Project

General

Profile

Writing Definitions » History » Version 2

Version 1 (Luke Murphey, 04/02/2010 06:48 PM) → Version 2/14 (Luke Murphey, 04/02/2010 07:46 PM)

h1. Writing Definitions

h2. Definition Types

NSIA supports two types of definitions:

* ThreatPattern
* ThreatScript

|ThreatScript |ThreatPattern|
|Written in ECMAScript / JavaScript |Written in a format similar to Snort|
|Somewhat complex to create |Simple and easy to create|
|Can auto-baseline tune itself |Cannot baseline itself (can only be disabled entirely)|
|Is stateful (remember things from previous scans) |Is stateless (cannot remember things from previous scans)|
|Are slower than ThreatSignatures |Are faster than ThreatScripts|
|Very flexible detection logic; can be used to detect nearly anything |Functionality limited to what regular expressions|

h2. General Notes Before You Start

h3. Definition IDs

Custom definitions must have an ID of 1000000 or more; only official definitions can have IDs of less than 1000000.

h3. Identifying Definition Errors

NSIA will parse definitions before they are saved in order to identify syntax and some semantic errors. Errors that are discovered during runtime are noted on the definition errors page (e.g. http://127.0.0.1:8080/Definitions/Errors) and in the event logs (e.g. http://127.0.0.1:8080/System/Eventlog).

ThreatScript definitions may be flagged as having an error if they fail to complete within 10 seconds (see ScriptDefinition.MAX_SCRIPT_RUNTIME).


Creating a ThreatSignature

Creating a ThreatScript