My script in the cluster resource load script for DHCP reads:

CLUSTER DHCP CN=fptest.O=logan.T=clustertest
DHCPSRVR -d3 --servaddr=10.1.15.11

10.1.15.11 is the address of my server

when I yank the power chord of 10.1.15.11 the resource fails over to my gwtest server which is 10.1.15.12

The log shows a request from the mac address
The dhcp server tries to send a reply back as a subnet broadcast, but it is not getting a response from the workstation. This is happening all on the same gigabit switch.

I have read all the documentation
I understand that the documentation was pretty screwed up on clustering and dhcp. I am wondering if the tid to fix it, 10090710, is telling me everything.