Critical Control 5: Boundary Defense

Critical Control 5: Boundary Defense
Controlling the flow of information is critical to properly protecting critical information. Systems and their respective data must be broken down into trust levels or classifications. Any connectivity between networks of different trust must be through a robust boundary defense that properly protects and controls information flow and limits exposure.
Key rules to always follow when designing a robust boundary defense
1) Any system that is accessible from the Internet must be on the DMZ and DMZ systems contain no sensitive data;
2) Any system with sensitive data must be on a private network with no direct access from the Internet;
3) The only way DMZ systems can communicate with private network systems is through a middleware proxy.  For additional protection add additional middleware tiers.
Combining control 5 with the previous controls, organizations need to ensure that any device that is directly on a boundary and exposed to the Internet or untrusted networks have proper protection and carefully guarded against attacks. Simple changes to a device can caused unexpected vulnerabilities to be introduced into the system and can lead to either direct compromise of the device, or compromise of a device behind the perimeter.
Key steps organizations should take to properly implement this control:
1) Make sure ALL connections go through a firewall or a robust perimeter;
2) Validate the ruleset on any filtering device to make sure it is implementing a principle (continue reading...)