Mega-lab - Part 3: Using Ansible to gather information from hosts and save to files

In this post, we will set up an Ansible playbook to collect some information from hosts and save the outputs to different files depending on the hostname. The following steps show how to setup and run an Ansible playbook:

• Creating the inventory file
• Modifying the ansible.cfg
• Writing the playbook

1. The ansible folder structure

We will need to create the ansible folder according to the following structure:

.
├── ansible.cfg
├── gather_commands.yml
├── inventory
│   └── hosts.yml
└── output
    ├── R1.txt
    └── R2.txt

The ansible.cfg is as follows:

[defaults]
inventory = hosts.yml
host_key_checking = False
ANSIBLE_SSH_ARGS = -oKexAlgorithms=+diffie-hellman-group1-sha1 -oHostKeyAl
gathering = explicit
#interpreter_python = .venv/bin/python

[paramiko_connection]
#host_key_auto_add = True
look_for_keys = False
host_key_checking = False

[persistent_connection]

connect_timeout = 30
#connect_retry_timeout = 15
command_timeout = 30

2. Create the hosts inventory file

all:
  children:
    CORE1:
      children:
        Switch1:
          hosts:
            R1:
              ansible_host: 10.15.1.1
            R2:
              ansible_host: 10.15.1.2

  vars:
    ansible_user: cisco
    ansible_password: cisco
    ansible_connection: network_cli
    ansible_network_os: ios

We can check the connection to a groups of routers defined in the inventory file.

ansible all -i inventory -m ping

Output:

R1 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/home/doanh/.pyenv/shims/python3.7"
    },
    "changed": false,
    "ping": "pong"
}
R2 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/home/doanh/.pyenv/shims/python3.7"
    },
    "changed": false,
    "ping": "pong"
}

We can use the Ansible ad-hoc CLI tool to run a command on a remote host.

ansible -i inventory -m setup -u cisco R1

Here cisco is the user on the router R1.

The setup module, which is part of Ansible core, enables users to view facts gathered on remote hosts. By default, Ansible collects these facts when we use the Ansible tool to manage hosts.

Suppose that we want to run a command on dozens of routers. Instead of opening a terminal session to each router and scraping the output, we can use the Ansible ad-hoc CLI with group definition in an inventory file.

ansible  -i inventory -m command -a "df -k" -u cisco Switch1

3. Create the Ansible playbook

---
- name: Gather commands
  hosts: Switch1
  vars:
    commands:
      - show version
      - show ip int br
      - show memory statistics
      - show arp
      - show ip route
  tasks:
    - name: Gather commands
      ios_command:
        commands: "{{ commands }}"
      register: output

    - name: save output to a file
      copy:
        content: "{{output.stdout | zip(commands) |
        flatten | join('\n') }}"
        dest: "output/{{ inventory_hostname }}.txt"

4. Running the Ansible playbook

We run the Ansible playbook to gather the following information:

• show version
• show ip int br
• show memory statistics
• show arp
• show ip route

The command to run is as follow:

ansible-playbook -i inventory gather_commands.yml

Output:

PLAY [Gather commands] ************************************************************************************************************

TASK [Gather commands] ************************************************************************************************************
ok: [R1]
ok: [R2]

TASK [save output to a file] ******************************************************************************************************
changed: [R2]
changed: [R1]

PLAY RECAP ************************************************************************************************************************
R1                         : ok=2    changed=1    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0   
R2                         : ok=2    changed=1    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0 

The outputs file will be saved in the output folder: R1.txt and R2.txt.