Modern scanner architectures are built as inference based systems, not requiring any agent software on the target systems. They learn about each target system individually while selective probes are being exchanged with the target system. The inference based architecture is centred around highly multi-threaded engines for scanning thousands of vulnerabilities simultaneously on any system on a network. Modern scanner architectures also support multiple levels (trusted and un-trusted) of vulnerability assessment against any given target system. Un-trusted vulnerability assessments simulate the scenario of an attacker without prior knowledge about the target system, while trusted assessments leverage credentials to log into the target systems for auditing configuration and patch information. An important criterion for measuring the effectiveness of a vulnerability scanner is the comprehensiveness and accuracy of its vulnerability knowledge base. Also, the ability to report and communicate vulnerability findings in a standardized manner from the vulnerability scanner to other applications (i.e. patch distribution or configuration management) is a critical requirement.
The link for this article located at Security Park is no longer available.
