Skip to content

CSP setting continuously reapplying #38302

Description

@mason-buettner

customer-redqueen GONG snippet: https://us-65885.app.gong.io/call?id=4695135856071655463&highlights=%5B%7B%22type%22%3A%22SHARE%22%2C%22from%22%3A1637%2C%22to%22%3A1763%7D%5D

Fleet version: 4.78.3

Web browser and operating system: Windows 11 Education/Windows 11


💥  Actual behavior

When enrolled in Fleet for MDM, Windows custom settings are generating a high volume of 4719 event logs that indicate frequent changes. These logs trigger SIEM alerts and are flagged as IoCs.

customer-redqueen sees roughly 80 4719 events per minute/host with the changes Failure added and Failure removed.

I have previously observed similar behavior on https://dogfood.fleetdm.com/hosts/1489/

🛠️ To fix

Timeboxed to investigate what if anything differently we can do or document possible side effects of this CSP.

🧑‍💻  Steps to reproduce

These steps:

  • Have been confirmed to consistently lead to reproduction in multiple Fleet instances.
  • Describe the workflow that led to the error, but have not yet been reproduced in multiple Fleet instances.
  1. Enroll the Windows host to Fleet.
  2. Add custom setting (Audit policy CSP).
  3. Profile is stuck in "Verifying".
  4. In the Windows events, audit policy is constantly switching states. (see gong snippet on the top of the issue).

🕯️ More info (optional)

Slack thread with redqueen: https://fleetdm.slack.com/archives/C08UZKFR7LM/p1767897717326719

Metadata

Metadata

Assignees

Labels

#g-power-to-pcPower to the PC working group:releaseReady to write code. Scheduled in a release. See "Making changes" in handbook.bugSomething isn't working as documentedcustomer-redqueen~aging bugBug has been open more than 90 days~assisting qaThis issue can be QA'd by anyone outside the QA team when capacity allows

Type

No type

Fields

No fields configured for issues without a type.

Projects

Status
Done

Milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions