We have a device discovery protocol which uses a UDP broadcast to find devices on the local network, and I'm trying to understand the behavior of windows XP networking in a corner case. When writing to the UDP broadcast address, without specifying an IP, the machine will normally send out the broadcast packet on all NICs, as desired. However, if the IP addresses and subnet masks on the two NICs overlap (example below) it will only send out the broadcast on one card. example: Eth0 IP: 10.0.x.y Subnet: 255.255.255.128 Eth1 IP: 10.a.b.c Subnet: 255.0.0.0 In this case, the broadcast will only be sent out on Eth0. If the subnet mask for Eth1 is changed to a typical 255.255.255.0, then both NICs will be used. Is this expected (and explainable) behavior? Is there a resource somewhere to explain the inner workings of windows routing in scenarios like this? Anyone run into something like this before? Thanks, yalurker

You have to correct your subneting to get it to work. In a well configured subneting, there is no overlapping. When you've corrected you subnet configuration, each subnet has it's own broadcast address and will work as desired. Please send a reply, if you solved the problem !!!

? 255.0.0.0 has 10.255.255.255 for broadcast 255.255.255.128 has 10.0.0.127 for broadcast Even the subnet ids are different with the 10.x.y.z I don't see your logic here.

I think you expect both nics to reply to broadcast? "Best Practices", Event viewer, host file, perfmon, antivirus, anti-spyware, Live CD's, backups, are in my top 10