why does IGRP change the mask when using secondary ip address with a further subneted mask w/o regard with split-horizon implications ?
I have:
Int s0
ip add 10.4.0.34 255.255.255.248
ip add 10.4.240.1 255.255.255.0 secondary
Int loop20
ip add 10.4.1.1 255.255.255.0
router Igrp
netw 10.4.0.0
________
The loopback gets advertised as 10.4.1.0/28 in the remote router directly connected to the serial0, completely the opposite of what you would expect.