Skip to content

mtr

Combines ping and traceroute: measures the way to a host continuously and shows packet loss and response times for every router.

mtr output

Syntax

mtr [OPTION]... HOST

Without options, mtr shows a live view that updates every second. Quit with Q.

Options

Option Description
-r, --report Report mode: measure, then print the result once
-c NUMBER Number of rounds, default 10 in report mode
-n Don't resolve router names
-w Wide report: show full host names
-T -P PORT Use TCP to this port instead of ICMP
-4, -6 Force IPv4 or IPv6

Examples

Create a report for a host:

mtr -r example.com
Start: 2026-10-26T10:42:17+0100
HOST: tuxdex-srv                  Loss%   Snt   Last   Avg  Best  Wrst StDev
  1.|-- _gateway                   0.0%    10    0.3   0.3   0.3   0.4   0.0
  2.|-- core1.isp.example.net      0.0%    10    1.1   1.1   1.0   1.2   0.0
  3.|-- ???                       100.0    10    0.0   0.0   0.0   0.0   0.0
  4.|-- ix1.fra.example.net        0.0%    10    4.9   4.9   4.8   5.0   0.1
  5.|-- edge2.example.net         20.0%    10   11.4  11.5  11.3  12.1   0.2
  6.|-- example.com                0.0%    10   11.6  11.7  11.6  11.9   0.1
Column Meaning
Loss% Lost packets at this hop
Snt Packets sent
Last, Avg, Best, Wrst Last, average, best and worst response time in ms
StDev Fluctuation of the times: high values mean an unstable connection

Hop 5 loses 20 %, but the target loses nothing. The router only answers mtr slowly and still forwards all packets. Only loss that continues up to the target is a real problem.

Without name resolution and with more rounds:

mtr -r -n -c 10 example.com
Start: 2026-10-26T10:42:17+0100
HOST: tuxdex-srv                  Loss%   Snt   Last   Avg  Best  Wrst StDev
  1.|-- 192.0.2.1                  0.0%    10    0.3   0.3   0.3   0.4   0.0
  2.|-- 198.51.100.1               0.0%    10    1.1   1.1   1.0   1.2   0.0
  3.|-- ???                       100.0    10    0.0   0.0   0.0   0.0   0.0
  4.|-- 198.51.100.65              0.0%    10    4.9   4.9   4.8   5.0   0.1
  5.|-- 198.51.100.129            20.0%    10   11.4  11.5  11.3  12.1   0.2
  6.|-- 198.51.100.7               0.0%    10   11.6  11.7  11.6  11.9   0.1

Measure through firewalls with TCP:

sudo mtr -r -T -P 443 example.com
Start: 2026-10-26T10:42:17+0100
HOST: tuxdex-srv                  Loss%   Snt   Last   Avg  Best  Wrst StDev
  1.|-- _gateway                   0.0%    10    0.3   0.3   0.3   0.4   0.0
  2.|-- core1.isp.example.net      0.0%    10    1.1   1.1   1.0   1.2   0.0
  3.|-- ???                       100.0    10    0.0   0.0   0.0   0.0   0.0
  4.|-- ix1.fra.example.net        0.0%    10    4.9   4.9   4.8   5.0   0.1
  5.|-- edge2.example.net         20.0%    10   11.4  11.5  11.3  12.1   0.2
  6.|-- example.com                0.0%    10   11.6  11.7  11.6  11.9   0.1

When do I need it?

  • When a connection is slow or drops now and then
  • To find out whether packet loss happens in your own network, at your provider or at the target
  • As evidence for your hosting provider: the report shows exactly where packets are lost

Measure longer for sporadic problems

mtr -r -c 300 example.com measures for five minutes. Problems that only occur now and then show up much more reliably than with ten rounds.

Loss in the middle isn't always a problem

Many routers answer test packets with low priority. Loss at a single hop that doesn't continue to the following hops can usually be ignored.

Distribution differences

Debian Ubuntu RHEL Arch
Package mtr-tiny (or mtr with a graphical interface) mtr-tiny mtr mtr
Installed by default no yes no no

traceroute ยท ping