Target-Based Defragmentation Policies
A host's operating system uses three criteria to determine which packet fragments to favor when reassembling the packet:
-
the order in which the fragment was received by the operating system
-
its offset (the fragment's distance, in bytes, from the beginning of the packet)
-
its beginning and ending position compared to overlap fragments.
Although every operating system uses these criteria, different operating systems favor different fragments when reassembling fragmented packets. Therefore, two hosts with different operating systems on your network could reassemble the same overlapping fragments in entirely different ways.
An attacker, aware of the operating system of one of your hosts, could attempt to evade detection and exploit that host by sending malicious content hidden in overlapping packet fragments. This packet, when reassembled and inspected, seems innocuous, but when reassembled by the target host, contains a malicious exploit. However, if you configure the IP defragmentation preprocessor to be aware of the operating systems running on your monitored network segment, it will reassemble the fragments the same way that the target host does, allowing it to identify the attack.