We are working on Pritunl deployment. However we are facing to some issues.
Using Windows Pritunl client (up to date version), for users connected with 1Gb wire, the metrics applied to route define on Pritunl virtual interface and physical interfaces are the same when configured with automatic metric: value=25 (The Automatic Metric feature for IPv4 routes - Windows Server | Microsoft Learn).
We are using Pritunl to route all traffic from client (route 0.0.0.0/0 set)
So it means DNS server announced by Pritunl will not be prioritize because it’s not applied on the lowest interface metric.
If I manually set a low metric value on the “Pritunl 1” interface, everything work as expected. I’ve also identified that a metric value can be configured on the route setting in Pritunl (in Advanced UI), but I don’t find any documentation about this, and it seems having no impact on the metric applied on the client.
Excepted enforcing a static value for metric on the client virtual interface, what is the most appropriate solution to achieve prioritize Pritunl to always route all traffic ?
Is anyone know something about this ?
We are observed that metric seems to be automatically reset to “Automatic value”. So it automatically go back to the default metric value…
This is a real issue because it means VPN cannot properly work for users with 1G wire connection…
Is there a possibility to automatically run a command or apply a config once the VPN is connected ?
The code for this feature has been added to the repository. It will be available in the next release under the Advanced Settings in the top right menu.
Thanks @zach for the update, we need to spent more time to test, but it looks very promising !
Could you point us to the best solution to apply this setting during deployment on all our devices ?