Routing Loops

A routing loop is a fault in which packets circle endlessly between routers because their routing tables disagree.

What is Routing Loops?

Distance-vector protocols such as RIP guard against loops with several techniques: a maximum hop count (16 in RIP means unreachable), split horizon (not advertising a route back to where it was learned), route poisoning (advertising a failed route as unreachable), and hold-down timers (ignoring updates for a while after a route fails). The IP time-to-live field stops looping packets lasting forever.

← Back to the A–Z definitions