We have a temporary office (3 months) with it's own cable modem, Linksys
router WRT54GS, and 3 wireless laptops, all configured the same. They use
DHCP and NAT to share a public address, and they use a 192.168.115.x
address internally at the remote site, which is not used anywhere in the
local network. When they vpn in, it seems we never have problems when 2
machines are connected, but when the 3rd connects, one of the two existing
gets "knocked off"(and it can be any of the 3 machines). By that I mean
they still have a vpn connection, they just lose and cannot reconnect to
the citrix server at the local site (the purpose of the vpn connection),
nor can the disassociated machine ping anything inside the local network.
The vpn connection is never lost, and if you look at the VPN monitor in
nwadmin, you can see the 3 machines at the remote site still have valid
connections, each with their own 192.168.115.x ip address. if we log off
all three and start over, we may have a problem getting the 3rd connected
again, or maybe all 3 connect for a while, but eventually one will lose
it's connection. And it seems that we never have a problem with 2 losing

Was this never meant to work? Should it? We didn't want to go to the
expense or trouble of buying another copy of BM, Novell, and a server to
set up a site-to-site vpn for such a short period of time, and when we
tested with 2 laptops before we went live it was never a problem.

We're using BM 3.7 on our server. The VPN tunnel subnet is 192.168.1.x
The laptops are WIndows xp SP2 with the firewall enabled. we did make an
exception on the laptops for UDP 353 to start, or they would get a timeout

Thanks in advance...