I am creating a disaster recovery plan for our company. Currently I have 2
netware servers off-site. They are in our production tree in a different
subnet but have a read/write replica of the Master at HQ. We thought this
would be good because it will have a copy of the master and all the objects
in edirectory. The problem is when are network connection is severed for
testing purposes any changes I make in Nwadmin/Console one at are DR-site do
not get propagated. Is this because it can not talk to the master server in
the tree?

I have read the option below and was wondering if this would be the
recommended way instead? With this way you would have to have servers up in
a different tree. When the disaster happens would you need to put the server
in the same tree name the original was in. Then restore the Directory
structure with the users objects and rights.

Is there a way to keep the DR servers in our production tree and have it
with a read/write replica of our master. This way when the production
servers go down we will be able to continue on the DR-Site without having to
restore the NDS structure?

Also if the DR-servers are on their own subnet but are part of the replica
ring. How do I keep the workstation from our production environment from
trying to ping or attach to those servers.