Duckdns issue. Network crashes and modem / router restarts

Hello. Please help. I have hassio installed in RPI 3b + in docker. If I set it locally as 192.168.0.101:8123 all works ok. Supervisor instantly opens and addons the same, but if I set hassio remotely as mydomain.duckdns.org:8123 (Port forwarding 8123,80,443,1880) the supervisor has a delay of about 10s and when I open some addons like Adgurad or Nodered my router restarts.

Why are you forwarding all those ports?

Are you comfortable having anyone in the world access running whatever is on 80, 443 and 1880?

ok…now is only 8123 and the issue is the same.

Take a look at your router logs. Does it point to what is happening? Kind of sounds your router is getting overloaded.

i’ve got from router info : Illegal - Dropped INPUT packet: SRC=xxx.xxx.xxx.xxx MAC=xx:xx:xx:xx:xx:xx (mac of RPI) and when i made a network diagnostic i’ve got Your home network has a few problems . -- Checking Ethernet Connections and without rpi the network test passes ok.

This has been wrecking my head for some time also but seems to be happening more and more. Initially it was just opening up camera streams.

Now when I try download snapshots, open camera stream, open adGuard it happens alot.

Do you have a Virgin media superhub 3 by any chance? Someone mentioned its due to the hub not supporting hairpin NAT (which I have no clue about actually)

EDIT:… actually decided to read up on what that is and seems quite possible its the same issue. “Without NAT hairpin, you cant access your external IP from inside your network.” is what i’m summing up. I just stress tested a lot of stuff and no problems when I access from 192.168.0.xx but when I try to access things from external IP it causes problems.

Would love to know if there is a workaround without having to reorganise modems