k3s agent stuck on HA env

[alfredodeluca]

Expected Behavior

On HA env I would expect 2 master e 2 node running smoothly.

Current Behavior

it seems that 2 worker nodes are stuck with the following logs entries
Oct 15 12:46:43 ubumba01 k3s[1596]: time="2024-10-15T12:46:43+02:00" level=error msg="failed to get CA certs: Get "https://127.0.0.1:6444/cacerts\": read tcp 127.0.0.1:43760->127.0.0.1:6444: read: connection reset by peer"

Steps to Reproduce

1. 2 master in HA
2. 2 workers nodes

Context (variables)

Operating system: mix of Debian and Ubuntu

Hardware: mix of laptop

Variables Used

all.yml

k3s_version: ""
ansible_user: NA
systemd_dir: ""

flannel_iface: ""

#calico_iface: ""
calico_ebpf: ""
calico_cidr: ""
calico_tag: ""

apiserver_endpoint: ""

k3s_token: "NA"

extra_server_args: ""
extra_agent_args: ""

kube_vip_tag_version: ""

kube_vip_cloud_provider_tag_version: ""
kube_vip_lb_ip_range: ""

metal_lb_speaker_tag_version: ""
metal_lb_controller_tag_version: ""

metal_lb_ip_range: ""

Hosts

host.ini

master]
10.116.1.22 
10.116.1.233

[node]
10.116.1.189 
10.116.1.27  

[k3s_cluster:children]
master
node

Possible Solution

• I've checked the General Troubleshooting Guide