Treat board onboarding as a controlled configuration change. Validate the platform entry, the extracted compatible strings, and the scheduler rule together, then confirm the lava target actually matches the declared profile before relying on the test output.
What “governance” means for new KernelCI board onboarding
Board onboarding should be treated like a release-gated change, not a casual platform update. The important governance question is whether the new board is represented consistently across the platform entry, its compatible strings, and the scheduler logic, so that KernelCI tests the hardware profile you think it is testing.
That means the onboarding record needs to be internally consistent before the board is allowed into routine use. If the declared board profile and the underlying LAVA target diverge, the result is not just a bad test run, it is a false signal about kernel compatibility, regressions, and readiness.
For teams running board fleets, the practical rule is simple: every new board definition should be reviewable as a change package. The platform entry, device-tree or compatible identifiers, and execution target should all be checked together so that naming, matching, and scheduling reflect one controlled configuration.
Why configuration consistency matters more than the individual file change
The risk is usually not in any single field. A board can look valid in isolation while still being misrouted at runtime because the scheduler rule matches too broadly, the compatible string is incomplete, or the LAVA target name points to a different profile than the one described in the board entry.
When that happens, teams may trust test output that came from the wrong hardware class, wrong boot path, or wrong lab target. That can hide real regressions, create noisy failures, or make a board appear supported before the full boot and test path has actually been proven.
Governance should therefore focus on alignment checks, not just authorisation to add the board. The useful control is a simple one: if the platform metadata, selector logic, and lab target do not agree, the onboarding should remain in review until the mismatch is resolved.
How to make onboarding auditable and safe to rely on
Good onboarding has an evidence trail. Teams should be able to show which board profile was added, which compatible strings were accepted, which scheduler rule matched, and which LAVA target executed the test, so that a later failure can be traced back to the exact onboarding decision rather than to an opaque lab state.
That auditability also supports change review. A reviewer should be able to confirm that the new board is not reusing an older target name or inheriting a scheduler rule that was written for a different device family, because those shortcuts are where silent misclassification tends to creep in.
When the change is managed this way, onboarding becomes repeatable. The board definition is not “done” when it exists in the repository, it is done when the declared profile and the actual execution path have been shown to agree under test.
Risk and Threat Considerations
Board onboarding failures most often produce integrity risk rather than obvious outages. A misbound target or overly broad scheduler rule can make results look legitimate while quietly testing the wrong device, which weakens confidence in the entire validation pipeline.
Failure mechanism: The platform entry, compatible string matching, and scheduler selection diverge, so the board routes to an unintended LAVA target or executes under the wrong declared profile.
Impact: Teams may ship on the basis of false-positive compatibility data, miss real board-specific regressions, or spend time chasing failures that are actually caused by onboarding mismatch rather than kernel behaviour.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
NIST CSF 2.0 and NIST SP 800-53 Rev 5 set the technical controls, while ISO/IEC 27001:2022 defines the regulatory obligations.
| Framework | Control / Reference | Relevance |
|---|---|---|
| NIST CSF 2.0 | GV.PO-01 — Policy | Board onboarding is governed as a controlled change process. |
| Recommendation — Define onboarding policy and require consistent approval before new board profiles are used. | ||
| NIST SP 800-53 Rev 5 | CM-3 — Configuration Change Control | New board onboarding is a configuration change that needs formal review and approval. |
| CM-6 — Configuration Settings | The board profile, compatible strings, and scheduler rule are configuration settings that must align. | |
| AU-2 — Event Logging | Onboarding needs traceable evidence of which target and rule actually ran. | |
| Recommendation — Require review and approval for each board onboarding change before it reaches production tests. Standardise and verify board configuration settings so the declared profile matches execution. Log onboarding decisions and runtime target selection to support later audit and troubleshooting. | ||
| ISO/IEC 27001:2022 | A.8.9 — Configuration management | The subject is about controlled configuration updates and consistency checks. |
| Recommendation — Apply configuration management to board definitions, matching rules, and lab targets. | ||
Practitioner Guidance
What to verify: Review the board entry, compatible strings, scheduler rule, and LAVA target as one onboarding package, then confirm the selected target matches the declared profile before accepting results as authoritative.
Decision rule: If any part of the matching chain is ambiguous, treat the board as not yet governed, and keep it out of routine test reliance until the mapping is explicit and repeatable.
Common mistake: Teams often validate the YAML or metadata update but do not validate the runtime selection path. That is where the control breaks, because the board can be “configured” yet still be tested through the wrong execution route.
Practitioner takeaway: The safest onboarding model is not approval by existence, it is approval by observed match, the board should be trusted only after its declared identity and its executed target have been proven to be the same thing.
Related resources from NHI Mgmt Group
Deepen Your Knowledge
Free weekly newsletter
Subscribe to the NHI & AI Identity Journal
The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.
Bonus 33% off our NHI Course when you subscribe.
Reviewed and updated by the NHIMG editorial team on October 11, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org