Question about VPN Configurations

[NETRiNA]

Hi there,

I'm using AdGuard Pro and I'd like to know how it uses the NetworkExtension API in iOS. I understand how a split-tunnel VPN configuration work in a traditional use case but what's up with the AdGuard Pro case?

Does using split-tunnel configuration mean that only DNS requests (port 53) or specified domains (per list used) are getting through the virtual VPN (127.1.1.1) interface? All other IP traffic is getting straight through the real (WiFi/Mobile data) interfaces?

When using full-tunnel mode, does that mean that all IP traffic is routed through the virtual VPN?

Is there a difference in performance/overhead between the two?

The question came to my mind because while I was using YouTube, I noticed that with AdGuard Pro VPN enabled, the Cast to TV button is not enabled, even when using split-tunnel.

Is that because no matter which configuration, the App reads the "fake VPN" ip as it's assigned IP? (Instead of my LAN's client IP?)

Thanks in advance

 

[Jheeze]

I’m also interested in this. YouTube will show the cast to TV button, but won’t show any of the TVs. Just a button for airplay & Bluetooth devices whilst the adguard VPN is on.

 

[vasily_bagirov]

In both modes, only traffic that has our fake DNS addresses (121.121.121.12x) as edstination will be going through he tunnel. The rest of the traffic will not be tunneled. So the difference between modes is in the tunnel configuration.

In the split mode, we tell the system that we will only process 121.121.121.121 requests, and in the full mode we declare that we will process all requests, save for exceptions; the trick is that exceptions are all addresses except for 121.121.121.121. This move helps increase the tunnel priority (it can't be ignored as with split mode).