local host

The address of the local computer; references the local loopback address

Adding Static Entries

The route command is used, as detailed in Table 3.2, to add static entries to a routing table.

Table 3.2 Commands Used for Adding Static Entries

To add or modify a static route

Function

route add [network address] mask [gateway address]

Adds a route

route delete [network address] [gateway address]

Deletes a route

route change [network address] [gateway address]

Modifies a route

route print [network address] [gateway address]

Prints a route

route -s [gateway address]

Adds a route to a smart gateway

route -f

Clears all routes

It works like this: To add a static route, enabling communications between a host on network 160.1.89.0 from a host on network 160.1.66.0, you would run the following command:

route add 160.1.24.0 mask 255.255.255.0 160.1.16.1

Note: Static routes are stored in memory unless the -p parameter is used. No, it doesn't stand for permanent; it stands for persistent. Persistent routes are stored in the registry. If you restart your NT Server or workstation, you will have to re-create all nonpersistent routes if you're using a static routing table.

If your internetwork has more than two routers, at least one of which is a static router, you'll need to configure static routing table entries to all known networks on a table at each multihomed computer, as shown in Figure 3.6.

Cc750576.implip06(en-us,TechNet.10).gif

Figure 3.6: Proper routing table configuration

A static routing table for Network C, IP address 160.1.43.0, is created on multihomed Host 1, directing the router to send the packets to interface 160.1.89.1. This will permit packets to move from Network A to Network C.

Also, a static route is configured on Host 2 to allow Network C to send packets to Network A. This is accomplished by referencing the network address 160.1.66.0, directing the router to send packets on to IP address 160.1.89.2, which can directly deliver packets destined for Network A.

The static routing table is always checked before a packet is routed. If there is no static route to a particular host, the packet is sent to the configured default gateway. For a host to communicate with other hosts on the internetwork, its default gateway address must be configured to match the IP address of the local router's interface.

Integrating Static and Dynamic Routing

As mentioned above, breeds of routers stick to their own and do not speak with those of a different feather. Static routers do not trade routing information with dynamic routers unless they're forced to. As they say, where there's a will, there's a way! To route from a static router through a dynamic router, such as a RIP or OSPF-enabled IP router, one must first add a static route to the routing tables located on both the static and dynamic routers, as shown in Figure 3.7.

Cc750576.implip07(en-us,TechNet.10).gif

Figure 3.7: Static and dynamic router integration

In order to route packets from Network A to the Internet, Host 2 requires that a route be added to its routing table. This addition includes the IP address of the closest interface (in this case, 160.1.66.2) that can access the dedicated IP router to the Internet.

To route packets from Networks B and C to the Internet, a static entry must be added to Computer B's routing table. This entry includes the IP address of the nearest interface (160.1.89.2) on the dedicated IP router to the Internet.

To allow computers on the Internet to communicate with hosts on Networks 1 and 2, one must statically configure the dedicated IP router with the IP address of the interface to Host 1. Host 1 then acts as a gateway to other subnets.

Note: It's important to mention here that some implementations of RIP do not propagate static routing tables. Should you find yourself beset with this dilemma, you'd have to statically configure the remote routers in the Internet cloud to achieve your routing goals. Also important to note here is that the exact method for configuring a static route on a RIP router varies with each kind of router. It is therefore very wise to refer to your particular router's vendor documentation for more information.

In Exercise 3.1, you'll view and configure the routing table.

Exercise 3.1: Viewing and Configuring the Routing Table

  1. From a command prompt, type route print, and then press Enter to view the route table.

    Cc750576.implip08(en-us,TechNet.10).gif

  2. Under Gateway Address, you should find your default gateway address—the router interface address attached to your local network interface.

  3. Remove your default gateway address from your computer. This will prevent any packets being sent to the default gateway for routing and require all routing to be done from existing route entries.

  4. Access the Microsoft TCP/IP Properties dialog box.

  5. Delete the Default Gateway address.

  6. Click OK in the Microsoft TCP/IP Properties box.

  7. Click OK.

  8. Switch to a command prompt, and use the route print command. The default gateway should not be listed.

Next, attempt communication with both local and remote hosts.

  1. Ping the IP address of a local host. This should be successful.

  2. Ping the IP address of a host on a remote address. This should fail, with a "Destination host unreachable" error.

  3. Add a route entry to your computer.

  4. Type the following command: route add remote_network_id mask subnet_mask your_default_gateway . (For example, if your workstation is on network 150.150.28.0 and the remote network is 150.150.40.0, the command would look like this: route add 150.150.40.0 mask 255.255.252.0 150.150.28.200. This means that to get to network 150.150.40.0, you would use address 150.150.28.200, which should be the default gateway.)

