Project

General

Profile

Actions

Bug #4941

closed

alerts: 5.0.8/6.0.4 count noalert sigs towards built-in alert limit

Added by Victor Julien almost 3 years ago. Updated about 2 years ago.

Status:
Closed
Priority:
Normal
Target version:
Affected Versions:
Effort:
Difficulty:
Label:
Needs backport to 5.0, Needs backport to 6.0

Description

Changes in alerting in 5.0.8/6.0.4 store noalert sigs in the packet alert array before removing them when finalizing the alerts. This solved several issues (#4663, #4670), however it introduces a new issue.

When many noalert rules are used, for example for flowbit "setter" logic, these rules now consume space in the alert array, leaving less space for "real" alerts that should be outputted. Since there is a built-in limit of 15 (see #4207) its not hard to reach this limit.


Subtasks 4 (0 open4 closed)

Optimization #4207: Use configurable or more dynamic @ PACKET_ALERT_MAX@ClosedJuliana Fajardini ReichowActions
Optimization #5178: detect/alert: improve packet alert queue handlingRejectedJuliana Fajardini ReichowActions
Task #4942: alerts: SV test for noalert issueClosedShivani BhardwajActions
Optimization #4943: alerts: use alert queing in DetectEngineThreadCtxClosedJuliana Fajardini ReichowActions

Related issues 4 (0 open4 closed)

Related to Suricata - Bug #4663: rules: drop rules with noalert not fully droppingClosedVictor JulienActions
Related to Suricata - Bug #4670: rules: mix of drop and pass rules issuesClosedVictor JulienActions
Copied to Suricata - Bug #5120: alerts: 5.0.8/6.0.4 count noalert sigs towards built-in alert limit (6.0.x backport)ClosedJuliana Fajardini ReichowActions
Copied to Suricata - Bug #5124: alerts: 5.0.8/6.0.4 count noalert sigs towards built-in alert limit (5.0.x backport)ClosedJuliana Fajardini ReichowActions
Actions

Also available in: Atom PDF