I am trying to configure load balancing on two 3020 concentrators. When I configured it, I keep getting the following messages

LBSSF received GRAT-ARP from duplicate master[0003a08ab42b]

6167 12/04/2007 16:15:15.240 SEV=3 LBSSF/85 RPT=527

LBSSF detected duplicate master[0003a08ab42b] and staying MASTER

6168 12/04/2007 16:15:18.450 SEV=4 LBSSF/49 RPT=529

LBSSF process dead peer[x.x.x.x (IP address of the secondary box)]

Does anyone know what is causing this?