Hello
I updated the Calico CNI because there were issues where new Pods would not deploy or communication between nodes was cut off after returning from sleep mode.
■ Cause of the Problem
Calico uses short-term tokens for API server authentication and CNI configuration. Leaving the laptop in sleep mode causes the following two issues.
- Returning from sleep mode after more than 1 hour → calico-node fails API server authentication, unable to renew CNI token
- Returning from sleep mode after more than 24 hours → CNI token expired, unable to deploy new Pods
■ Action Taken
Upgraded Calico from v3.29.2 to v3.31.2 and applied a non-expiring Static Secret token.
Applied Courses:
- Kubernetes for Easy Start
- Kubernetes Learning Through Illustrations
- Prometheus Learning through Hands-on Practice
- Learning Grafana through Hands-on Practice
■ Test Status
- k8s_learning.kit : Completed (4 nodes Ready, no static token expiration)
- k8s_starter.kit : Completed (4 nodes Ready, No static token expiration)
- prom_learning.kit : Completed (4 nodes Ready, No static token expiration)
- graf_learning.kit : Completed (4 nodes Ready, No static token expiration)
■ For those currently using the existing environment
Method 1. Reconfigure the entire environment (Recommended)
vagrant destroy -f && vagrant up
Method 2. Re-apply Calico only (after connecting to cp-k8s)
kubectl delete -f https://raw.githubusercontent.com/sysnet4admin/IaC/main/k8s/CNI/calico-quay-v3.31.2.yaml
kubectl apply -f https://raw.githubusercontent.com/sysnet4admin/IaC/main/k8s/CNI/calico-quay-v3.31.2.yaml
For those setting up a new environment, it will be applied automatically without any additional action.
I hope this helps with your learning.
Best regards, Hoon Jo.