We are trying subnet a class C with a 27 bit netmask (.224) on a
Livingston PM2e to provide 8 subnets of 30 hosts per subnet, so that
our sys admin can dial in via an Ascend Pipeline 25 on an ISDN port and
have his own 30 host subnet. We have also assigned addresses for 20 or so
dial-in users in an other subnet. The PM2e is configured by adding
a netmask to the class C in question (one address of which is on ether0)
and adding a static route to the Pipeline's Ethernet interface, an address
on his private subnet:
add netmask 199.199.99.0 255.255.255.224 ;our class C
add route 199.199.99.96 (his subnet) 199.199.99.97 (his WAN port on the
P25)
The Pipeline has a LAN port address as well as a WAN (serial) port
address-- should the gateway for his subnet be the LAN or serial port?
The ISDN link works fine, and telneting into the PM2e, he can ping the
LAN port on the P25, but when he pings the serial port on the P25, he can
see packets
arriving and transmitting but the PM2e never registers the echo packet and
thinks it's unreachable.
Do we need to be dealing with two separate class C's-- one subnetted for
the assignment pools and one for the segment on ether0? Our
default gateway is an address on the class C in question which is also a
secondary address on a Cisco 2503 ethernet port which routes to our ISP
and our other class C networks (secondaries on the same Cisco ethernet
interface).When we apply the netmask to the class C, packets get lost for other
dial-up hosts on the PM2e and all heck breaks loose.
Any thoughts or suggestions are appreciated...
Mick Maldonado