Managing TCP/IP RoutingWith Routing and Remote Access (RRAS), a computer running Windows Server 2003 can function as a network router, which routes IP packets between networks. This router service allows LANs and WANs to be interconnected easily. The routing technology is built into the operating system, providing small and large businesses with a cost-effective and secure way of interconnecting their networks. IP RoutingRouting is the process of sending a packet from the source address to the destination address. Because all IP packets have a source and destination IP address, it is possible to deliver them to the proper location. Of course, how the routing of IP packets actually occurs is much more complex. Because each IP packet has addressing information within the header, routers can use this information to determine where a packet should be sent so that it can reach the destination host. Routers maintain information about the physical network, such as the path to a destination network and the metric associated with the route. The metric is the cost associated with using a route. Typically, this is the number of hops to the destination network. If there are multiple routes to a destination network, the route with the lowest metric is favored. For routers to know the location to which packets must be forwarded, they must also know about their neighboring routers. This information is stored within the routing table. When a router receives a packet, it checks the routing table to determine where the packet must be sent to reach the destination host. The information within a routing table can be generated statically or dynamically. Static RoutingWith static routing , an administrator must manually configure the routing table by adding entries that tell the router how to reach other networks. Using the route command, an administrator updates the routing table by specifying the network addresses, the subnet masks, and the metric associated with each route. When deciding whether to use static routing, keep in mind that it works best for networks that do not change on a regular basis. If the network configuration is constantly changing, the administrative overhead associated with having to update the routing tables every time something changes increases greatly because the changes must be made on each router. In such cases, it might be more beneficial to implement dynamic routing. Dynamic RoutingDynamic routing eliminates the overhead associated with manually updating routing tables. Routers can dynamically build their own routing tables by communicating with other routers on the network. With dynamic routing, the routing tables are built automatically through router communication. Using a routing communication protocol, routers periodically exchange messages containing location information about routes through the network. This information is used to build and update routing tables. The major advantage of dynamic routing is that it reduces the administrative overhead associated with manually updating routing tables. For example, if a router goes down, the change is automatically propagated to all routers on the network so that they are all aware of the change in the network topology. One of the major disadvantages is the high amount of traffic it generates. Managing Routing ProtocolsFor routers to share information and dynamically update their routing tables, a routing protocol must be used. As already mentioned, the two routing protocols supported by Windows Server 2003 are Routing Information Protocol (RIP ) and Open Shortest Path First (OSPF). Although both routing protocols are used for dynamic routing, there are some distinct differences between the two. Routing Information Protocol (RIP)The RIP is designed for small to medium-size networks. One of the main benefits in choosing RIP is that it's easy to configure and deploy. One of the major drawbacks associated with this protocol is that it is limited to a maximum hop count of 15. This means that any networks more than 15 hop counts away are considered unreachable. Also, as a network increases in size, excessive traffic can be generated from RIP announcements.
When a router is first configured as a RIP router, the only entries in the routing table are for those networks to which it is physically connected. It then begins to send announcements of its availability to notify other routers of the networks it services. RIPv1 sends the announcements as broadcasts, whereas RIPv2 can broadcast multicast packets for the announcements. When changes occur to the network topology, RIPv2 uses triggered updates to communicate the changes to other routers. With triggered updates, the change to the network topology can be propagated immediately.
Open Shortest Path First (OSPF)OSPF is designed for large internetworks ( especially those spanning more than 15 router hops). The disadvantage of OSPF is that it's generally more complex to set up and requires a certain amount of planning.
OSPF uses the Shortest Path First (SPF) algorithm to calculate routes. The shortest path (the route with the lowest cost) is always used first. Unlike RIP, which uses announcements only to update and share routing information, OSPF maintains a map of the network known as the link-state database . This map is synchronized between adjacent routers, or those neighboring OSPF routers. When a change is made to the network topology, the first router to receive the change sends out a change notification. Each router then updates its copy of the link state database, and the routing table is recalculated.
One of the main differences between OSPF and RIP is that OSPF divides the network into different areas. Each of the routers maintains information in the link-state database only about those areas to which it is connected. Another difference is that OSPF replicates the changes only to the routing table, not to the entire table.
Installing and configuring RIP and OSPF is discussed in the section entitled "Managing Remote Access." Manage Routing TablesIn some instances you need to add a static route to your Windows Server 2003 router. This, of course, has its advantages and disadvantages. Creating a static route is simple; however, the routes you configure are not shared between routers. Static routes specify the network address and subnet mask that tell the router how to reach a certain destination. The router uses the information to determine which gateway to forward the packet to so that the packet can reach the destination host. Static routes can be configured in one of two ways. First, they can be configured using the route add command. Using the route add command, you can add static entries to the local routing table on a router. (You can also use the p parameter to specify whether they should be persistent routes, meaning that they will remain in the routing table when the system reboots.) The syntax for the route add command is as follows : route add <destination> mask <netmask> <gateway> metric <interface> The second option is to configure a static route within the RRAS management console. To configure a static route using this method, perform the following steps:
|