K
Q

Setup kubernetes using ansible

9/5/2019

I am setting up a kubernetes setup using ansible. To setup master i have written a playbook as shown below.

- hosts: master
  become: yes
  tasks:
    - name: initialize the cluster
      shell: kubeadm init --pod-network-cidr=10.244.0.0/16 

    - name: create .kube directory
      become: yes
      become_user: ubuntu
      file:
        path: $HOME/.kube
        state: directory
        mode: 0755

    - name: copy admin.conf to user's kube config
      copy:
        src: /etc/kubernetes/admin.conf
        dest: /home/ubuntu/.kube/config
        remote_src: yes
        owner: ubuntu

    - name: install Pod network
      become: yes
      become_user: ubuntu
      shell: kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/a70459be0084506e4ec919aa1c114638878db11b/Documentation/kube-flannel.yml >> pod_network_setup.txt
      args:
        chdir: $HOME
        creates: pod_network_setup.txt

The problem is that when i run the playbook, it doesn't wait for the initialization to complete i.e. for command "kubeadm init " to return and runs each task one after another. Since, the initialization takes time and one it is completed then only the file "/etc/kubernetes/admin.conf" is created. As ansible doesn't wait for it to complete it exists giving error in task 3 that "/etc/kubernetes/admin.conf" is not found.

If i run following playbook then ansible waits till the initialization is complete i.e. the control hangs till "kubeadm init " returns.

- hosts: master
  become: yes
  tasks:
    - name: initialize the cluster
      shell: kubeadm init --pod-network-cidr=10.244.0.0/16

How can i make ansible wait for command "kubeadm init " to complete and then only start another command ?

-- Prakhar
ansible
kubeadm
kubernetes

Similar Questions

Rollout deployment of akka cluster in kubernetes with a failing (old) pod: cluster never created
Using initcontainers in a job to do some stuff after pod initialization
Azure Kubernetes Containers to other internal vlans?
Automaticaly execute script inside Kuebrentes pod A when i update pod B with ansible

2 Answers

1/16/2020

When I use command module instead of shell module it appears to wait as it has no trouble making the copy of admin.conf. Also my next step after creating the .kube/config is to apply flannel overlay and that also works. Here's what my tasks looks like.

  - name: Initialize the Kubernetes cluster using kubeadm
    command: kubeadm init --config /etc/kubernetes/kubeadminit.yaml

  - name: create .kube in root home
    file:
      path: /root/.kube
      state: directory

  - name: copy kubernetes admin.conf to root home dir
    copy:
      src: /etc/kubernetes/admin.conf
      dest: /root/.kube/config
      remote_src: yes
-- Levi Silvertab
Source: StackOverflow
9/5/2019

There are three ways:

  1. wait_for some time after executing kubeadm init ansible task.
- hosts: master
  become: yes
  tasks:
    - name: initialize the cluster
      shell: kubeadm init --pod-network-cidr=10.244.0.0/16 
    - name: sleep for 20 seconds
      wait_for:
        timeout: 20
  1. Make use of until in ansible.
- hosts: master
  become: yes
  tasks:
    - name: initialize the cluster
      shell: kubeadm init --pod-network-cidr=10.244.0.0/16
      register: result
      until: result.stdout.find("Your Kubernetes master has initialized successfully!") != -1
      retries: 1
      delay: 20

NOTE: Here we retry kubeadm init until we get the string Your Kubernetes master has initialized successfully! in the output.

  1. Check if the file /etc/kubernetes/admin.conf exists after executing kubeadm init ansible task.
- hosts: master
  become: yes
  tasks:
    - name: initialize the cluster
      shell: kubeadm init --pod-network-cidr=10.244.0.0/16
    - name: create .kube directory
      become: yes
      become_user: ubuntu
      file:
        path: $HOME/.kube
        state: directory
        mode: 0755
    - name: Check admin.conf file exists.
      stat:
        path: /etc/kubernetes/admin.conf
      register: k8s_conf
    - name: copy admin.conf to user's kube config
      copy:
        src: /etc/kubernetes/admin.conf
        dest: /home/ubuntu/.kube/config
        remote_src: yes
        owner: ubuntu
      when: k8s_conf.stat.exists

NOTE: Here we execute admin.conf copy only when the k8s config file /etc/kubernetes/admin.conf exists.

Hope this helps.

-- mchawre
Source: StackOverflow