fleetdm
VM Hosts with Duplicate UUIDs do not report correctly in Fleet
Fleet version: 4.49
Web browser and operating system: N/A
💥 Actual behavior
When a VM Host with a duplicate hardware UUID enrolls into Fleet, it assumes the role of the previously enrolled host, causing an enroll loop between all hosts with the same UUID. To the user this appears to be hosts not enrolling, or hosts disappearing.
🧑💻 Steps to reproduce
- Enroll 2 hosts both using the same hypervisor
Suggestion: Fleet to detect an abnormal amount of host enrollments from the same UUID. Fleet will tell these hosts to send an alternate identifier (osquery ID). This has a negative affect on macOS MDM, so limit the solution to Linux/Windows.
Currently, when enrolling hosts w/ the same UUID, the best practice is to create a fleetd agent w/ an alternative identifier.
it assumes the role of the previously enrolled host, causing an enroll loop between all hosts with the same UUID.
Hey @mostlikelee and @sharon-fdm, while confusing, this is the expected behavior.
I think this is a feature request and not a bug. Moving this to feature fest.
Let me know if you disagree!
We're improving this iteratively in the "Improve error message when hosts enroll with duplicate identifiers" story here: #16393
I think this issue and suggested solution could be a good next iteration.
Looking at this issue again, I think it meets the definition of an engineering initiated story.
Adding the ~engineering-initiated label and removing ~feature-fest
FYI @lukeheath