My little rural network consists of a "Main" RT-N16 running Toastman 7483.2 RT Std, 6 WRT54GL secondaries running Toastman 7625 or 7633 ND Std, and 1 RT-N16 secondary running Toastman 7483.3 USB Ext. All routers are configured as AP+WDS, G Only, Channel 11, WPA+AES. I'm running QoS on Main's WAN port. Here's the problem... Until about 2 months ago everything was relatively stable. But lately, several of the secondaries have been periodically dropping their WDS connections. Real-time diagnosis is not possible, because for geographical reasons I can only connect via the RT-N16 secondary - when it's down, I can't get to Main, and I can't see what's going on network-wide. I have to wait for it to reconnect (sometimes 10-20 minutes), after which I can view the logs and bandwidth stats of all routers, which show that every time there's a connection drop, the bandwidth usage on one particular WRT54GL secondary shows a huge spike. I've contacted the homeowner to see what's going on at the time - evidently the kids in the household are playing an online game called "Fantage". Questions... 1. Is "Fantage" a P2P application? I can't find a definitive answer despite diligent web searching. 2. Could the WDS connection drops be caused by a sudden storm of P2P connections overloading Main's CPU and/or memory? I've reduced Max Connections to 2048, hash table size to 512. I've also implemented an ugly QoS rule to force the client in question to the "crawl" class, which seems to have reduced the number of drops, but certainly isn't the ideal solution. 3. Any other suggestions for troubleshooting this issue? Ironically the offending WRT54GL secondary never drops its connection, and has the weakest RSSI of all the secondaries - around 68-72 dBm - so I don't think signal strength is an issue.