Traceroute — Every Hop Between You and a Host
Traceroute reveals the sequence of routers a packet passes through to reach a destination, with timing for each. It works by sending probes with deliberately short lifetimes so that each router in turn is forced to report itself. The result is a map of the path and a rough sense of where time is being spent along it.
Why run it
When something is slow or unreachable and the endpoints look healthy, the problem is in between, and traceroute is how you find out where. It answers whether traffic is even taking the route you expect — a surprising amount of unexplained latency turns out to be traffic crossing a continent unnecessarily because of a routing change nobody announced. Read it with care, though: the intermediate timings are notoriously misleading, and most people misdiagnose from them.
Why automate it
Routes change, and they change without any signal to you. A path that was direct last month may now traverse three extra networks because a peering arrangement lapsed. Capturing the path on a schedule gives you a history, and the history is what makes a change obvious — you can point at the day the route changed and correlate it with the day latency rose. Trying to reconstruct that after the fact from a single traceroute run during an incident is guesswork, because you have nothing to compare to.
What it gives you
The location of a problem rather than just its existence. Recorded over time it also becomes evidence, which is what you need when the fix is someone else's to make and they need convincing it is theirs.