STUDENT NUMBER: 500687685
COURSE TITLE: CN8810
SUBMISSION DATE: October 14, 2015
QUESTION
What are Routing Loops?
The following techniques are used to prevent routing loops:
1. Split Horizon
2. Route poisoning
3. Hold-down timer
Explain two of the above techniques briefly with diagrams.
SOLUTION
ROUTING LOOPS
Routing loops occur as a result of inconsistent routing table. When using distance-vector routing, routers can learn “bad” routes and send them to other routers thereby forming a loop. In simplest terms it is a packet that never reaches its destination and just keeps traversing between the same routers in an endless circle.
SPLIT HORIZON
It is used along with other techniques …show more content…
In other words, it means a router should not send routing updates out through the interface it was received or learnt. Split Horizon has another variation called Split Horizon with poison reverse which allows a node advertise routes as unreachable over the interface which they were learned by setting the route metric or minimum cost to infinite.
The diagram below with example explains Split Horizon.
NETWORK A NETWORK B S0/2 S0/2 S1/0 S1/0 NETWORK C
FIGURE 1
Split Horizon rule for R1 and R3 in FIGURE 1
R1 should not advertise the route to Network B and Network C back to R2. if there is no split Horizon and the link between R2 and R3 goes down. R1 will keep informing R2 about an alternative route to Network B (thru R2) and without sufficient Artificial intelligence on R2, a loop can be created.
R3 should strike out the route to Network A when sending routing updates to R2. If there is no split Horizon and the link between R1 and R2 goes down. R3 will continue to inform R2 it can get to Network A (thru R2) and without sufficient Artificial intelligence on R2, it would use R3’s route as the alternative route to network A, thereby causing Routing …show more content…
If R3 sends network c with a metric of 1 to R2, then its interface s1/0 goes down and the invalid timer for R2 expires, then during hold down time on R2, S1/0 on R3 comes on again and sends network c with a metric of 1 again, R2 will discard it during its hold-down.
Even if a triggered message is sent from R3 to R2 regarding network c, the route would be discarded immediately, with a poison reverse and the bad news is passed to R1
If R3 stops sending updates, and R2 hits the invalid timer for that route, it will begin the hold-down, and send a poisoned route to R1. Because R1 got the poisoned route from R2, R1 will simply delete the route, and not go into hold down.
NOTE: Invalid timer is how many seconds, since seeing a valid update, to consider the route invalid, and placing the route into hold down
NOTE: triggered update is sent when a change occurs outside the update interval.