Hello,
Most enterprise wireless APs seem to record the IP address of a connected
client -- either/both in an 'association' table and/or in a RADIUS
accounting record as a 'Framed-IP-Address'.
It always seems correct but I'm wondering where it comes from?
Obviously, in the old* dial-up days, the access server knew the address
because it supplied it to the connecting device via PPP or whatever.
However, when a wireless access point set up in a bridge mode (as ours
are), it's really just operating at layer 2 and, I assume, the only way
it's going to find out the address of the client is by snooping ARP or
DHCP packets.
Does anyone know how this actually works and (more importantly) can I
trust it to detect the address of the connecting device, or is it just
useful as a hint to avoid an extra step of lookup of MAC -> IP in ARP or
DHCP logs?
- Bob
--
Bob Franklin <[log in to unmask]> +44 1223 748479
Network Division, University of Cambridge Computing Service
|