Skip to content

[CP 1471] feat(rbac): grant metrics-exporter events:create for profiler-disable…#561

Merged
sajmera-pensando merged 1 commit into
ROCm:mainfrom
ci-penbot-01:CP.O2O.pensando.gpu-operator.1471.rocm.gpu-operator.main
May 28, 2026
Merged

[CP 1471] feat(rbac): grant metrics-exporter events:create for profiler-disable…#561
sajmera-pensando merged 1 commit into
ROCm:mainfrom
ci-penbot-01:CP.O2O.pensando.gpu-operator.1471.rocm.gpu-operator.main

Conversation

@ci-penbot-01
Copy link
Copy Markdown
Contributor

@ci-penbot-01 ci-penbot-01 commented May 28, 2026

cp of pensando/gpu-operator#1471

Source PR Description (pensando/gpu-operator#1471):

…d K8s warnings

Add events:create to the metrics-exporter ClusterRole so DME can emit K8s Warning events when the GPU profiler is auto-disabled (KUBE-16).

Mirrors DME PR pensando/device-metrics-exporter#1306.

$ sudo k3s kubectl get pod default-metrics-exporter-gqpb9 -n kube-amd-gpu \
    -o jsonpath='{.spec.serviceAccountName}'      
amd-gpu-operator-metrics-exporter
$   sudo k3s kubectl get events -n kube-amd-gpu --field-selector reason=ProfilerDisabled
LAST SEEN   TYPE      REASON             OBJECT                               MESSAGE
4m27s       Warning   ProfilerDisabled   pod/default-metrics-exporter-gqpb9   GPU profiler metrics (gpu_prof_*) disabled: rocpctl process core dumped/aborted. Restart the pod to re-enable.

Cherrypick triggered by: ACP-Automation

@bhatturu @ci-penbot-01
feat(rbac): grant metrics-exporter events:create for profiler-disable…
c7e1792 
…d K8s warnings (#1471)

Add events:create to the metrics-exporter ClusterRole across all install
paths so DME can emit K8s Warning events when the GPU profiler is
auto-disabled (KUBE-16).

Updated locations:
- helm-charts-k8s/templates/metrics-exporter-rbac.yaml (Helm/K8s)
- hack/k8s-patch/template-patch/metrics-exporter-rbac.yaml (Helm patch source)
- config/rbac/metrics_exporter_cluster_role.yaml (kustomize/OLM source)
- bundle/manifests/amd-gpu-operator.clusterserviceversion.yaml (OLM CSV)

Mirrors DME PR pensando/device-metrics-exporter#1306.

(cherry picked from commit f1575d751b7675360c9728e44b86d3ec0a8aec55)
spraveenio
spraveenio approved these changes May 28, 2026
View reviewed changes
Copy link
Copy Markdown
Contributor

@spraveenio spraveenio left a comment

Choose a reason for hiding this comment

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

LGTM

@sajmera-pensando sajmera-pensando merged commit cbe6ed9 into ROCm:main May 28, 2026
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@spraveenio spraveenio spraveenio approved these changes

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@ci-penbot-01 @spraveenio @sajmera-pensando @bhatturu