  5. View the route table.

  6. Ping a remote host. This should be successful.

  7. Restore your default gateway.

  8. Test communication by pinging your default gateway.

Implementing a Windows NT Router

Windows NT originally shipped with the ability to act as a static IP router. The static IP router was enabled by creating a multihomed system and enabling routing either through the registry or through a checkbox in the Advanced TCP/IP Configuration dialog box.

Static routing can work well for small networks and remote sites, but for large internetworks, the overhead of manually maintaining routing tables is significant. By enabling the RIP for IP routing protocol, Windows NT Server 4.0 can be a dynamic IP router. Windows NT 4.0 RIP for IP eliminates the manual configuration of routing tables. RIP for IP is suitable for medium-size internetworks, but it is not suitable for large IP internetworks because of the significant amount of broadcast traffic it generates.

To implement a Windows NT Router:

  1. Install multiple adapter cards and appropriate drivers, or configure multiple IP addresses on a single card.

  2. Configure the adapter card(s) with a valid IP address and subnet mask.

  3. On the Routing tab of the Microsoft TCP/IP Properties dialog box, select the Enable IP Forwarding checkbox.

    Depending on which version of Windows NT you are running:

    • On the Services tab of the Control Panel Network tool, add the RIP for Internet Protocol service.

    • Add static routes to the static routers routing table for all networks to which the computer has no configured interface.

The TRACERT Utility

The TRACERT utility is essentially a verification tool. It's used to substantiate the route that's been taken to a destination host. TRACERT is also highly useful in isolating routers and identifying WAN Links that are not functioning and/or are operating too slowly. Here's the relevant command syntax for deploying TRACERT:

tracert 160.1.89.100

where 160.1.89.100 is the remote computer.

Cc750576.implip09(en-us,TechNet.10).gif

Below is an example of the output that would result from entering the command if the computer was on the network:

Tracing route to 160.1.89.100 over 2 hops1 <10 ms <10 ms <10 ms¬ 160.1.89.22
<10 ms <10 ms <10 ms 160.1.89.100

TRACERT can aid in determining if a router has failed by the degree of success the command enjoys. For example, if the command is unsuccessful, it is possible to assess router or WAN link problems and identify the point at which routing failed. The response time for the command is returned in the output. The information contained in the output can be readily compared to that recorded for another route to the same destination. This greatly facilitates identifying a slow or ineffective router or WAN link.

For example, the command tracert 160.1.66.11 would display the path taken from the local host to the destination host: 160.1.66.11. The output from the preceding command would confirm that the router address 160.1.66.1 was the route taken from the local host to the destination host. Here's what that output would look like:

Tracing route to 160.1.66.11 over a maximum of 30 hops1 <10 ms¬ <10 ms <10 ms
160.1.66.12 <10 ms <10 ms <10 ms 160.1.66.11

Let's go to the Internet and do a tracert to the White House. This is the output we receive:

C:\WINDOWS>tracert www.whitehouse.gov
Tracing route to www.whitehouse.gov [198.137.240.92]
over a maximum of 30 hops:
1 191 ms 136 ms 138 ms 38.1.1.1
2 142 ms 132 ms 137 ms 38.17.3.1
3 157 ms 151 ms 152 ms 38.1.42.3
4 152 ms 149 ms 147 ms 38.1.42.3
5 206 ms 285 ms 338 ms se.sc.psi.net [38.1.3.5]
6 286 ms 290 ms 265 ms rc5.southeast.us.psi.net [38.1.25.5]
7 278 ms 317 ms 268 ms ip2.ci3.herndon.va.us.psi.net [38.25.11.2]
8 268 ms 264 ms 313 ms 198.137.240.33
9 288 ms 279 ms 346 ms www.whitehouse.gov [198.137.240.92]
Trace complete.
C:\WINDOWS>

IPv6 (IPng)

This protocol used to be referred to as "IP Next Generation" or "IPng." IP version 6 (IPv6) is a new version of the Internet Protocol designed as a successor to IP version 4 (Ipv4, RFC-791). The current header in IPv4 hasn't been changed or upgraded since the 1970s! The initial design, of course, failed to anticipate the growth of the Internet, and the eventual exhaustion of the Ipv4 address space. Ipv6 is an entirely new packet structure, which is incompatible with IPv4 systems.

