2.2 RIP Overview

CCNP RS 300-101, Implementing RIPng, Uncategorized

RIPv1 was released in the 1980s, and featured classful addressing, routing updates by broadcast, and lack of security. These shortcomings were addressed by RIPv2, which provided classless addressing, multicast updates at 224.0.0.9, and routing protocol authentication.

RIPng was created for IPv6 and has much in common with RIPv2. It is, however, not backwards compatible with either of its predecessors. In a dual-stack configuration, RIPng can run alongside RIPv1 or RIPv2.

Here are the characteristics of RIPv1, RIPv2, and RIPng:

  • Easy to configure
  • Standardized IGP that enables mixed-vendor support
  • The hop count (sole metric) is maxed at 15 hops (16 is infinite/unreachable)
  • RIPv1 is classful; RIPv2 and RIPng are classless and support VLSM
  • RIPv1 and RIPv2 use UDP 520
  • RIPng uses UDP 521
  • RIPv2 multicast address = 224.0.0.9 (RIPv1 broadcast)
  • RIPng multicast address = FF02::9
  • RIP is a distance vector protocol and will send its entire routing table to neighbors every 30 seconds (or with a topology change)
  • RIPv2 and RIPng support authentication
  • RIP has a default administrative distance of 120, the worst of all IGPs

There are two fundamental flaws that ensure RIP is not widely used.

  1. The maximum 15 hops does not allow it to scale to larger networks
  2. The use of hop count as the sole metric allows the protocol to make erroneous assumptions about the underlying network infrastructure, for example, when a router prefers a 1-hop 128Kbps link versus a 2-hop 1Gbps link.
Published