I am wondering how an Ethernet-to-CAN converter works. CAN is a multi-master broadcasting network where messages are associated with IDs. Each message type has a unique ID. Devices listen to all messages on the bus and decide whether an ID is of interest. Ethernet, however, uses point-2-point addressing, meaning that sender and receiver are specified. These information cannot be mapped on CAN, because there is not sender or receiver field.
Let's say, device A sends message m to device B. Device B is connected via CAN to the converter. Device A communicates over Ethernet.
How does the CAN-to-Ethernet converter select the message ID when converting an Ethernet frame to a CAN fame? Are the same CAN IDs used for packets to the same recipient?
How does the CAN-to-Ethernet convert a CAN message to an Ethernet frame? A CAN message has no recipient, but an Ethernet frame knows the recipient address. How does the mapping work?