[Release-1.24] - GO 1.20.6 breaks cri-dockerd

[brandond]

Backport fix for GO 1.20.6 breaks cri-dockerd

• GO 1.20.6 breaks cri-dockerd #8089

[mdrahman-suse]

Validated on release-1.24 branch with commit 9fc75e7

Environment Details

Infrastructure

• Cloud
• Hosted

Node(s) CPU architecture, OS, and Version:

PRETTY_NAME="Ubuntu 22.04.2 LTS"

Cluster Configuration:

1 server

Config.yaml:

write-kubeconfig-mode: 644
cluster-init: true
docker: true

Additional files

• nginx-deployment.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
spec:
  selector:
    matchLabels:
      app: nginx
  replicas: 2 # tells deployment to run 2 pods matching the template
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.14.2
        ports:
        - containerPort: 80

Testing Steps

1. Copy config.yaml

$ sudo mkdir -p /etc/rancher/k3s && sudo cp config.yaml /etc/rancher/k3s

2. Install k3s
3. Ensure k3s is up and running
4. Deploy nginx-deployment.yaml
5. Check cri-dockerd version
6. Check deployments

$ kubectl get pods -A | grep nginx
default       nginx-deployment-6595874d85-kk8x6         1/1     Running     0          4m
default       nginx-deployment-6595874d85-vf78c         1/1     Running     0          4m

Replication Results:

• k3s version used for replication:

k3s version v1.24.16+k3s1 (ccae12ed)
go version go1.20.6

• cri-dockerd version from logs

cri-dockerd version v0.3.2-0.20230123224751-84abd1adca2f (HEAD)

• Execute command in Pod's container via kubectl exec, observed the error http: invalid Host header

$ kubectl exec -it -n default nginx-deployment-6595874d85-kk8x6 -- bash
error: Internal error occurred: error executing command in container: http: invalid Host header

Validation Results:

• k3s version used for validation:

k3s version v1.24.16+k3s-9fc75e74 (9fc75e74)
go version go1.20.6

• cri-dockerd version from logs

cri-dockerd version v0.3.2-0.20230123224751-84abd1adca2f (HEAD)

• Execute command in Pod's container via kubectl exec, observed no error

$ kubectl exec -it -n default nginx-deployment-6595874d85-kk8x6 -- bash
root@nginx-deployment-6595874d85-kk8x6:/# 

Additional context / logs:

• Tested on new install and upgrades