Skip to content

Conversation

@rajatchopra
Copy link
Contributor

Description

If the device plugin restarts, it will attempt to update the existing nodefeatures object
Also need to update the nodefeatures object for the 'gpu.count' label

Checklist

  • No secrets, sensitive information, or unrelated changes
  • Lint checks passing (make lint)
  • Generated assets in-sync (make validate-generated-assets)
  • Go mod artifacts in-sync (make validate-modules)
  • Test cases are added for new code paths

Testing

Manual testing

- get
- list
- watch
- update
Copy link
Contributor

@tariq1890 tariq1890 Feb 9, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Are you able to patch instead to update your NFD CRs ? update in general is a riskier operation as it is more sensitive to race conditions.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

No. The GFD seems to use the Update call:

E0210 20:21:14.746771       1 main.go:127] failed to update NodeFeature object "nvidia-features-for-dellr760tdx": nodefeatures.nfd.k8s-sigs.io "nvidia-features-for-dellr760tdx" is forbidden: User "system:serviceaccount:gpu-operator:nvidia-sandbox-device-plugin" cannot update resource "nodefeatures" in API group "nfd.k8s-sigs.io" in the namespace "gpu-operator"

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Got it. I meant updating the code to use patch instead of update

@cdesiniotis cdesiniotis merged commit 8548d88 into NVIDIA:main Feb 13, 2026
16 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants