This is an overarching issue for HA-dpservice.

The two dpservice instances are working in active-backup mode, the second one needs to receive updates about NAT table entries. Otherwise when taking over it is unable to resume NAT communication properly.