Understanding mwan3 Behavior: Deactivation of WAN Interfaces and Network Isolation

I've been working on implementing a network switching function using mwan3 for multiple WAN interfaces. I have a question regarding the behavior of mwan3 when WAN interfaces are deactivated.

Is mwan3 a purely software-based solution, or does it involve any hardware-level actions such as physical disconnection of WAN interfaces?


  • WAN1, WAN2, and WAN3 are all network-activated and utilized with mwan3 policies.
  • If I deactivate all WAN interfaces except WAN1, does mwan3 ensure that there is no packet communication between WAN2 and WAN3?

I want to clarify whether mwan3, when deactivating WAN interfaces, actively prevents packet communication at the software level or if there can still be some low-level packet communication or signaling at the hardware level between the inactive WAN interfaces.