[Release-1.32] - rke2 ingress upgrade fails 1.31.4 -> 1.31.5

[brandond]

Backport fix for rke2 ingress upgrade fails 1.31.4 -> 1.31.5

• rke2 ingress upgrade fails 1.31.4 -> 1.31.5 #7652

[mdrahman-suse]

Validated using commit 411fdda on release-1.32 branch

Environment

Ubuntu 24.04, Single server

Testing steps:

• config.yaml

write-kubeconfig-mode: 644
node-name: server1

• Install rke2
• Ensure cluster is up
• Update /var/lib/rancher/rke2/server/manifests/rke2-ingress-nginx.yaml with defaultBackend.enabled: "true"
• Ensure helm-install-rke2-ingress-nginx-* pod is in Completed state

Replication

rke2 -v
rke2 version v1.32.1+rke2r1 (c0f7be4407cf2c437cacfe735e5c943e827f2ff8)
go version go1.23.4 X:boringcrypto

• Checked rke2-ingress-nginx charts version

$ helm list --all-namespaces | grep rke2-ingress
rke2-ingress-nginx          	kube-system	1       	2025-02-13 22:57:06.493496818 +0000 UTC	deployed	rke2-ingress-nginx-4.12.003         	1.12.0

• Updated rke2-ingress-nginx manifest file

$ sudo cat /var/lib/rancher/rke2/server/manifests/rke2-ingress-nginx.yaml | grep defaultBackend
    defaultBackend.enabled: "true"

• helm-install-rke2-ingress-nginx-* pod is in CrashLoopBackOff state

kubectl get pods -A | grep rke2-ingress
kube-system   helm-install-rke2-ingress-nginx-k9zxf                   0/1     CrashLoopBackOff   5 (2m20s ago)   5m22s
kube-system   rke2-ingress-nginx-controller-gb8tt                     1/1     Running            0               42m

Validation

rke2 -v
rke2 version v1.32.1+dev.411fdda2 (411fdda2fe224d9c751ecffbd78229c1340699d9)
go version go1.23.4 X:boringcrypto

• Checked rke2-ingress-nginx charts version

 helm list --all-namespaces | grep rke2-ingress
rke2-ingress-nginx          	kube-system	2       	2025-02-13 23:44:16.000399795 +0000 UTC	deployed	rke2-ingress-nginx-4.12.004         	1.12.0

• Updated rke2-ingress-nginx manifest file

$ sudo cat /var/lib/rancher/rke2/server/manifests/rke2-ingress-nginx.yaml | grep defaultBackend
    defaultBackend.enabled: "true"

• helm-install-rke2-ingress-nginx-* pod is in Completed state

kubectl get pods -A | grep rke2-ingress
kube-system   helm-install-rke2-ingress-nginx-cv66g                   0/1     Completed      0               68s
kube-system   rke2-ingress-nginx-controller-lr4d6                     1/1     Running        0               46s

NOTE: A new pod gets created with ErrImagePull state and it is expected as the image is invalid

kube-system   rke2-ingress-nginx-defaultbackend-6678546969-fjk4d      0/1     ErrImagePull   0               58s