What is a Routing Loop?
Learn what a routing loop is, why distance-vector protocols cause it, and how split horizon and TTL prevent it — with interview Q&A.
Expected Interview Answer
A routing loop is a condition where two or more routers each believe the best path to a destination is through the other, so a packet keeps bouncing between them (or around a cycle of routers) instead of ever reaching its destination, typically until its TTL expires and it is dropped.
Routing loops usually appear after a link or route fails and the routers involved have inconsistent, stale views of the topology — a distance-vector router may keep advertising a route it learned from a neighbor back to that same neighbor, so both sides think the other still has a valid path. Symptoms include packets circulating indefinitely, rapidly incrementing hop counts, and TTL-exceeded ICMP messages flooding back to senders. Distance-vector protocols like RIP are especially prone to this because they lack full topology awareness; link-state protocols like OSPF avoid it because every router computes shortest paths from the same complete topology graph. Loop-prevention techniques include split horizon, route poisoning, hold-down timers, and TTL as a last-resort safety net.
- Explains why packets never reach their destination during the loop
- Shows how stale distance-vector state causes mutual mis-belief
- Distinguishes distance-vector (loop-prone) from link-state (loop-resistant)
- Points to concrete prevention mechanisms interviewers expect
AI Mentor Explanation
A routing loop is like two fielders each thinking the other has called for a catch, so both run toward the ball and then away from it in a confused circle while it never actually gets caught. Each fielder’s belief about who is closer to the ball is based on stale information shouted a second ago, not the current position. The ball (the packet) keeps moving between them until the umpire calls a dead ball (TTL expiry) and stops play. Clear, single-source calling — the equivalent of consistent routing state — is what prevents this chaos.
Step-by-Step Explanation
Step 1
Topology change
A link or route fails, and routers begin propagating updated reachability information.
Step 2
Stale advertisement
A distance-vector router advertises a route back to the neighbor it originally learned it from, unaware the neighbor lost it.
Step 3
Mutual mis-belief
Both routers now believe the other has a valid path, so packets are forwarded back and forth between them.
Step 4
TTL expiry
Each looped packet’s TTL decrements on every hop until it hits zero and is discarded, generating ICMP Time Exceeded messages.
What Interviewer Expects
- Clear definition: packets circulate instead of reaching the destination
- Explains why distance-vector protocols are more loop-prone than link-state
- Names concrete prevention techniques (split horizon, route poisoning, hold-down timers)
- Mentions TTL as the last-resort safety net that bounds loop damage
Common Mistakes
- Describing it only as “slow routing” instead of packets actually cycling
- Not distinguishing distance-vector loop-proneness from link-state resistance
- Forgetting TTL is what ultimately terminates a looped packet
- Confusing a routing loop with a broadcast storm at Layer 2
Best Answer (HR Friendly)
“A routing loop happens when routers get confused about who is closer to a destination and end up sending a packet back and forth between each other instead of forwarding it onward. It usually happens right after something in the network changes and the routers have not all caught up on the new information yet. Eventually the packet gets thrown away, but until then it wastes bandwidth and can slow down or break connectivity, which is why protocols include specific safeguards to prevent it.”
Code Example
# A healthy traceroute shows increasing, distinct hops toward the destination
traceroute 198.51.100.20
# A routing loop shows the same set of hops repeating with rising latency
# 1 10.0.0.1 1.2 ms
# 2 10.0.1.1 2.1 ms
# 3 10.0.0.1 3.4 ms <- same router as hop 1, looping
# 4 10.0.1.1 4.6 ms
# ...
# 30 * * * <- TTL exceeded, trace gives up
# Check a router’s routing table for suspicious mutual next-hops
ip route showFollow-up Questions
- How does split horizon prevent a routing loop?
- Why is OSPF less prone to routing loops than RIP?
- What is route poisoning and how does it complement split horizon?
- What role does TTL play in bounding the damage of a routing loop?
MCQ Practice
1. What ultimately stops a packet stuck in a routing loop?
The TTL field decrements at each hop; once it hits zero the packet is dropped and an ICMP Time Exceeded is sent.
2. Which class of routing protocol is most prone to routing loops?
Distance-vector protocols like RIP rely on neighbor-advertised distances rather than full topology awareness, making loops more likely.
3. What does split horizon specifically prevent?
Split horizon stops a router from re-advertising a route to the same neighbor it learned it from, a primary cause of two-node loops.
Flash Cards
What is a routing loop? — A condition where packets circulate among routers instead of reaching their destination, due to inconsistent routing state.
Why are distance-vector protocols loop-prone? — They only know distances reported by neighbors, not the full topology, so stale advertisements can create mutual mis-belief.
What finally stops a looped packet? — TTL decrements to zero and the packet is dropped, triggering an ICMP Time Exceeded message.
Name two loop-prevention techniques. — Split horizon and route poisoning (also hold-down timers).