| | 1 | = Design Noteds for nodehandler = |
| | 2 | |
| | 3 | == Underlying model == |
| | 4 | * Broadcast |
| | 5 | nodehandler broadcasts each command to all the nodes using a C program |
| | 6 | |
| | 7 | == System Parameters == |
| | 8 | * Underlying traffic model: Figure out |
| | 9 | * how many packets |
| | 10 | * of what size |
| | 11 | * and with what interarrival |
| | 12 | are generated by a "typical" experiment script |
| | 13 | |
| | 14 | * For this traffic model, test the performance in terms of packet loss, and avg. latency |
| | 15 | |
| | 16 | == Building reliability into the protocol == |
| | 17 | Some thoughts on building reliability into the protocol |
| | 18 | 1. Cumulative ACK by nodeagent every 5 packets |
| | 19 | 2. Stagger sending of ACK's to reduce the collision domains based on |
| | 20 | {{{ |
| | 21 | N* backoff interval, where N = row no. |
| | 22 | }}} |
| | 23 | 3. nodehandler separate thread maintains a bitmap for all packets and fills a "one" for every missing packet reported |
| | 24 | by the nodeagent |
| | 25 | 4. nodehandler broadcasts these missing packets after a ''timeout'' |
| | 26 | 5. This timeout is based on the time taken by all nodes to send an ACK to the nodehandler using the convention of Step 2 |
| | 27 | |
