I
(think that I...) would like to set up
one OS/X machine to act as
two different OpenVPN servers at the same time:
- On the standard TCP port, for use by "road warriors," I would set up a bridge so that warriors can connect to the office network just as though they were really in the office. I think I would use TAP "bridging" for this.
- Meanwhile, we have a MySQL database-replication server, which will need to reach a remote server which should not see any part of the internal network. Therefore, I think that I need to set up a separate TUN. I understand that, to do this, I must allocate a second (non-standard) port on the server for a second instance of OpenVPN to be listening to.
Is this simply a matter of defining two server configurations (that is to say, "we are a server") with different configuration particulars such as TUN vs. TAP and port-number? Will TunnelBlick start and maintain both of these, if I ask it to?
To elaborate on the second case: I'd tunnel between, say, port "10.0.111.111" on the server and "10.0.222.222" on the remote. The MySQL remote server (master) would open an OpenVPN connection to us, and thereby be able to communicate through its port 10.0.222.222. Meanwhile, the replication server (slave) would connect to its side of the pipe, 10.0.111.111. The traffic between the two SQL servers would therefore be handled by the second OpenVPN, but managed by one always-running instance of TunnelBlick.
Will this work?