Cilium 1.18 to 1.19/1.20 upgrade stuck at Init:CrashLoopBackOff: Working workaround
While upgrading Cilium from 1.18.16 to 1.19.10, the worker agent got stuck in Init:CrashLoopBackOff . The same issue occurred with 1.20.4.
The failing config init container was running cilium-dbg build-config and reporting:
Unable to contact k8s api-server Get "https://10.96.0.1:443/api/v1/namespaces/kube-system": dial tcp 10.96.0.1:443: i/o timeout
Interestingly, the worker could reach the API ClusterIP using curl , but Cilium couldn't.
What fixed it
First, check whether you're using CiliumNodeConfig :
kubectl get ciliumnodeconfigs.cilium.io -A
If you're not using per-node configuration overrides, add this to your existing Helm values:
daemon: configSources: "config-map:cilium-config"
Apply the configuration using Helm with your current target version:
helm upgrade cilium cilium/cilium \ --version 1.20.4 \ -n kube-system \ -f my-values.yaml
This removes the config init container and makes Cilium read its…