[Cu-wireless] Re: Cu-wireless digest, Vol 1 #158 - 2 msgs

David Young dyoung at ojctech.com
Thu Jun 27 14:39:20 CDT 2002


That will be two different interfaces, but dhcrelay and dhcpd still
conflict. There is this notion of a "fallback" socket in dhcpd/dhcrelay.
Each creates a "fallback" socket bound to 0.0.0.0.bootps. That is the
problem. I am hacking the relay and server to get around that.

Sort of an esoteric IP & IP implementation question that comes up
as I hack the relay and server is, If I've bound UDP sockets to
192.168.1.1.bootps and 0.0.0.0.bootps, will both sockets receive
UDP packets addressed to 192.168.1.1.bootps? Which socket or sockets
receive UDP packets addressed to 255.255.255.255.bootps? Addressed to
192.168.1.255.bootps (i.e., the broadcast for 192.168.1/24) ?

Dave

On Thu, Jun 27, 2002 at 02:09:18PM -0500, stephane_alnet at ureach.com wrote:
> 
> > A station will run both the DHCP server and the relay when it is both
> > a non-leader member of a pod and the leader of an uplink network.
> 
> So that will be on different interfaces; "dhcrelay -i <if-my-pod>", "dhcpd
> <if-uplink>", I'd guess?
> 
> S.

-- 
David Young             OJC Technologies
dyoung at onthejob.net     Engineering from the Right Brain
                        Urbana, IL * (217) 278-3933




More information about the CU-Wireless mailing list