Ok, I've read all the previous threads, searched in vain for a solution, and still can't get my SCO box to respond to a ping thru a VPN tunnel. Here's the particulars:
Router is a Netgear FVS318
Local VPN Router: 192.168.0.1 (on a static ip)
Remote Client: 192.168.1.95
Once the tunnel is established, the client can see all devices EXCEPT for the SCO box. When I veiw the VPN status log on the router to see what devices are available, I see all the Windows PC's, but not the Unix machine.
What is it about SCO's TCP/IP and/or ICMP responses that prevents any off-network machine from seeing it? All the local PC's on the same network subnet (192.168.0.x) can see it. But a remote machine on a different subnet can not.
I've added the IP's to the hosts file for both the router and the remote PC and I've performed a route add to point SCO to the router
route add -net 192.168.1.95 192.168.0.1
and the routing table reflects the entry.
Does SCO do something funky in the kernel parameters that could prevent this type of recognition?
I'm totally tapped out with no where else to search, have been looking for more than two days. Can I also say that this is not unique to a particular client. I have experienced this on several occasions with different clients. Some have gone to the extreme of adding a Linux box attached to the SCO box, then they can telnet to the Linux machine (because it can be recognized) and then step across to SCO. Not a good solution, but it does work.
Any new light would be greatly appreciated.
I'll be glad to provide more details, just tell me what you need.
Thanks
Joel
piperent
Router is a Netgear FVS318
Local VPN Router: 192.168.0.1 (on a static ip)
Remote Client: 192.168.1.95
Once the tunnel is established, the client can see all devices EXCEPT for the SCO box. When I veiw the VPN status log on the router to see what devices are available, I see all the Windows PC's, but not the Unix machine.
What is it about SCO's TCP/IP and/or ICMP responses that prevents any off-network machine from seeing it? All the local PC's on the same network subnet (192.168.0.x) can see it. But a remote machine on a different subnet can not.
I've added the IP's to the hosts file for both the router and the remote PC and I've performed a route add to point SCO to the router
route add -net 192.168.1.95 192.168.0.1
and the routing table reflects the entry.
Does SCO do something funky in the kernel parameters that could prevent this type of recognition?
I'm totally tapped out with no where else to search, have been looking for more than two days. Can I also say that this is not unique to a particular client. I have experienced this on several occasions with different clients. Some have gone to the extreme of adding a Linux box attached to the SCO box, then they can telnet to the Linux machine (because it can be recognized) and then step across to SCO. Not a good solution, but it does work.
Any new light would be greatly appreciated.
I'll be glad to provide more details, just tell me what you need.
Thanks
Joel
piperent