Common issues and Debugging
This page provide the most common pitfalls users might have with VXLAN tunnel. This page is largely under construction as we are gathering more edge cases. Your feedback is much appreciated.
Last updated
Was this helpful?
This page provide the most common pitfalls users might have with VXLAN tunnel. This page is largely under construction as we are gathering more edge cases. Your feedback is much appreciated.
Last updated
Was this helpful?
It is crucial to ensure that you can directly connect to the services or games you are hosting using their IPv4 address and port. For instance, in this setup documentation, the was accessible via 108.61.149.182:6900
. If your VXLAN tunnel is not functioning correctly, several potential causes may exist. Below are some best practices to follow to confirm that your service is properly set up and running.
After running the setup script, you can run the following commands to check your tunnel configuration:
Example output:
You should see 0 under the errors and dropped columns. Any other numbers would indicate packets being dropped (possibly from a firewall).
To verify that your VXLAN tunnel has been properly established, you can ping the Private IP Address
of your tunnel. You should see responses like the example below:
The local IP address here corresponds to the VXLAN interface running on our anycast server, so the latency you observe will vary depending on your backend’s location. For further debugging, run:
as it can confirm whether the virtual link is properly resolving IP-to-MAC, something ping
won't be able to show:
In this example, the private interface IP resolves to a MAC address within the tunnel, which is 02:ab:cd:34:56:78
As for your tunnel's Public IP Address
, the reply should looks something like this (with <1ms latency):
You can also check the output of:
If the tunnel was properly created, you will see your private IP address in the output.
There are two potential scenarios:
Server Firewall: If you are using a server firewall such as netfilter
, iptables
, or ufw
, you will need to open both the backend port and the VXLAN port to accept external connections, or, in rare cases, explicitly whitelist the VXLAN interfaces.
Using iptables
to Accept Connections to Your Backend Port
Notes: To remove these rules, simply replace -A
with -D
to delete them.
Using UFW
:
To delete these rules, use the following command:
sudo ufw delete allow /udp
sudo ufw delete allow /tcp
Whitelisting VXLAN Interface (Optional)
Although this step is typically not necessary, you can whitelist the VXLAN interface explicitly using iptables
:
Using UFW
:
This error can be encountered by customers using the Pterodactyl panel or other Docker services (though extremely rare). It indicates that the VXLAN tunnel failed to bind to your local interface, most likely because Pterodactyl is currently reserving the exact IP address for its own use.
To debug:
Check the Local IP Address for VXLAN: It should resemble something like 172.18.132.2
.
Run ip route
and you may see an entry like:
In this case, the 172.18.132.2
IP address might conflict with the Docker network configurations. For example, if the pterodactyl0
interface is using 172.18.0.1
, this could block the gateway you are trying to use (172.18.132.2
), as it lies within the Docker subnet.
In case you need to delete the tunnel and re-run the setup script, here is how:
You can also remove the iptables rules in the script, by replacing -A
and -I
with -D
, example:
Sometimes, after rebooting your VM or dedicated server, the VXLAN script may not re-run, causing the service to stop working. To fix this, you can set up a persistent service to automatically run the VXLAN script after each reboot.
Steps to Setup Persistent VXLAN Tunnel:
Edit the rc.local
file:
Add the following lines to execute the VXLAN script at boot:
Save and exit the editor (Ctrl+X
, then Y
to confirm).
Make the script executable:
Now, the VXLAN script will run automatically after each reboot, ensuring that the tunnel is re-established.
Cloud Provider Firewall: Major cloud providers like Linode, AWS, and Azure have default security group or firewall settings that may block traffic before it even reaches your VPS. In this case, you can navigate to your provider's control panel and allow traffic from to your backend ports.