Expanded Addressing Capabilities

IPv6 has a128-bit source and destination IP addresses. With approximately five billion people in the world using 128-bit addresses, there are 2128 addresses, or almost 296 addresses per person! An IPv6 valid IP address will look something like this:

3F3A:AE67:F240:56C4:3409:AE52:220E:3112

IPv6 uses 16 octets; when written, it is divided into eight octet pairs, separated by colons and represented in hexadecimal format.

Header Format Simplification

The IPv6 headers are designed to keep the IP header overhead to a minimum by moving nonessential fields and option fields to extension headers after the IP header. Anything not included in the base IPv6 header can be added through IP extension headers placed after the base IPv6 header.

Improved Support for Extensions and Options

IPv6 can easily be extended for unforeseen features by adding extension headers and option fields after the IPv6 base header. Support for new hardware or application technologies is built in.

Flow Labeling Capability

A new field in the IPv6 header allows the pre-allocation of network resources along a path, so time-dependent services such as voice and video are guaranteed a requested bandwidth with a fixed delay.

Note: Not to worry, they don't ask you to subnet IPv6 on the NT4.0 TCP/IP test.

Supernetting

The Internet is running out of IP addresses. To prevent the complete depletion of network IDs, the Internet authorities have come up with a scheme called supernetting. The main difference between subnetting and supernetting is that supernetting borrows bits from the network ID and masks them as the host ID for more efficient routing. For example, rather then allocate a Class B network ID to a company that has 1800 hosts, the InterNIC allocates a range of eight Class C network IDs. Each Class C network ID gives 254 available hosts for a total of 2,032 host IDs.

However, this means that the routers on the Internet now must have an additional eight entries in their routing tables to route IP packets to the company. To prevent this problem, a technique called Classless Inter-Domain Routing (CIDR) is used to collapse the eight entries used in the above example to a single entry corresponding to all of the Class C network IDs used by that company.

Let's take a look at how this would work with eight Class C networks IDs starting with the network ID 220.78.168.0 through the network ID 220.78.175.0. In the routing table the entry then would be:

Network ID

Subnet Mask

Subnet Mask Binary

220.78.168.0

255.255.248.0

11111111.11111111.11111000.00000000

A typical Class C network would have a subnet mask of 255.255.255.0. Only the fourth octet would be available for subnetting and hosts. However, in supernetting, we can use bits in the usually reserved third octet to combine networks. With a 255.255.248.0 subnet, counting the zeros, we have three bits to work with. Each bit can be either a 1 or a 0. So, for each bit we get two choices. With three bits, it then becomes 23=8 subnets (don't subtract 2 in supernetting like you would in subnetting). Because the first network ID is 220.78.168.0, the router know to count up eight networks starting with 220.78.168.0 and going to 220.78.175.0.

Summary

Well, we hope you've found all of this helpful and enlightening. To make sure none of it got lost in the shuffle, let's take a moment to recap things.

We explained the difference between static and dynamic IP routing, and also the host configuration requirements to communicate with a static or dynamic IP router. We configured a computer running Windows NT to function as an IP router. We also talked about building a static routing table and the use of the TRACERT utility, which allows us to isolate route or network link problems.

In regards to the NT TCP/IP 4.0 test, we hope you were paying attention in this chapter. Microsoft seems to want you to understand how routing works. So, to make sure you were paying attention, go through the exercises below.

Review Questions

You have three Class C network addresses: 203.200.5.0, 203.200.6.0, and 203.200.7.0. You want to combine these addresses into one logical network to increase the host IDs that you can have on one subnet. Which subnet mask should you assign?

  1. 255.255.252.0

  2. 255.255.254.0

  3. 255.255.255.252

  4. 255.255.255.254

You can ping all the computers on your subnet and your default gateway, but you cannot ping any of the computers on a remote subnet. Other users on your subnet can ping computers on the remote subnet. What could be the problem?

  1. The subnet mask on the router is invalid.

  2. The subnet mask on your computer is invalid.

  3. The default gateway address on your computer is invalid.

  4. The computers on the remote subnet are not WINS enabled.

  5. The route to the remote subnet has not been established on the router.

Using Windows NT Explorer, your Windows NT Workstation can connect to a remote server, but not to a server on your local subnet. What is most likely the cause of the problem?

