asked on

Proxy and Tunneling

I assume the following is true:

I would be ill advised to tunnel (as in SSL) at the application level, but proxy in the DMZ

... because if you tunnel at the app level, as the packets leave the app to the DMZ where the proxy is, the proxy can not truly collect the packets into a message, and based upon the content of that message make decisions, like blocking IP's from going to certain destinations, or in the case of incoming messages, blocking IP's from coming in.

Therefore you either want to proxy and tunnel at the DMZ level, or proxy and tunnel at the app level, but not proxy at the DMZ, and tunnel at the app level

Is this correct?

Rich Rumble

You seem to be very interested in proxies and DMZ's... I think your definition of DMZ may be incorrect however. The DMZ is an isolated Network, not a layer in the TCP/IP or OSI stacks, like Application, Link, and Transport etc...
Let's put it this way, a http proxy that scans the traffic for malicious content cannot scan httpS connections, UNLESS the proxy injects it's own SSL certificate between the Client and itself. Effectively the proxy is translating between the two, the end website and the client.
SSL traffic operates at two different layers of the two standard stacks
The SSL/TLS library operates above the transport layer (uses TCP) but below application protocols. In the OSI model it's at layer 6 (presentation).
In both however you can fully view/read the IP information so decisions can still be based on DNS or IP addresses, but not the content unless that content is intercepted and then passed on.
-rich

Anthony Lucia

ASKER

OK, a proxy cant filter out content unless it scans the data.

So why would a proxy NOT scan the data. IF it doesn't scan the data, wouldn't you simply replace it with port forwarding ?

ASKER CERTIFIED SOLUTION

Rich Rumble

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial