Protect Arrikto EKF Pods¶
This guide describes the necessary steps to patch an existing Rok cluster on Kubernetes, in order to protect essential Rok External Services Pods and the Arrikto EKF Pods from being terminated in case of a CPU pressure scenario.
Overview
What You’ll Need¶
- A configured management environment.
- An existing Kubernetes cluster.
- A working Rok deployment.
- A working Kubeflow deployment.
Procedure¶
Run the rok-protect script that will patch the Pods with the appropriate CPU
requests:
root@rok-tools:~# rok-protect
Verify¶
Ensure that all Pods in the cluster are up and running:
root@rok-tools:~# kubectl get pods -A
NAMESPACE NAME READY STATUS RESTARTS AGE
auth dex-7c9b56d8f-whmjn 1/1 Running 0 2h
cert-manager cert-manager-cainjector-c5cc9b5c6 1/1 Running 0 2h
cert-manager cert-manager-dfcd64965-29v2g 1/1 Running 0 2h
...
Summary¶
You have successfully patched all Rok External Services Pods and all Arrikto EKF Pods with the appropriate Kubernetes CPU requests and have protected them against CPU starvation under CPU intensive scenarios.
What’s Next¶
Check out the rest of the maintenance operations that you can perform on your cluster.