Traceroute / MTR

Run MTR from real public probes and inspect hop addresses, ASNs, average latency, jitter and loss.

No measurement yetSubmit a public target to see the probes and measurement evidence that actually return.

How to read an MTR trace

MTR combines route tracing with repeated probes to show where latency and packet loss appear. Enter a public IP or hostname, then select a probe location to inspect each hop.

Two examples, explainedIllustrative · not live results

Loss at one intermediate hopICMP

Example observation: Hop 5: 20% lossLater hops and destination: 0%
How to interpret it: Probe replies may be rate-limitedThis hop alone does not prove a fault

Loss continues to the destinationICMP

Example observation: Loss begins at hop 5: 10%Loss also appears at later hops
How to interpret it: Investigate the path furtherCompare locations and test times

Probe perspectiveResults start at the selected probe, not your device; return paths may differ.

Common questions

Does an asterisk mean the route is broken?

No. A router may ignore probes while still forwarding traffic. Check whether later hops and the destination respond.

Should I use Ping or MTR first?

Start with Ping for overall latency and loss, then use MTR for path detail. Use TCP or HTTP tests for application access issues.

How should I compare average and worst RTT?

Average RTT summarizes the sample; worst RTT is its peak. One spike does not establish long-term quality. Review jitter, loss and sample size too.

Does the first slow hop identify the fault?

Not necessarily. Routers prioritize replies differently, and replies may take different routes. An increase that persists to the destination deserves closer investigation.

Why can the route to the same target change?

Routing policy, load balancing and maintenance can change the path. Keep the probe, destination address and protocol consistent when comparing tests.

Does a clean trace mean the website works?

No. Route probes do not validate TLS, login or application responses. Follow up with TCP and HTTP/TLS checks when a website still fails.