How this check works
The check validates the target, refuses private and reserved addresses, and asks the Globalping network to run a traceroute from the probe locations you choose. Each probe sends packets with an increasing TTL (hop limit). Every router that discards a packet because its TTL ran out sends back an ICMP "time exceeded" message, which reveals that hop's address and the round-trip time to it. The trace ends when the target answers or the hop limit is reached.
Results show each hop with its address, reverse DNS name where one exists, the network (ASN) it belongs to, and the round-trip times. Globalping measurement data, including the target, is public.
How to read the results
- Rows of
*mean that hop didn't answer. Many routers don't send "time exceeded" messages, or rate-limit them. If later hops answer, the path is fine. - A jump in latency that stays high for every hop after it points to the link where delay starts, such as an ocean crossing or a congested peering point.
- A jump on one hop only, with lower times after it, is usually the router answering slowly because it treats traceroute replies as low priority. It isn't slowing traffic that passes through it.
- The trace stops before the target when a firewall drops the probe packets or the host doesn't reply. That doesn't prove the target is down: compare with the Ping Check.
- Different paths from different locations are normal. Each probe's provider chooses its own route to the target.
Common questions
Why doesn't the last hop match the target? Some networks answer from a load balancer or a firewall in front of the host.
Is a traceroute from here the same as one from my office? No. Routing depends on where you start. Run tracert (Windows) or traceroute (macOS, Linux) from the affected network to see that path, and use this check to compare it with paths from elsewhere.
Why not use mtr? mtr repeats a traceroute many times and is better for intermittent loss. The ping and ICMP guide explains when to use each.
Updated 26 October 2026