Maintaining packet order in a parallel processing network device

A network device, comprising:

• a plurality of network ports configured to receive and to transmit packets on a network;
• a plurality of packet processing elements configured to process packets received via the network ports;
• and
• a plurality of ordering including at least a first ordering unit and a second ordering unit, the first ordering unit configured to:
• distribute a plurality of packets among two or more packet processing elements for processing of the packets, wherein the plurality of packets belong to a data flow;
• queue a first instance of a packet of the plurality of packets according to an order of the packet within the data flow;
• cause the first instance of the packet to be transmitted when both of (i) processing of the first instance of the packet is completed and (ii) the first instance of the packet is at a head of a queue at the first ordering unit;
• and
• cause a second instance of the packet to be queued at a second instance of the packet according to an order of the packet within the data flow;
• and
• wherein the second ordering unit is configured to cause the second instance of the packet to be transmitted when both of (i) processing of the second instance of the packet is completed and (ii) the second instance of the packet is at a head of a queue at the second ordering unit.