Skip to content

OKE_NPD_DEPLOY.md

Latest commit

 

History

History
39 lines (24 loc) · 1.59 KB

OKE_NPD_DEPLOY.md

File metadata and controls

39 lines (24 loc) · 1.59 KB

Installing OKE Node Problem Detector (NPD) DaemonSet with OCI GPU Scanner Service

OKE NPD is an extension of https://github.com/kubernetes/node-problem-detector that processes GPU health check failures reported by GPU Scanner service and sets conditions on the affected nodes. This feature enables proactive monitoring of GPU node health and early detection of issues.

Install

These actions should only be performed after a successful OCI GPU scanner service installation of control plane and data plane (plugin) components on individual GPU nodes.

❗❗IMPORTANT: NPD will only start processing GPU health check events when OCI GPU Scanner Service data plane plugin is actively running.

Label the target GPU nodes so they can host the NPD DaemonSet.

Only run this command on the GPU nodes that you would like to run the NPD feature on.

kubectl label node <nodeIP e.g 10.0.65.72> oci.oraclecloud.com/oke-node-problem-detector-enabled="true"

Install the NPD DaemonSet.

kubectl apply -f https://github.com/oracle-quickstart/oci-gpu-scanner/blob/main/existing_cluster_deploy/oke-node-problem-detector.yaml

Verify that NPD DaemonSet has been installed successfully and running.

kubectl get pods -l app=oke-node-problem-detector -o wide -n kube-system

Results should show oke-node-problem-detector in running state for all targeted GPU nodes.

Uninstall

To remove the NPD from an OKE cluster, run the below command

kubectl delete -f https://github.com/oracle-quickstart/oci-gpu-scanner/blob/main/existing_cluster_deploy/oke-node-problem-detector.yaml