$ Linux Command

Network

mtr

Network diagnostic tool combining ping and traceroute.

networkpingtraceroutediagnosticlatency

Additional Notes

mtr (My Traceroute) combines the functionality of ping and traceroute into a single diagnostic tool. It continuously probes each hop along the path to a destination and displays real-time statistics including packet loss and latency for each hop.

It provides a much clearer picture of network problems than ping or traceroute alone. By monitoring each hop, mtr can identify where packet loss or high latency is introduced in the route.

Syntax

mtr [options] hostname

Parameters

  • hostname: The target hostname or IP address.

Common Options

  • -r, --report: Run in report mode (send 10 packets to each hop, then print statistics and exit).
  • -c count, --report-cycles count: Set the number of packets sent per hop in report mode.
  • -i seconds, --interval seconds: Set the interval between probes.
  • -n, --no-dns: Show numeric IP addresses only, no DNS resolution.
  • -b, --show-ips: Show both hostnames and IP addresses.
  • -s size, --psize size: Set the packet size.
  • -t, --tcp: Use TCP SYN packets instead of ICMP.
  • -u, --udp: Use UDP datagrams instead of ICMP.
  • -4: Use only IPv4.
  • -6: Use only IPv6.
  • -o field-order: Customize the output field order.
  • --report-wide: Report mode with wider output.
  • -j, --json: Output in JSON format.
  • -x, --xml: Output in XML format.

Examples

mtr google.com

Run interactive mtr to google.com. Use arrow keys to navigate, q to quit.

mtr -r -c 20 8.8.8.8

Run a report with 20 probes per hop to 8.8.8.8.

mtr -n 192.168.1.1

Run mtr without DNS resolution for faster results.

mtr -r -c 10 -i 1 example.com

Report mode with 10 probes and 1-second intervals.

mtr -t -r github.com

Use TCP SYN packets and run a report.

Understanding Output

  • Loss%: Packet loss percentage at each hop.
  • Snt: Number of packets sent to that hop.
  • Last: Latency of the last packet in ms.
  • Avg: Average latency in ms.
  • Best: Best (lowest) latency in ms.
  • Wrst: Worst (highest) latency in ms.
  • StDev: Standard deviation of latency.

Practical Notes

  • Packet loss at a hop may not indicate a problem with that router; routers often prioritize routing over ICMP responses.
  • Packet loss at the final destination is the most significant indicator of connectivity issues.
  • High latency at the first hop suggests a local network problem (cable, Wi-Fi interference, switch issue).
  • Use -r mode for scripted diagnostics or when sharing results with support teams.
  • Some networks block ICMP; use -t (TCP) or -u (UDP) to work around this.