  1. Invalid default gateway address on your workstation

  2. Invalid default gateway address on the local server

  3. Invalid subnet mask on your workstation

  4. Invalid subnet mask on the remote server

Using Windows NT Explorer, your Windows NT Workstation cannot connect to a local server, but all other users can. When you run Network Monitor, you notice that each time the workstation attempts to connect to the server, it broadcasts an ARP request for the default gateway. What is most likely the cause of the problem?

  1. Invalid default gateway address on the workstation

  2. Invalid subnet mask on the workstation

  3. The workstation has a duplicate IP address.

  4. The workstation is not configured to use WINS.

Your NT Workstation cannot connect to a remote server, but all other workstations can. The network is configured as follows: two subnets, one router. The router has two interfaces: Network A with 131.107.32.1 and Network B as 131.107.64.1. All computers use a subnet mask of 255.255.240.0. You are located on Network B. When you run IPCONFIG on your workstation, you receive the following output:

  • IP Address

131.107.82.17

Subnet Mask

255.255.240.0

Default Gateway

131.108.64.1

What is most likely the cause of the problem?

  1. IP address on workstation is invalid

  2. Subnet mask on workstation is invalid

  3. Default gateway on workstation is invalid

  4. IP address on server is invalid

  5. Default gateway on server is invalid

You are troubleshooting a Windows NT Server computer on a TCP/IP network. The server is located on a subnet with a network ID of 142.170.2.0. The default gateway address is 142.170.2.1. Users on a remote subnet cannot access the server. You run ipconfig /all at the server and receive the following output:

  • Host Name

pdctest

DNS Servers

 

Node Type

Hybrid

NetBIOS scope ID

 

IP Routing Enabled

no

WINS Proxy Enabled

no

NetBIOS Resolution Uses DNS

no

Physical Address

00-20-AF-CA-E5-27

DHCP Enabled

no

IP Address

142.170.2.223

Subnet Mask

255.255.0.0

Default Gateway

142.170.2.1

Primary WINS Server

142.170.2.46

What is most likely the cause of the problem?

  1. Incorrect subnet mask

  2. NetBIOS scope ID is incorrect

  3. NetBIOS node type is incorrect

  4. IP address is out of range for this subnet

You have five Windows NT Server computers all configured as static routers. Which utility should you use to identify the path that a packet takes as it passes through the routers?

  1. Network Monitor

  2. ROUTE.EXE

  3. TRACERT.EXE

  4. IPCONFIG.EXE

  5. NETSTAT.EXE

You have an NT Server on a remote subnet. You cannot ping this server by its IP address, but you can ping your default gateway and all other computers on the remote subnet. What are two likely causes of the problem?

  1. The server is not WINS enabled.

  2. The server has an invalid subnet mask.

  3. The server has an invalid default gateway address.

  4. Your workstation is configured with an invalid subnet mask.

  5. Your workstation is configured with an invalid default gateway address.

You can connect through NT Explorer to all workstations on your local subnet, but cannot connect to any workstations or servers on a remote subnet. Which IP address should you ping first to diagnose the problem?

  1. The local server

  2. The default gateway

  3. The remote server

  4. None, TCP/IP isn't loaded

  5. None, reboot the router

Your Windows NT Workstation cannot connect to a remote Windows NT Server using NT Explorer. All other computers can connect to the remote NT Server. You run Network Monitor and notice that each time the workstation attempts to connect to the server, the workstation broadcasts an ARP request for the remote server's IP address. What is most likely the cause of the problem?

  1. The workstation is configured with an invalid default gateway address.

  2. The workstation is configured with an invalid subnet mask.

  3. The workstation is configured with a duplicate IP address.

  4. The workstation is not configured to use WINS.

You have a RAS server that is connected to an Internet Service Provider via ISDN. You want your Windows 95 workstations to use the RAS server to access the Internet. How should the default gateway addresses be configured?

  1. The default gateway address on the RAS server specifies the IP address of the ISP's router interface to your internal network.

  2. The default gateway address on the RAS server specifies the IP address of the ISP's router interface to the Internet.

  3. The default gateway address on each Windows 95 computer specifies the IP address of the ISP's router interface to your internal network.

  4. The default gateway address on each Windows 95 machine specifies the IP address of the RAS server's network interface to your internal network.

Your NT Server has three network adapters. You configure the server to route TCP/IP packets and you want it to be able to automatically update its routing tables by using routing information from other routers on the network. Which service must you install on the Windows NT Server computer?

