Automate Routine Tasks With an Ad-Hoc Ansible Script
No one wants to spend an entire day running a manual script on a thousand servers. Fortunately, many processes like this can be automated using Ansible.
For example, imagine you are given a task to run a command to detect the Timezone on 1,000 servers. A quality engineer detected that some transactions are dated in the future, and the engineers suspected that some servers might have incorrect time zones. You are given a file that contains the IP addresses of 1,000 servers.
I will walk through completing this task by running an ad-hoc script to dynamically generate an inventory file using Ansible. All of this code is available on GitHub, and this tutorial assumes you already have an understanding of Python, Ansible and Bash.
Step 1
Establish a passwordless SSH connection with the target host. This will enable Ansible to securely log into the target host without having to input the password.
Step 2
Add the IP address to the file servers.txt and ensure the IP address is valid and follows the format in the servers.txt file.
Step 3
Extract the server’s IP address using Python to dynamically generate the inventory file:
| #!/usr/bin/env python3 | |
| import re | |
| import json | |
| import os | |
| # Get the current directory | |
| current_directory = os.getcwd() | |
| # Concatenate the current directory with the file name | |
| server_file = os.path.join(current_directory, 'servers.txt') | |
| def read_servers_file(server_file): | |
| """Reads the server file and extracts the IP addresses.""" | |
| ips = [] | |
| with open(server_file, 'r') as f: | |
| lines = f.readlines() | |
| for line in lines: | |
| if 'ip_address=' in line: | |
| match = re.search(r'ip_address=([\d\.]+)', line) | |
| if match: | |
| ips.append(match.group(1)) | |
| return ips | |
| def generate_inventory(ips): | |
| """Generates the inventory in JSON format.""" | |
| inventory = { | |
| '_meta': { | |
| 'hostvars': {} | |
| }, | |
| 'all': { | |
| 'hosts': ips | |
| } | |
| } | |
| return inventory | |
| def main(): | |
| """Main function.""" | |
| ips = read_servers_file(server_file) | |
| inventory = generate_inventory(ips) | |
| print(json.dumps(inventory)) | |
| if __name__ == '__main__': | |
| main() |
Step 4
Save the following Ansible playbook as ansible-playbook.yml:
| --- | |
| - name: Extract ctlplane IP addresses and run command on servers | |
| hosts: all | |
| gather_facts: yes | |
| become: yes | |
| remote_user: ec2-user #change this to the remote user | |
| tasks: | |
| - name: Run command on servers and save output locally | |
| ansible.builtin.shell: "date" | |
| register: command_output | |
| run_once: yes | |
| - name: Debug command output | |
| ansible.builtin.debug: | |
| msg: "{{ command_output.stdout }}" | |
| - name: Create your local file on master node | |
| ansible.builtin.file: | |
| path: "report.txt" | |
| state: touch | |
| mode: '0644' | |
| delegate_to: localhost | |
| become: no | |
| - name: Create report.txt file or append to existing file | |
| ansible.builtin.lineinfile: | |
| path: "report.txt" | |
| line: "{{ item }} - {{ command_output.stdout }}" | |
| loop: "{{ ansible_play_batch }}" | |
| delegate_to: localhost | |
| become: no |
Step 5
Run the Ansible playbook, which will run a command date on the target servers and display the output in a file called report.txt , with:
| ansible-playbook -i dynamic_inventory.py ansible-playbook.yml |
Step 6
Your output should look similar to this screenshot.
Conclusion
Ansible simplifies complex tasks, such as instance infrastructure provisioning, configuration management and software deployment across a large-scale environment. You can find the full code for this tutorial on GitHub.
