MTR (My Traceroute)

MTR (My Traceroute) combines the functionality of two classic network diagnostic tools, ping and traceroute, into a single continuous diagnostic. Where a traditional traceroute prints one snapshot of the path to a destination, MTR repeatedly probes every hop along the route and aggregates the results, building a live statistical picture of each router between you and the target. The output shows each hop's hostname or IP address alongside per-hop metrics like packet loss percentage, the number of packets sent, and latency figures (best, average, worst, and standard deviation).

The real value of MTR over a one-off traceroute is that intermittent problems become visible. A single traceroute might look perfectly healthy, but a connection that drops 5 percent of packets at a specific router will only reveal itself over many probes. By watching the loss column across dozens of cycles, you can pinpoint exactly which hop is introducing latency or dropping traffic, distinguishing a flaky intermediate ISP router from a problem at the destination itself. High standard deviation in the latency column signals jitter, which matters for real-time applications like VoIP and gaming.

Common use cases include diagnosing slow or unreliable connections to a specific website or game server, providing evidence to an ISP or hosting provider when reporting a routing problem, and verifying that a network path change (such as a new BGP route) actually improved performance. Network engineers also use MTR to confirm whether packet loss is symmetric or only appears on the return path. Because it identifies the responsible network segment, MTR output is often the first thing support teams request.

When reading MTR results, remember that loss appearing at an intermediate hop but not at later hops usually means that router is simply rate-limiting ICMP responses to itself, not actually dropping your traffic. Loss is only meaningful if it persists through to the final destination. Pair MTR with a basic Ping test to confirm baseline reachability and a Traceroute for a clean single-pass view of the route, and use a DNS Lookup if hop hostnames are not resolving.