Skip to main content

VPS: Netzwerkprobleme mit WinMTR/MTR erkennen

This guide was created with the following products:

(Details may vary with products from different providers but the main concepts remain the same)

VPS

Introduction​

Network problems can be disruptive and, of course, should not be part of everyday life. However, when they do occur, it is important to quickly and efficiently identify the cause in order to resolve the issue. Detailed information about network problems can be effectively analyzed using the WinMTR (Windows) or MTR (Linux/Mac OS) application.

img

WinMTR/MTR is a network diagnostic tool that combines the functions of Ping and Traceroute. It allows for detailed analysis of both the outbound route (Client → Server) and the return route (Server → Client) by tracking the path of data packets in both directions. In doing so, it captures important information such as latency and packet loss, which are crucial for accurately diagnosing and addressing network problems.

Outbound route (Client → Server): A report for the outbound route is useful when issues such as connection drops, slow connections, or difficulties establishing connections arise. This analysis helps identify potential problems on the path from the client to the server, such as network congestion, packet loss, or faulty routing.

Return route (Server → Client): A report for the return route only makes sense once a stable and functional connection on the outbound route has been confirmed. This means the return route report becomes particularly relevant when the outbound route shows no issues, but problems such as slow server responses, delayed loading times, or incomplete data transmissions still persist.

In summary, if noticeable network problems occur, the outbound route (Client → Server) should be analyzed first. If the outbound report shows no anomalies and the issues persist, it is necessary to check the return route (Server → Client) to identify possible asymmetric problems. Asymmetric network issues can arise when data packets are transmitted without problems in one direction, while delays or packet loss occur in the opposite direction.

Installation​

Now that it has been clarified when reports for the inbound and outbound routes are useful, we can move on to the installation on both the client and server levels.

Client​

For the outbound route, the application must be installed on the client (your computer). The following steps will explain how to install the application for your operating system.

When using a computer with a Windows operating system, the network analysis is conducted through the WinMTR application. To install WinMTR on your computer, first download the application from the ZAP-Hosting website. After downloading, extract the file. As a result, you will obtain the executable WinMTR.exe file.

ApplicationDownload
WinMTRLink

Server​

For the return route, the application must be installed on the server. The following steps will explain how to install the application for your server's operating system.

Connect to your server via SSH. To install MTR on your system, execute the appropriate installation command for your server's operating system:

Debian

apt install mtr -y

Ubuntu

apt install mtr -y

OpenSUSE

zypper addrepo https://download.opensuse.org/repositories/openSUSE:11.4/standard/openSUSE:11.4.repo
zypper refresh
zypper install mtr

Fedora

yum install mtr -y

Create Report​

In the following, a report will be created using the WinMTR/MTR application. This will be done for both the outbound route (Client → Server) and the return route (Server → Client).

Create report during active issues

The report should be generated when the issue is actively occurring and noticeable. Only in such a case will the report provide useful information that can aid in troubleshooting and resolving the problem.

Outbound route (Client → Server)​

To analyze the route from your computer to the server, open the WinMTR/MTR application on your computer and follow the steps outlined below.

The WinMTR application can be launched by double-clicking the WinMTR.exe file. In the Host field, enter the IP address or hostname of the server. Then, start the report. Let the application run for at least one minute to collect enough data. Once you have sufficient data, stop the analysis and save the result by clicking on Export TEXT.

img

Return route (Server → Client)​

To analyze the route from the server to your computer, connect to your server and follow the steps outlined below.

The MTR application can be used on your server by executing the following command in the SSH terminal:

mtr <your client-ip-address>

Let the application run for at least one minute to collect enough data. Once you have sufficient data, stop the analysis and save the result.

info

Don't know your IP address or unsure how to find it? There are several methods to determine your IP address. The quickest way is through an online service like WhatIsMyIPAddress.

Evaluate report​

When evaluating the results, there are several key points you should keep in mind. The following section will explain these points in more detail and clarify their significance for the analysis. Paying close attention to these aspects is crucial for making an accurate diagnosis and identifying potential causes effectively.

