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

What is MTR?

MTR (My Traceroute) combines the functionality of ping and traceroute into a single network‑diagnostic tool. It sends ICMP packets by default (unlike traceroute’s UDP) and provides a continuous view of each hop between the source and destination.

Installation

Windows : download the BestTrace executable from https://cdn.ipip.net/17mon/besttrace.exe or obtain the portable MTR client from the GitHub release page https://github.com/oott123/WinMTR/releases.

Linux :

# Debian/Ubuntu
apt install mtr

# RedHat/CentOS
yum install mtr

macOS : search for “Best NetTools” in the App Store and install.

Android : download TracePing from the provided short link https://dwz.cn/KCdNPH4c (or Google Play where available).

Basic Usage

To view the route to qq.com, simply run:

mtr qq.com

Understanding the Output Columns

IP address – the hop’s address.

Loss – packet loss percentage.

Snt – number of packets sent.

Last – latency of the most recent packet.

Avg – average latency.

Best – minimum latency.

Wrst – maximum latency.

StDev – standard deviation (stability) of latency.

Key Command‑Line Options

-r

or --report: generate a static report (default sends 10 ICMP packets). Without it, MTR runs continuously. -c: specify the number of packets to send (overrides the default 10). -s or --packetsize: set the size of each ICMP packet (e.g., mtr -s 100 qq.com). -n: suppress hostname resolution and display raw IP addresses only.

Analyzing MTR Results

The report is divided into three logical sections: the local ISP hops, intermediate routing hops, and the destination ISP hops. Pay special attention to the Loss and Avg/Best/Wrst columns.

Packet loss : If a hop shows a loss percentage but the next hop does not, the loss is likely due to an ICMP rate‑limit imposed by that router rather than a real network problem. Consistent loss across multiple subsequent hops indicates a genuine issue.

Latency : Large jumps in Avg or high StDev values suggest unstable or congested links. A high StDev means latency varies widely, and the average may not reflect typical performance.

When a specific hop shows both high loss and high latency, investigate that node—if it is within your ISP’s network, contact them; if it is beyond, reach out to the upstream provider.

Practical Troubleshooting Steps

Run mtr -r qq.com to capture a snapshot report.

Identify hops with significant loss or latency spikes.

Determine whether loss is due to ICMP throttling (isolated to a single hop) or a real fault (persistent across hops).

Document the report and share it with the relevant ISP or network administrator for further investigation.

When Issues Are Temporary

Most routing problems are transient and resolve within 24 hours. If the issue persists, provide the ISP with the MTR report and any relevant packet‑size or count parameters you used.

Additional Resources

For deeper Linux system administration knowledge, refer to the comprehensive book mentioned in the original source (removed here for brevity).