Saturday, September 12, 2009


After a while (some time of not using), the OpenVPN server broke down.
The secured connection istablishes, but IP does not go.

I guess, the problem in the IP routing. I don't know what I did, but after server rebooted, the connection start working.

The routing table on the server looks as below:

C:>route print 10.8*

IPv4 Route Table
Interface List
0x1 ........................... MS TCP Loopback interface
0x2 ...00 ff 3d 7b 61 e5 ...... TAP-Win32 Adapter V8
0x10004 ...00 13 72 69 2d 0c ...... Broadcom BCM5708C NetXtreme II GigE (NDIS VBD Client)
Active Routes:
Network Destination Netmask Gateway Interface Metric 30 1 30
Default Gateway:
Persistent Routes:

There are still questions for me. Why routing to via two gateways - and ? - is the address of the TAP IP interface, but what is then?
The address of the "modem pool"?

Same way, on the client the interface address, but - P-t-P - what's that?

Good instructions about setting OpenVPN up are located here:

And don't forget, when you want to connect a machine behind the server it has to have route to the network as well.


  1. new version of OpenVPN does not possess that problem, and no extra route add is necessary.

  2. unfortunately, but that's not true.
    I don't know why, but after a while I had to add the following route manually:

    route add mask if 0x2

  3. I had to start the RRAS service to make the routing work properly.