Packet loss​

If you see packet loss in the results, this indicates potential network problems. A minor temporary packet loss of 1-2% might not be problematic, but higher values suggest more serious issues. Packet loss can cause delays or interruptions in the services' connections. If the loss is evenly distributed across all hops, the problem could lie within your own network or on the server side. However, if the loss occurs only at a specific hop or area, the issue is likely with that node or the connection to the next one. It's also important to know that small losses in the first hops, which belong to your local network, are not necessarily critical, as these devices often prioritize ICMP requests (such as ping) lower and may discard them.

Latency (ping times)​

The latency values (Avg, Best, Worst) give you insight into the speed and stability of the connection. If you notice that latency remains consistently high at a particular hop, this could indicate network congestion or a slow router. A sudden increase in latency between two hops points to a possible bottleneck. Normally, latency gradually increases along the route to the destination. However, be on the lookout for sudden, significant spikes, which are often a sign of an issue. Comparing the outbound and return routes can also be helpful in identifying asymmetric problems, which indicate that traffic in one direction is encountering difficulties.

With careful analysis of these factors, you can accurately determine where the network problem lies—whether in your own network, with the internet service provider, or somewhere along the path to the server.

Diagnostic examples​

To enhance your understanding, let's clarify the previous information with a few diagnostic examples. We have compiled a few scenarios that illustrate different situations and possible causes. These examples will help demonstrate how to interpret the data and diagnose network issues more effectively.

info

The following diagnostic examples illustrate various scenarios based on fictional WinMTR/MTR reports. The IP addresses, hostnames, and routes contained within are entirely fictional and have no connection to real networks or hosts.

Example: Packet Loss at the Client

Description

In this example, the situation describes packet loss occurring on the client side, which is visible on both the outbound and return routes.

Outbound Route (Client → Server)

Host                               Loss%   Snt   Last   Avg  Best  Wrst StDev
1. 192.168.1.1                     15.0%   20    1.1    3.4   1.0   10.2  2.8
2. 10.0.0.1                        10.0%   20   12.3   15.7  10.5   30.8  5.5
3. 172.16.0.1                       0.0%   20   25.1   26.0  24.8   27.2  0.6
4. 203.0.113.5                      0.0%   20   35.4   34.5  33.2   36.1  0.9
5. be1234.ccr42.isp.net             0.0%   20   50.6   49.8  49.1   51.5  0.8
6. be5678.ccr21.isp.net             0.0%   20   60.1   59.9  59.5   61.0  0.4
7. 198.51.100.10                    0.0%   20   70.2   70.5  70.0   71.3  0.5
8. Ziel-IP (Server)                 0.0%   20   80.2   80.0  79.5   81.2  0.6

Return Route (Server → Client)

Host                               Loss%   Snt   Last   Avg  Best  Wrst StDev
1. Ziel-IP (Server)                 0.0%   20   80.2   80.0  79.5   81.2  0.6
2. 198.51.100.10                    0.0%   20   70.1   70.4  69.8   71.2  0.5
3. be5678.ccr21.isp.net             0.0%   20   60.2   60.0  59.6   60.7  0.4
4. be1234.ccr42.isp.net             0.0%   20   50.1   50.2  49.5   51.0  0.5
5. 203.0.113.5                      0.0%   20   35.5   35.0  34.5   36.0  0.4
6. 172.16.0.1                       0.0%   20   25.2   26.1  24.9   27.3  0.7
7. 10.0.0.1                        10.0%   20   15.8   16.5  15.0   35.2  6.2
8. 192.168.1.1                     15.0%   20    2.0    4.0   1.2   12.0  3.8

Evaluation

The evaluation reveals that packet loss is occurring on the client side. Both routes show losses at the first hops (Hop 1 and 2), indicating an issue on the client's side, either within the local network or with the router.

Reporting Issues​

Although ZAP-Hosting continuously monitors the network, it is important to contact our support team promptly in case of network issues. To ensure an accurate diagnosis and a quick resolution, we kindly ask you to provide us with the results of your WinMTR/MTR reports for both the outbound and return routes. This information helps us to precisely identify potential problem areas and resolve the issue quickly.