Packet Routing in Normal Operation

In normal operation, packets are routed downstream from the controller to nodes, and upstream from nodes to controller.

Downstream Packet Routing

Downstream Packets proceed as indicated from controller to all nodes. Each node forwards (repeats) each downstream packet as it is received. All packets eventually reach all nodes.

The redundant link is essentially ignored in this situation. However, it is critical for fault recovery. To make fault recovery possible, the controller is actually sending identical packet streams on both the IN and OUT ports, however Node 1 is aware that its OUT port is redundant, therefore that stream is ignored.

Upstream Packet Routing

Upstream packets proceed as indicated from each node to the controller. Each node forwards (repeats) each packet as it is received from node to node. All packets eventually reach the controller (but all nodes do not see all upstream packets).

The redundant link plays no part in the flow of upstream packets. However, it is critical for fault recovery. To make fault recovery possible, the controller will accept packets arriving at either the OUT or IN ports. SynqNet timeslot allocation ensures that packets will not collide by arriving on both ports at the same time.