Pacemaker and Corosync cluster ansible role

Ansible role to install and configure a cluster with Corosync and Pacemaker

Introduction

Pacemaker and Corosync is one of the mostly used high availability cluster stack. Pacemaker/Corosync Configuration System on the other hand is a tool to easily configure both Pacemaker and Corosync.

Ansible

This role was tested against Ansible version 2.8 2.9 2.10 . The supported platforms are

• Ubuntu
  • focal
  • bionic

Mixing different distros might end up with different package versions being installed. It is strongly recommended to use a single version across all cluster. For example focal comes with corosync 3.0, while bionic with version 2.4

Variables

Configuration example

group_vars:
  all:
    corosync_hacluster_password: 1q2w3e4r5t
    corosync_cluster_settings:
      - key: stonith-enabled
        value: "false"
      - key: no-quorum-policy
        value: ignore
      - key: start-failure-is-fatal
        value: "false"
      - key: symmetric-cluster
        value: "false"
    corosync_cluster_defaults:
      - key: resource-stickiness
        value: 100

If you want to use an internal network

group_vars:
  all:
    corosync_use_internal_ip: true
host_vars:
  node1:
    internal_ip: 10.0.0.1
  node2:
    internal_ip: 10.0.0.2
  node3:
    internal_ip: 10.0.0.3

Multi-site cluster configuration

all:
  children:
    newyork:
      vars:
        totem_cluster_name: 'NewYork'
        hacluster_password: 'randomgeneratedpassword'
        ansible_connection: local
        booth_type: 'site'
        booth_ip: '10.254.1.15'
        booth_config: 'setup'
      hosts:
        ny01:
          ansible_host: '10.254.1.15'
          ansible_hostname: 'ny01'
    london:
      vars:
        totem_cluster_name: 'London'
        hacluster_password: 'randomgeneratedpassword'
        ansible_connection: ssh
        booth_type: site
        booth_ip: '10.254.2.15'
        booth_config: 'pull'
      hosts:
        ld01:
          ansible_host: '10.254.2.8'
          ansible_hostname: 'ld01'
        ld02:
          ansible_host: '10.254.2.9'
          ansible_hostname: 'ld02'
    munich:
      vars:
        totem_cluster_name: 'Munich'
        hacluster_password: 'randomgeneratedpassword'
        ansible_connection: ssh
        booth_type: arbitrator
        booth_ip: '10.254.3.15'
        booth_config: 'pull'
      hosts:
        muc01:
          ansible_host: '10.254.3.9'
          ansible_hostname: 'muc01'
        muc02:
          ansible_host: '10.254.3.10'
          ansible_hostname: 'muc01'
        muc03:
          ansible_host: '10.254.3.11'
          ansible_hostname: 'muc01'

The cluster site with booth_config 'setup' generates the booth configuration. Subsequently the other cluster sites will pull the booth configuration. There should be only one cluster site configured as 'setup' and all others as 'pull' When deploying is required to prioritize the booth_config 'setup' site first to ensure configuration is performed on the 'pull' sites accordingly.

Testing

Molecule with docker is being used.

Running the tests:

pipenv install
pipenv run molecule test