RIP Version 2 Protocol Applicability Statement
also
- std-57
Extracted elements (13)
Authentication was added to RIP-2 to prevent insertion of bogus routing information by parties who cannot directly observe the network. Confidentiality of routing messages is considered unimportant; integrity is the goal. The authentication field allows for extensible authentication types, though only 16 octets are available due to the original RIP packet format.
RIP-2 is an extension to RIP-1, not a new protocol. Its applicable environments are a superset of RIP-1 use cases, meaning operational experience with distance-vector routing within an autonomous system carries over directly.
RIP-2 is not intended as a substitute for OSPF in large autonomous systems. The same restrictions on AS diameter and complexity that applied to RIP-1 still apply to RIP-2; RIP-2 fills the gap for smaller, simpler environments that need authentication or VLSM but do not require a link-state protocol.
RIP-2 uses IP multicast for periodic route advertisements to reduce load on systems that do not support RIP-2. This also ensures that RIP-1 routers never receive RIP-2 routes, which is desirable when correct route use depends on knowing the precise subnet mask that RIP-1 would ignore.
Subnet masks were added to RIP-2 because VLSM, CIDR, and supernetting made it impossible for heuristics to reliably distinguish network, subnet, and host routes. Including the 32-bit mask immediately after the IP address in each routing entry effectively creates a 64-bit address that eliminates the ambiguity.
The Next Hop field was added to eliminate packets traversing unnecessary extra hops, particularly in mixed environments where RIP-2 is not running on every router. A RIP-2 speaker that knows of a better next hop for a route can advertise that gateway directly, allowing routers to bypass intermediate speakers.
The Next Hop mechanism is explicitly non-recursive and only eliminates a single extra hop. Extensions to handle larger optimizations were considered but not specified because the potential for routing loops had not been sufficiently analyzed.
By advertising RIP-2 routes via IP multicast rather than broadcast, RIP-2 speakers guarantee that RIP-1 routers are never exposed to routes whose correct interpretation requires knowing the subnet mask. This prevents RIP-1 nodes from misrouting traffic based on incomplete information.
When a Next Hop value specifies a router not directly connected from the receiver's perspective, the receiving router ignores the specified next hop and falls back to using the advertising router as the next hop, preserving correct (if suboptimal) routing behavior.
Each RIP-2 routing entry includes a 32-bit subnet mask field immediately following the IP address. This field unambiguously identifies the prefix length of every advertised route, enabling VLSM and CIDR support.
The authentication field in a RIP-2 packet provides 16 octets for authentication data. The format is extensible to support multiple authentication types, though the space constraint limits strength; the initial type is a simple cleartext password.
The Next Hop field in a RIP-2 route entry allows the advertising router to specify a better next-hop gateway for that route. A value of 0.0.0.0 indicates the advertising router itself should be used as the next hop.
The authentication mechanism in RIP-2 prevents remote parties from injecting bogus routing information, but only as long as they cannot sniff the network to obtain the password. The 16-octet constraint on the authentication field limits the strength of authentication that can be carried; stronger authentication types are allowed by the specification but constrained by this space.