Poison Reverse vs Split Horizon
Split Horizon and Reverse Poisoning both are techniques also called loop prevention mechanisms used in the network layer. In this article, we will discuss what split horizon, poison reverse, and the main difference between poison reverse and split horizon.
Distance vector protocols employ the split horizon technique to avoid network routing loops. The fundamental rule is straightforward: Never transmit routing information back in the direction it came from.
Consider 3 routes, A—–B—–C which are linked by point-to-point links.
Node A does not advertise its path for C (i.e., A to B to C) back to B in accordance with the split-horizon rule. This initially appears pointless because B will never route via node A since the route is more expensive than the direct route from B to C. If the connection between B and C fails, on the other hand, and B had been given a path from A to C, B might end up taking that route via A.
An implemented algorithm called Poison Reverse is frequently used in distance-vector routing. To solve the count-to-infinity problem, employ poison reverse. Practically speaking, poison reverse is the opposite of split horizon. The main goal of poison reverse is to prevent paths from reversing into the same node when a network cost changes.
Consider 3 routes,
The cost of going from A—-B is 1, from B—-C is 2 and from C—-A is 30.
Node A travels to destination C by way of node B. The count-to-infinity problem will arise as the cost between Y and C rises. We use the poison reverse technique to prevent it. A will tell B a white lie by declaring to B that traveling to destination C will cost an unlimited amount of money as long as A uses node B to reach C.
Split horizon is sometimes used with a move known as poison reverse. This is the equivalent of telling all routers that a certain packet’s path back to the originating node has an infinite metric, or route poisoning all potential reverse pathways.
The difference between Poison Reverse and Split Horizon:
|Split Horizon||Poison Reverse|
|It states that a route cannot be promoted out of an interface if the next hop for the advertised route is located on that interface.||It states that routes that are received via one interface must be broadcast back out from that interface with an unreachable metric.|
|By not broadcasting the route out of the incorrect interface, Split Horizon passively attempts to prevent the routing loop.||In order to access the networks, Poison Reverse aggressively inhibits neighboring routers from using it.|
|Routers will utilize a hop count of 16 instead of immediately removing faulty path information from the routing table.||The device does not transfer updates through the same interface out when it receives an update from an interface.|
|It simply states that you shouldn’t advertise a path out of an interface where you first learned the way.||It states that it will advertise that route to you, but with an infinite metric if you’ve told me this route is not reachable through you.|
|It is used alone in loop prevention.||It is generally used with Split Horizon for loop prevention.|
In networks with many routing pathways, split horizon with poison reverse outperforms ordinary split horizon, even if it generates more network traffic. However, in networks with just one routing line, a split horizon with poison reverse offers no advantage over a plain split horizon.