Free Traceroute Online Tool

What Is Traceroute Online?

Traceroute online is a network diagnostic tool that lists every router between a probe and a destination host, with the round-trip time to each one. It sends packets with an increasing IP time to live value and records which router returns the ICMP Time Exceeded message defined in RFC 792, revealing the path one hop at a time.

Every IP packet carries a TTL field. RFC 791 requires each router to decrement it and to destroy the packet when it reaches zero, and RFC 792 defines the Time Exceeded reply as type 11, code 0 ("time to live exceeded in transit"). A traceroute sends a probe with TTL 1, then 2, then 3, so each router in turn discards one probe and identifies itself, until the destination answers and the trace stops.

The TestMu AI traceroute runs from probes in up to 16 regions at the same time, five of them (New York, London, Frankfurt, Singapore, and Sydney) selected by default, plus any custom city, country, or ASN, all operated by the Globalping API that also powers our Ping IPv6 and Port Checker tools. All parsing, table rendering, and export happen in your browser, and the only data that leaves it is the target hostname or IP address sent to the probes. TestMu AI maintains this tool alongside its testing cloud of 3000+ browsers and 10,000+ real devices, where region-specific latency is a daily question.

Why Does a Traceroute Test Matter for QA and DevOps Teams?

Ping tells you a host answers; a traceroute test tells you where the time goes. Here is where a traceroute earns its place in a testing and release workflow:

• Flaky cloud test runs: when automated suites on HyperExecute time out against a staging host, a trace from the runner region shows whether the delay sits in your network, a transit provider, or the destination.
• Regional performance bugs: a page that is fast in London and slow in Sydney is usually a routing or CDN problem, not an application bug, so compare the region tabs before filing it against the frontend team.
• Firewall and CDN verification: a TCP trace on port 443 confirms the request reaches your CDN edge or load balancer, and the resolved IP shows which point of presence each region hits.
• Mobile app backends: before a release, trace the API host from the regions where you run tests on the real device cloud so a slow path is not mistaken for a slow app.
• Incident evidence: download the raw output and attach it to the ticket, because ISPs and hosting providers ask for exactly this when you report a routing issue.

How Do You Run a Traceroute Online?

Running a traceroute online takes one input and one click. The five steps below map to the controls in the widget above, and the result normally arrives within a minute:

• Enter the target: Type a hostname, IPv4, or IPv6 address in the Target box, or click Load sample to load example.com. A pasted URL is reduced to its hostname.
• Pick a protocol: Keep ICMP for a general trace, or choose TCP and set a port such as 443 when ICMP is filtered. UDP mirrors the classic Unix behavior.
• Choose regions: Five regions are selected by default. Click any of the 16 region chips to add or remove it, use the All, Default 5, and Clear links, or type a city, country, or ASN in the custom location box.
• Click Run Traceroute: The probes start within a few seconds and the full result normally arrives within a minute.
• Read, copy, or download: Each region gets its own tab with a hop table, and the tab shows the latency to the destination. Copy puts the raw text on your clipboard and Download saves it as a .txt file.

What Are the Features of the Traceroute Online Tool?

As a network tool, this traceroute online utility adds a few things a plain command-line trace does not. Here are the features of the tool:

• Up to 16 regions at once: presets from New York to Johannesburg run in parallel, plus any custom city, country, or ASN, and each region gets its own tab with a destination latency badge.
• Protocol selector: ICMP echo, TCP SYN to a port you choose (default 80), or UDP, matching the three probe methods in the Linux traceroute manual.
• Latency bars and jump notes: each hop's average is drawn against the slowest replying hop in the region, and a rise of more than 50 ms over the previous hop is flagged.
• No reply and destination status: silent hops are marked instead of shown as bare asterisks, and every tab states whether the destination itself answered.
• Summary tiles: fastest region, slowest region, hop range, and how many regions reached the target appear above the tabs for a one-glance comparison.
• Raw output, copy, and download: the untouched probe output is one click away, copies to the clipboard, and saves as a .txt file named after the target.
• URL and IPv6 handling: a pasted URL is reduced to its hostname, and IPv4 addresses, IPv6 addresses, and hostnames are all accepted as targets.

How Do You Read Traceroute Results?

