FortiSIEM Discussions
bhinangt
New Contributor III

Detecting 3-Day Sustained NAC Breach Using Rule Logic in FortiSIEM

 

Overview:

In NAC-based environments, it's common to see compliance violations or host breaches reported daily. However, not all of them require immediate action—especially if they self-remediate. The real concern arises when a host remains non-compliant for consecutive days [Taking 3 days as an example].

 

In this post, I’ll walk you through how to design a correlation rule that ignores isolated Day-1 and Day- 2 alerts but fires only when the same host is non-compliant over 3 days—a common compliance logic for risk-aware environments.

 

 

Objective:

Trigger an alert only if a host remains in NAC breach for 3 continuous days. As Saturday and Sunday is Holiday- Thursday, Friday and Monday would be only trigger an alert.

 

Rule Design Philosophy:

  • Skip isolated or sporadic violations.

  • Only trigger if breach continues beyond 48 hours (i.e. on the third day).

  • Prevent noise from multiple Day-1 alerts or flapping.

 

Pattern Configuration:

 

Screenshot 2025-06-14 at 11.33.07 AM.png

Time Window:

Set a time window wide enough to allow detection across 3 days, e.g.:

259200 seconds = 3 days
You can adjust slightly based on event granularity and log delay.

 

Tip:

  • If your source system uses inconsistent naming (e.g. Day_1, Day_01, Day_001), make sure to include all formats in pattern logic.

  • Always test with historical logs to ensure detection is not too restrictive or too noisy.

 

Use Case Impact:

This rule is very useful in enterprise SOC environments where:

  • NAC breaches are frequent but mostly self-remediating

  • Security teams want to focus on persistent compliance failures

  • Helps prioritize response to meaningful threats instead of chasing false alarms

 

Final Note:

Feel free to copy the pattern logic into your own FortiSIEM. If you have different naming conventions, just update the subpattern values.

Let me know if you'd like a ready-to-import JSON or help tuning it to your log format.

Bhinang Tejani
Bhinang Tejani
0 REPLIES 0