  1. RIP for IP

  2. RIP for NWLink IPX/SPX Compatible Transport

  3. The DHCP Relay Agent

  4. The DHCP Service

You have five multihomed Windows NT Server computers running TCP/IP and routing TCP/IP packets. You want to configure the routing tables on these servers with a minimum of administrative effort. What should you do?

  1. Use NETSTAT.EXE to configure the routing tables.

  2. Use ROUTE.EXE to configure the routing tables.

  3. Install the DHCP Relay Agent.

  4. Install RIP for IP.

You have a multihomed Windows NT Server that you want to configure as a TCP/IP static router. What two steps must you complete?

  1. Enable IP forwarding.

  2. Configure each network adapter with a unique subnet mask.

  3. Configure each network adapter with an IP address, and ensure that each IP address is from a different subnet.

  4. Configure each network adapter with an IP address, and ensure that each IP address is from a different address class.

You have installed two routers on one subnet to provide redundancy. When one router fails, users start complaining that they cannot access remote subnets, even though you have a second router on the network. What should you do to prevent this problem from occurring the next time a router fails?

  1. Configure each workstation with multiple default gateway addresses.

  2. Configure each workstation with multiple IP addresses.

  3. Install a WINS server on each subnet.

  4. Install the DHCP Relay Agent on both routers.

The network is configured as follows: two subnets, one router. The router has two interfaces: Network A with 131.107.32.1 and Network B with 131.107.64.1. All computers use a subnet mask of 255.255.224.0. Your NT Workstation cannot connect to a remote server on Network A, but all other workstations on Network B can. You are located on Network B. When you run IPCONFIG on your workstation, you receive the following output:

  • IP Address

131.107.82.17

Subnet Mask

255.255.240.0

Default Gateway

31.108.32.1

What is most likely the cause of the problem?

  1. IP address on workstation is invalid

  2. Subnet mask on workstation is invalid

  3. Default gateway on workstation is invalid

  4. IP address on server is invalid

  5. Default gateway on server is invalid

The network is configured as follows: two subnets, one router. The router has two interfaces: Network A with 131.107.32.1 and Network B with 131.107.64.1. All computers use a subnet mask of 255.255.224.0. Your NT Workstation cannot connect to a remote server on Network A, but all other workstations on Network B can. You are located on Network B. When you run IPCONFIG on your workstation, you receive the following output:

  • IP Address

131.107.82.17

Subnet Mask

255.255.240.0

Default Gateway

131.108.64.1

What is most likely the cause of the problem?

  1. IP address on workstation is invalid

  2. Subnet mask on workstation is invalid

  3. Default gateway on workstation is invalid

  4. IP address on server is invalid

  5. Default gateway on server is invalid

Situation: You are installing an NT TCP/IP server with three network adapters. You also plan to use this server as a router.

Required result:

  • The new server must be configured to route TCP/IP.

Optional desired results:

  • The server must dynamically update its routing tables.

  • The server must provide IP addresses to all clients located on all subnets.

  • The server must be able to send trap messages across the network to a Windows NT workstation computer.

Proposed solution:

  • Install TCP/IP and configure one IP address for each of the servers network adapters.

  • Install PPTP on the server.

  • Install DHCP on the server and configure one scope for each subnet.

  • Install SNMP on the server and configure SNMP to forward trap messages to the workstation.

Which results does the proposed solution produce?

  1. The proposed solution produces the required result and produces all of the optional desired results.

  2. The proposed solution produces the required result and produces only two of the optional desired results.

  3. The proposed solution produces the required result but does not produce any of the optional desired results.

  4. The proposed solution does not produce the required result.

Situation: You are installing an NT TCP/IP server with three network adapters. You also plan to use this server as a router.

Required result:

  • The new server must be configured to route TCP/IP.

Optional desired results:

  • The server must dynamically update its routing tables.

  • The server must provide IP addresses to all clients located on all subnets.

  • The server must be able to send trap messages across the network to a Windows NT workstation computer.

Proposed solution:

  • Install TCP/IP and configure one IP address for each server's network adapters.

  • Enable IP forwarding on the server.

  • Install PPTP on the server.

  • Install the DHCP Relay Agent on the server.

Which results does the proposed solution produce?

