Use WinMTR Tool To Test Network 

What is MTR

When surfing the Internet, you often encounter that a website cannot be accessed or the access speed is very slow, you may doubt your own computer or network. At this time, you can use WinMTR to find out the abnormality of the network.


The tool integrates the graphical interface of tracert and ping commands. Using winmtr, you can directly see the response time and packet loss rate of each node, which is suitable for route tracing and PING testing under Windows. The use method is simple, WinMTR does not need to be installed, it can be executed after decompression. Ping and tracert are often used to detect network conditions and server status.

The ping command will send a packet to the specified server, and if the server has a response, it will send the packet back, and it will also tell us the time the packet goes back and forth. The tracert command is used to tell us which nodes (routes) and the response speed of each node will pass through from the user’s computer to the specified server.


After downloading WinMTR, double-click to run it. You can find the download link from the end of this article. After opening, we can see the text box in the Host column, enter the IP or domain name you want to track in the Host text box, and then press Start, then you can see the tracert and Ping results as shown below.

GUI Interface:

  • After tested ended for a few more minutes, we can export the test results.
  • Copy Text to clipboard– copy the results to the clipboard in text form
  • Copy HTML to clipboard– copy the results to the clipboard in HTML form
  • Export TEXT – export the results in text form
  • Export HTML– export the results in HTML form
  • Options– There are several options that can be set

WinMTR test result glossary

  • Hostname:The host IP or name of each node to pass through to the destination server.
  • Nr:The number of nodes passing through; the US computer room in Los Angeles in the above figure is an example: a total of 11 nodes must be passed through, the first of which is the gateway of the local broadband provider.
  • Loss%:pingThe percentage of data packets that fail to reply; from this, it can be judged that the node (line) is faulty, whether it is the computer room where the server is located or the international routing trunk.
  • Sent:The number of packets that have been transmitted.
  • Recv:Number of packets successfully received.
  • Best:Minimum response time.
  • Avrg:Average response time.
  • Worst:Maximum response time.
  • Last:The response time of the last packet.


When using the server connection to feel that there is network delay or packet loss, the client is required to provide the MTR test results from the local to the server and from the server to the local ip.

  • Windows version MTR download address:
  • Centos installation command
yum install mtr

From local to remote server

WinMTR tool test results.

Enter the server IPaddress and click startto start the test to send data packets. You need to let the program run more than one minute. If the time is too short , there will be not enough sampling data, then, there will be some errors in the analysis results.

A simple analysis to above result

The local IP in the above figure goes to This server needs to go through 11 nodes to arrive. From hop 1 to 9,there is no packet loss and the response time is very fast indicating that the network is very good.
But at the hop 10, sometimes there is packet loss and the network does not respond. Between hop 10 and 11, there is no packet loss at all and the response time is very fast.
This is a normal phenomenon, because the hop network device, there will be certain defense mechanisim which will filter some packets. This is the caused by network routing devices, not a network problem. (The small probability of packet loss of a single node can also be ignored)

So what can be counted as broken network or lost packets delay? There is a phenomenon of packet loss in multiple sequent hops, and the average responding time is long, which will be considered as a network problem if such a situation occurs.

When doing the test, the pakcet jumps from one network node to the next node of the server. It needs locate the hop to see if the packets entered into the server, where lost the packets or responding time too long. If the packet loss or the response time long happens in the inside of the server, then the problem should be resolved by network service provider. If not, there is a delay in the average access speed, then It should contact local network servie provider to deal with packet loss or response speed issues.

For example, in the figure above, the hop 9,IP is After entering this node, accessing data is officially entered into the service node provided by the server provider. If packet lost or responding time too long exists after this, then the server provider should troubleshoot the location and resolve it. If before that it already slow, you should contact the local network provider to solve it.

From server to local

Log in to the server to install MTR, and then run the MTR command. such as the test server to my local IP:
Tip: The local IP can be displayed directly by typing IP in Baidu or Google

Then run the command mtr, let the test run for a while and then check the data for analysis
as shown in the figure:

After the server has passed 11 nodes, sending data to my local. It is the same as the previous local to the server analysis, first to see if there is continuous packet loss or a problem with the response speed. If not, it proves that the network is normal. If there is a packet loss or a problem with the response speed, it also depends on whether there is a packet loss or a problem with the response speed during the transmission process of our server node, or a problem with the packet loss or the response speed after it has left our server.

Running MTR in the server clearly identifies our device name. The data transmitted before is still transmitted in our server node, and it is sent to others after arriving at of network operators. If there is a problem before this, then we should check whether there is a problem between the node devices. If there is continuous packet loss or a problem with the response speed after this, then we have to check the network settings, or contact the network provider to solve it for us question.

Windows Server

If the server is not a Linux system but a Windows Server, then download WinMTR in the server and do the same test to check the network node situation between the server and the local IP, and then analyze it.


Windows version MTR download address:

OneDrive Link:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.