Before the link failure, the converged distances to destination R are:
Distance from Q to R=1 (via direct link to R)
Distance from P to R=2 (via Q)
When link Q–R fails:
Q detects the failure and sets its direct cost to R to ∞.
Since P and Q send out routing updates independently at random times with the same average rate, the probability that P sends an update to Q first is 0.5, and the probability that Q sends an update to P first is 0.5.
Case 1: P sends update to Q before receiving an update from Q (Probability = 0.5):
P advertises its distance to R as 2. Router Q receives this and updates its distance to R via P as 1+2=3. This creates a routing loop between P and Q, triggering the count-to-infinity problem.
Case 2: Q sends update to P first (Probability = 0.5):
Q advertises distance ∞ to R. P updates its distance to R as ∞, avoiding a routing loop.
Therefore, the probability of a routing loop formation is 0.5.