Of all the Devops tools I have learnt this year, Ansible is my favourite. The ability to connect to different servers and perform different tasks in a way you can expect is superb. I have installed different flavour Linux Virtual Machines alongside my Raspberry Pi's and it connects to each one and does its thing as I expect.

The idea of automating everything, having it act in an expected and repeatable way is superb. I have written bash scripts that try to do this in the past, but fall over at scale and repeatability. Ansible is great for this.

In the process of learning ansible I have an inventory with 1 control node and 4 machines

[x] localhost (debian 10) [Control node]
[x] virtual machine (centos 8)
[x] virtual machine (ubuntu 20)
[x] virtual machine (alpine linux 3.12)
[x] raspberry pi (raspbian)

With this as my personal cloud, i can connect with ansible via ssh and run a whole range of tasks including:

[x] install and configure a webserver, mysql and install a laravel github repo
[x] run a git pull to update the repo
[x] update the OS and software
[x] install a docker container on a server 
[x] check the status of the machines 

Ansible provides a number of Facts for a machine which can be seen with this command:

ansible localhost -m setup --tree /tmp/facts

it is a json file of around 600 lines of Facts including for my localhost machine:

    "ansible_facts":
    {
        "ansible_os_family": "Debian",
        "ansible_architecture": "x86_64",
        "ansible_memtotal_mb": 7417,

    }

There is also extensive details about hardrives, IP addresses, python versions, users and the computer in general.

Similarly, Ansible provides details of the services running on the machine:

ansible localhost -m service_facts --tree /tmp/services

it provides a 1000 line Json file, with details of the services running on the machine, here is an example of the result:

    "ansible_facts":
    {
        "services":
            "apache2":
            {
                "name": "apache2",
                "source": "sysv",
                "state": "running"
            },
            "apache2.service":
            {
                "name": "apache2.service",
                "source": "systemd",
                "state": "running",
                "status": "enabled"
            },
    }

Also available is a list of package facts

ansible localhost -m package_facts --tree /tmp/package

This resulted in a json file with over 20000 lines, with details of everything installed on the machine, here is a section from the result:

    "ansible_facts":
    {
        "packages":
        {
            "apache2": [
            {
                "arch": "amd64",
                "category": "httpd",
                "name": "apache2",
                "origin": "Debian",
                "source": "apt",
                "version": "2.4.46-1"
            }],
            "zeal": [
            {
                "arch": "amd64",
                "category": "doc",
                "name": "zeal",
                "origin": "",
                "source": "apt",
                "version": "1:0.6.1-1+b1"
            }],
        }
    }

This provides a great deal of info about the machine, from which you could manage the the machines with conditionals like this.

- hosts: localhost
  roles:
     - role: debian_stock_config
       when: ansible_facts['os_family'] == 'Debian'

One thing I really like about Ansible is it is pretty obvious what the code is doing, and logical in how it does it. All these commands in this first part are ad-hoc, as in they are single commands you run in the command line.

---

If you would like to contact me with this form on londinium.com, ilminster.net or via Twitter @andylondon