Hello all,

I have a full mesh VPN configuration with (7) total sites. I am having
problems with the (2) newest sites. These (2) sites are both NetWare 6.5
with BorderManager 3.8 sp2. These (2) sites cannot communicate directly
with any other slave site (including each other). So, both of these sites
can communicate with the master site (any host at the master site), but
they cannot cummincate with any other slave site.

All static routes are built with the internal network of each slave site,
and a next hop of the tunnel address for the server at the respective
slave site (ex: All other slave sites are able
to communicate directly between each other, though not with either of
these (2) problematic sites.

I have verfied that IPX flows between these (2) sites and all other slave
sites. A "display servers" shows the Trees and services at all slave
sites. I can also ipxping between these (2) sites and all other sites.
My problem is that users need to be able to map drives from these (2)sites
to other slave sites, and it fails every attempt.

Running traces shows the ip ping attempts make it as far as the master
server, and then begin to drop (why are they even routing through the
master in a full mesh scenario?).

Does anyone have any thoughts as to why these (2) newest servers cannot
communicate directly with other slave sites in a full mesh design?

Thank you in advance for your time and assistance.

Aaron DiBari