  1. The proposed solution produces the required result and produces all of the optional desired results.

  2. The proposed solution produces the required result and produces only two of the optional desired results.

  3. The proposed solution produces the required result but does not produce any of the optional desired results.

  4. The proposed solution does not produce the required result.

Situation: You are installing an NT TCP/IP server with three network adapters. You also plan to use this server as a router.

Required result:

  • The new server must be configured to route TCP/IP.

Optional desired results:

  • The server must dynamically update its routing tables.

  • The server must provide IP addresses to all clients located on all subnets.

  • The server must be able to send trap messages across the network to a Windows NT workstation computer.

Proposed solution:

  • Install TCP/IP and configure one IP address for each of the servers network adapters.

  • Enable IP forwarding on the server.

  • Install RIP for IP on the server.

  • Install DHCP with scopes for all subnets.

Which results does the proposed solution produce?

  1. The proposed solution produces the required result and produces all of the optional desired results.

  2. The proposed solution produces the required result and produces only two of the optional desired results.

  3. The proposed solution produces the required result but does not produce any of the optional desired results.

  4. The proposed solution does not produce the required result.

Situation: You are installing an NT TCP/IP server with three network adapters. You also plan to use this server as a router.

Required result:

  • The new server must be configured to route TCP/IP.

Optional desired results:

  • The server must dynamically update its routing tables.

  • The server must provide IP addresses to all clients located on all subnets.

  • The server must be able to send trap messages across the network to a Windows NT workstation computer.

Proposed solution:

  • Install TCP/IP and configure one IP address for the server's network adapters.

  • Enable IP forwarding on the server.

  • Install DHCP with scopes for all subnets.

  • Install SNMP on the server and configure SNMP to forward trap messages to the workstation.

  • Install a third-party SNMP manager on the server.

Which results does the proposed solution produce?

  1. The proposed solution produces the required result and produces all of the optional desired results.

  2. The proposed solution produces the required result and produces only two of the optional desired results.

  3. The proposed solution produces the required result but does not produce any of the optional desired results.

  4. The proposed solution does not produce the required result.

Your boss frantically comes up to you and says he put two NIC cards in his NT Server, but the workstations on each segment can't see each other. He knows that to route IP packets to other networks each multihomed computer (static router) must be configured two ways, but he can't remember what they are. What are the two things you need to set on the NT Server?

You get a call from a company that thinks they need some routers. They have a small network and read in a magazine that they should employ static routing. They want to know more about how static routing would meet their networking needs. What do you tell them?

It's your first day on the job at Terrific Technology Teaching Center, and as a co-instructor you are asked to teach the enabling of IP routers. Take a moment to explain the procedure now.

Later, a confused student comes up to you and asks if she needs to add a routing table to a computer running as a multihomed computer and connecting two subnet segments. What do you tell her, and why?

You have two NT servers and a router to the Internet. Should you build a static router between the NT servers, or will the dynamic router to the Internet be sufficient?

What information would you put into the static routing table?

Using supernetting, assign the missing IP and subnet mask values for each customer below:

  • Beginning IP address

192.24.0.1

Ending IP address

192.24.7.8

Subnet Mask

 

Beginning IP address

 

Ending IP address

192.34.31.254

Subnet Mask

255.255.240.0

Beginning IP address

192.24.8.1

Ending IP address

 

Subnet Mask

255.255.252.0

Beginning IP address

192.24.14.1

Ending IP address

192.24.15.254

Subnet Mask

 

About the Authors

Todd Lammle is a Microsoft Certified Trainer (MCT) with over fifteen years of experience with LANs and WANs. He is president of GlobalNet Systems, a network integration firm in Colorado.

Monica Lammle is a Microsoft Certified Product Specialist (MCPS) in TCP/IP.

James Chellis, a Microsoft Certified Professional (MCP), is president of EdgeTek Computer Education, a national network training company and Microsoft Solution Provider.

Copyright © 1997 by Sybex, Inc.

We at Microsoft Corporation hope that the information in this work is valuable to you. Your use of the information contained in this work, however, is at your sole risk. All information in this work is provided "as -is", without any warranty, whether express or implied, of its accuracy, completeness, fitness for a particular purpose, title or non-infringement, and none of the third-party products or information mentioned in the work are authored, recommended, supported or guaranteed by Microsoft Corporation. Microsoft Corporation shall not be liable for any damages you may sustain by using this information, whether direct, indirect, special, incidental or consequential, even if it has been advised of the possibility of such damages. All prices for products mentioned in this document are subject to change without notice. International rights = English only.

International rights = English only.

Link
Click to order