Using Routing Information Protocol
Static routing table configuration proves efficient in case of limited routers and network configuration that is not subject to frequent changes. For networks that are large (with 15 or more routers) or networks where configuration is susceptible to changes 9example dynamic corporate networks) it is recommended that a routing protocol be used.
The Microsoft software-based routing solution is not best suited for networks that are large and complex in nature. Windows Server 2008 supports Routing Information Protocol (RIP). RIP is designed in a manner that allows exchange of routing related information. The size of the network should ideally be small or medium.
Configuration and deployment of RIP is a simple process. A major drawback with RIP is its unscalability to large networks. The maximum hop count that RIP routers can support is 15. In the case of large networks the announcements generated by the RIP counter seem like excessive traffic.
When compared to routing protocols that are considered sophisticated, RIP has a higher recovery time. As the topology of the network changes, it may take time before the RIP routers reconfigure. As the network reconfigures, routing loops may be formed resulting data getting lost or being treated as undeliverable. Inspite of all this, RIP still results in a lesser amount of delay and traffic that has been lost than in the case of manual reconfiguration.
To begin with, the route table for each router lists only those networks that are connected physically. Frequent announcements containing the route table entries are sent to other RIP routers informing them about the networks they can reach. IP broadcast packets are used by RIPv1 for announcements. Mulitcast or broadcast packets are used by RIPv2for announcements. The figure given below illustrates the process of route announcements.
Triggered updates are also used by RIP routers for communicating. Triggered updates occur as the topology of the network changes. Updated information is sent reflecting the changes. Triggered updates are transmitted immediately without waiting for any subsequent announcement. A router on receiving a triggered update changes its route table and incorporates the changes.
RIPv1 and RIPv2 are supported by Windows Server 2008. Multicast announcements, password authentication of a simple nature, greater flexibility in subnetted and classless interdomain routing (CIDR) are supported by RIPv2. It is also serves as the default routing protocol for Windows Server 2008.
RIP implementation in Windows Server 2008 environment has the following characteristics:
- The RIP version to be run on each interface for incoming and outgoing packets can be selected;
- Algorithms - triggered update, split-horizon and poison-reverse are used for avoiding routing loops. These are also used for speeding up recovery of the network when changes take place in topology.
- Route filters can be used for configuring networks for ignoring or accepting of announcements. This can be done on the Security tab of the RIP Properties dialog box.
- Peer filters can be used for choosing which router's announcements are to be accepted.
- Router announcements can be configured and aging timers can also be set.
- It is possible to support simple password authentication.
- Subnet summarization can be disabled.