Also see raster image processor (RIP).
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
The Routing Information Protocol (RIP) defines a way for routers, which connect networks using the Internet Protocol (IP), to share information about how to route traffic among networks. RIP is classified by the Internet Engineering Task Force (IETF) as an Interior Gateway Protocol (IGP), one of several protocols for routers moving traffic around within a larger autonomous system network -- e.g., a single enterprise's network that may be comprised of many separate local area networks (LANs) linked through routers.
Although once the most widely used IGP, Open Shortest Path First (OSPF) routing has largely replaced RIP in corporate networks.
Each RIP router maintains a routing table, which is a list of all the destinations (networks) it knows how to reach, along with the distance to that destination. RIP uses a distance vector algorithm to decide which path to put a packet on to get to its destination. It stores in its routing table the distance for each network it knows how to reach, along with the address of the "next hop" router -- another router that is on one of the same networks -- through which a packet has to travel to get to that destination. If it receives an update on a route, and the new path is shorter, it will update its table entry with the length and next-hop address of the shorter path; if the new path is longer, it will wait through a "hold-down" period to see if later updates reflect the higher value as well, and only update the table entry if the new, longer path is stable.
Using RIP, each router sends its entire routing table to its closest neighbors every 30 seconds. (The neighbors are the other routers to which this router is connected directly -- that is, the other routers on the same network segments this router is on.) The neighbors in turn will pass the information on to their nearest neighbors, and so on, until all RIP hosts within the network have the same knowledge of routing paths, a state known as convergence.
If a router crashes or a network connection is severed, the network discovers this because that router stops sending updates to its neighbors, or stops sending and receiving updates along the severed connection. If a given route in the routing table isn't updated across six successive update cycles (that is, for 180 seconds) a RIP router will drop that route, letting the rest of the network know via its own updates about the problem and begin the process of reconverging on a new network topology.
RIP uses a modified hop count as a way to determine network distance. (Modified reflects the fact that network engineers can assign paths a higher cost.) By default, if a router's neighbor owns a destination network (i.e., if it can deliver packets directly to the destination network without using any other routers), that route has one hop, described as a cost of 1. RIP allows only 15 hops in a path. If a packet can't reach a destination in 15 hops, the destination is considered unreachable. Paths can be assigned a higher cost (as if they involved extra hops) if the enterprise wants to limit or discourage their use. For example, a satellite backup link might be assigned a cost of 10 to force traffic follow other routes when available.
RIP has been supplanted mainly due to its simplicity and its inability to scale to very large and complex networks. Other routing protocols push less information of their own onto the network, while RIP pushes its whole routing table every 30 seconds. As a result, other protocols can converge more quickly, use more sophisticated routing algorithms, include latency, packet loss, actual monetary cost and other link characteristics, as well as hop count with arbitrary weighting.
Continue Reading About Routing Information Protocol (RIP)
Margaret Rouse asks:
Why use a self-learning protocol like RIP instead of having a single “master router” that knows all the routes and pushes them out to all participating routers?
1 ResponseJoin the Discussion