Each row is one router on the forward path. The Min, Avg, and Max columns summarize the round-trip times of the 2 probes sent to that hop, the Latency bar scales each average against the slowest replying hop in that region, and the Note column flags the patterns below:

• Latency jump: a row is highlighted when its average is more than 50 ms above the previous responding hop, and a jump that persists on every later row, including the last one, is a real delay such as a long-haul link.
• Jump that disappears: if the next rows drop back down, the slow router was only slow at generating ICMP replies, and your traffic is not delayed there.
• Last hop: the final row is the destination itself, and its latency is the number that matters for users in that region; intermediate values only explain where it comes from.
• Trace ends early: if the table stops with No reply rows and never shows the destination, a firewall is dropping the probe type, so re-run with TCP on the port the service listens on.

A worked example from the New York probe to example.com, captured in the screenshot above: hops 1 to 5 were DigitalOcean routers at 1.5 to 2.7 ms, hop 6 returned No reply, hop 7 was a Cloudflare address at 2.4 ms, and hop 8 reached example.com at 172.66.147.243 in 2.4 ms. The silent hop 6 changed nothing, because hops 7 and 8 both replied within a millisecond of the hops before it.

Pair the trace with the DNS Propagation Checker when the resolved IP differs between regions, and with the IP Address Lookup to identify the network that owns a slow hop.

What Is the Difference Between Traceroute and Ping?

Ping answers "can I reach this host and how fast", while traceroute answers "which path do my packets take and where does the delay start". Use ping first, then traceroute when ping shows loss or high latency.

Frequently Asked Questions (FAQs)

What do the asterisks or No reply rows in a traceroute mean?

An asterisk or No reply row in a traceroute means the router at that hop did not return an ICMP Time Exceeded message within the timeout. Most routers still forward traffic normally while ignoring probes. Treat the hop as a problem only when every hop after it also fails or shows the same delay.

Why does the traceroute show private IP addresses in the first hops?

Private IP addresses such as 10.x.x.x or 192.168.x.x in the first hops belong to the internal network of the probe running the test. Those hops are the data center gateway and switches before the public internet. They normally reply in under 2 ms and can be ignored when diagnosing an external route.

Does this traceroute run from my computer or from a remote server?

This traceroute online tool runs from remote Globalping probes in the regions you select, not from your computer. A web page cannot send packets with a custom TTL, so a browser-based traceroute always needs a remote vantage point. To trace from your own machine, run traceroute on Linux or macOS or tracert on Windows.

Which protocol should I choose: ICMP, TCP, or UDP?

Choose ICMP for a general traceroute, because ICMP echo is the most common probe method and the default here. Choose TCP with port 443 or 80 when a firewall drops ICMP but allows web traffic, since a TCP SYN probe looks like a normal connection. Choose UDP to match classic Unix traceroute behavior.

Why is high latency at one hop not always a problem?

High latency at a single traceroute hop is often caused by the router generating its ICMP reply on a slow control plane while forwarding real traffic in hardware. The reply time reflects how busy the router is, not the path. A hop only indicates a real delay when every later hop shows the same increase.

Does a traceroute show the return path from the destination?

A traceroute shows only the forward path from the probe to the destination. Each round-trip time still includes the return trip, but the reply can travel through different networks that the trace never lists. Asymmetric routing is common, so a slow return path can raise latency on hops that look fine on the forward route.

How many hops and probes does this online traceroute send?

This online traceroute sends 2 probes per hop and stops after 20 hops, the Globalping probe configuration shown in the raw output header. The Linux traceroute command defaults to 3 probes per hop and 30 hops. The example.com trace from New York finished in 8 hops, so the shorter limit rarely matters.

Is there a limit on how many traceroutes I can run?

Yes, the Globalping API that runs the probes allows 250 free measurements per hour for each unauthenticated IP address, and each selected region counts as one probe within a single measurement. If you hit the limit the tool shows an hourly limit message, and the allowance resets within the hour.

Can I trace an IPv6 address or a full URL?

You can trace an IPv6 address, an IPv4 address, or a hostname with this traceroute online tool. If you paste a full URL such as https://example.com/path, the tool strips the scheme, path, and port and traces the hostname. IPv6 traces need probes with IPv6 connectivity, so a region may occasionally return no result.

Why do traceroute results differ between regions?

Traceroute results differ between regions because each probe starts on a different network and BGP picks a different set of transit providers to reach the destination. A host served by a CDN often resolves to a different IP in each region. Comparing regions shows whether a slow path is global or limited to one geography.