Mastering MTR: Install, Use, and Analyze Network Paths on Linux, Windows, and macOS

MTR (My Traceroute) combines the functions of ping and traceroute into a single network diagnostic tool that uses ICMP packets, providing more convenient and detailed connectivity checks than the separate utilities.

Installation

Windows users can download besttrace.exe from https://cdn.ipip.net/17mon/besttrace.exe or obtain the portable WinMTR from the GitHub releases page. Linux users install via package managers:

# Debian/Ubuntu
apt install mtr
# RedHat/CentOS
yum install mtr

macOS users can install the “Best NetTools” app from the App Store, and Android users can obtain the TracePing client (or download it directly from https://dwz.cn/KCdNPH4c).

Basic Usage

Run mtr qq.com to display the route and latency from the local host to the target domain. The output includes columns such as IP address, loss percentage, packets sent (Snt), last latency (Last), average latency (Avg), best latency (Best), worst latency (Wrst), and standard deviation (StDev).

Parameter Explanation

-r or --report : Sends a fixed number of ICMP packets (default 10) and prints a static report instead of a continuously updating display.

-s or --packetsize : Sets the size of the ping packet, e.g., mtr -s 100 qq.com.

-c : Specifies the number of packets to send, e.g., mtr -c 100 qq.com.

-n : Disables hostname resolution, showing only IP addresses.

MTR Result Analysis

When analyzing the report, focus on two key metrics: loss and latency . A non‑zero loss percentage on a hop indicates possible packet loss; however, occasional loss may be caused by ICMP rate limiting by the provider. Compare subsequent hops to determine whether loss is artificial or genuine.

Latency values (Last, Avg, Best, Wrst) are measured in milliseconds. A high standard deviation (StDev) suggests unstable latency at that hop, meaning the average may not reflect typical performance. Large jumps in latency between consecutive hops often point to geographic distance, misconfigured routers, or congestion.

Both forward and reverse paths should be examined, as high latency can also stem from the return route, which may differ from the forward path.

Resolving Network Issues

Most routing problems are temporary and resolved by ISPs within 24 hours. If persistent loss or latency is observed, capture the MTR report and share it with the ISP, indicating the affected hops. For long‑distance links or peak‑time congestion, consider using a closer node or alternative routing.