Rancher Resolved Issues

Rancher Problems and Solutions

Quickly Clean a Node

Nodes need to be cleaned completely, use the following script.

curl https://gist.githubusercontent.com/Ileriayo/1bef407602208911e86f42d5d208c1fb/raw/af8fa882add9c0a7ccd72b92f1cfab5c95c355ba/nuke_rancher_kube_node.sh | sh

Waiting for API [Readey False xxx mins ago [Disconnected] Cluster agent is not connected]

The newest version of rancher brought this error to me, I’m not sure about the cause.

Waiting for API [Readey False xxx mins ago [Disconnected] Cluster agent is not connected].

Solution

• What I used: changed Rancher to 2.5.15(2.5x stable version), choose 19.x Kubernetes
• Another method suggested by others (but not working for me)
  • https://github.com/rancher/rancher/issues/36589

Waiting for Kubernetes API to be available

Solution

• Setting up hosts for all nodes.

  • ```bash
    hostnamectl set-hostname node1

    - Writing to the `hosts` file  for all nodes
    
      - ```bash
        vi /etc/hosts
        #192.168.2.121 node1
        #192.168.2.122 node2
        #192.168.2.123 node3

• Shut down firewall, selinux, and swap

  • sudo systemctl stop firewalld
    sudo systemctl disable firewalld
    
    #shut selinux： 
    sed -i 's/enforcing/disabled/' /etc/selinux/config
    setenforce 0
    
    #shut swap：
    swapoff -a  
    sed -ri 's/.*swap.*/#&/' /etc/fstab 
    

Master node unavailable after shutting down a worker node

I encountered this problem when I used only two nodes for the Kubernetes cluster, which is only one master node(with worker installed) and one worker node

Solution

• Having one more worker node. When a worker node is unavailable, after 5 minutes, a new pod will be created on another available working node automatically.