Discussion Locked

Remote Access from Private Networks

0 helpful
5 replies
Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.

Hello all.

I've got problem here that has me stumped, and I'm hoping someone out there has gone through this before and has a solution. It concerns remote access.

My main office is using private IP addresses, with a Cisco 2514 router doing Port Address Translation on a pair of global IPs.

I have static routes created for those servers who need a real identity in the outside world--my RAS server, for example.

Recently, we have been putting some regional people into executive office spaces. These places have their own networks that are beyond my control and invariably are also using private internal ranges. Problem is that everyone we put in one of these offices has trouble accessing certain things on our network.

For example,they cannot successfully connect to our NT4 RAS server (PPTP). Although they can ping it, they can't actually log on-- getting an error 650.

Additionally, things like NetMeeting (which we sometimes use for remote support) no longer work.

Run of the mill stuff, like e-mail, Web access, etc. always works fine.

I have tried to find out from the IT people who run these executive spaces what, if anything they are doing that would cause these problems, and the answer is always either "nothing" or "we don't know".

I cajoled one of these places into giving our people real IP addresses, and all of the problems went away. However, with this latest office we have, I do not think that will be an option.

Is anyone aware of any issues regarding private-to-private network communication that could be causing my woes? Any leads would be most appreciated.

5 replies

by stillatit

When you do NAT, all that the outside world sees is the address(es) which the router is using to provide access to your network. If your inside addresses are all non-routable, then the best a "static" address can do is to route all traffic coming tothe outside address on a particular port to a particular inside address. This typically works fine for web servers and mail servers.

The problem is the PPTP wants to "tunnel" to a particular server, and will insist on that server and no other. We solved this problem by having real IP addresses for our PPTP servers at our central site. At remote sites, rather than using a NAT router, we use an NT server running RRAS and Proxy servers. The remote site gets one real IP address, and that is theNT server. When any user at a remote site tries for the central site, RRAS opens a VPN connection (if not already open), and provides routing. If the user wants to browse the internet, Proxy provides the NAT services.

There is a good explanation of PPTP at
http://www.microsoft.com/technet/network/vpntwk.asp

by kasey

hmmm - take a look at the private address space you are using. now take a look at the networks that your remote users are using. i bet that some of them are using the same private networks. THEY CANNOT USE A NETWORK SPACE THAT IS THE SAME AS WHAT YOU HAVE IN YOUR NETWORK. IF THEY DO, THEIR TRAFFIC WILL NOT ROUTE VIA THE PPTP TUNNEL.

Now you can say - hey i am nat'ting - big deal. but you have to recognize the fact that you are using pptp. on the local machine, there are two addresses (loopback and the assigned ip) if a packet is destined for that addr space, then it will route to it. if not, it will follow the default route (bet that you are using the server supplied default which would be the PPTP connection.)

But you can get the pptp connection up and some apps work....right, until you try to access a device that is supposedly on your own network (but is actually over the PPTP tunnel).

This discussion is closed to new replies.