Ansible Playbook for Deploy 3 Node Glusterfs Cluster

This playbook has been tested on Centos7

Prerequisites

• 3 vm or physical machine installed centos7 for gluster node
• The gluster node has at least 1 new disk that is separate from the disk that is installed os
• Gluster node connected to the internet
• The control node (it could be localhost) has been installed ansible
• The gluster node has been installed ssh server
• Control node can access gluster node though ssh without password

Installation

• Clone The Repository

$ git clone https://github.com/prastamaha/glusterfs-ansible.git
$ cd glusterfs-ansible

• Inventory Configuration

Inventory Hosts file available on inventory/hosts. Edit the file and match it to your server environment

The default value will look like this

$ cat inventory/hosts

[gluster]
server1 ansible_user=root
server2 ansible_user=root
server3 ansible_user=root

• Playbook Variable Configuration

Adjust the variables in vars/config.yml to your server environment

The default value will look like this

$ cat vars/config.yml

# server1
server1_ip_address: 10.1.1.55
server1_hostname: server1

#server2
server2_ip_address: 10.1.1.60
server2_hostname: server2

#server3
server3_ip_address: 10.1.1.70
server3_hostname: server3

# note: must use a new disk that has not been partitioned / os installed
# gluster disk 
disk_device: /dev/vdb
partition_size: 2GiB

# brick 
brick_path: /bricks/brick1
brick_fstype: ext4

# volume 
volume_name: vol0

# choose between ( replicated / distributed )
volume_type: distributed

• Run The Playbook

$ ansible-playbook glusterfs.yml