In order to AI proof, System amdins must know of the confuration manger.I selected Anible.
Agenda for 14 days Ansible series
Days 1–2: Execution Engine Mastery
Must know cold
- Inventory (INI vs YAML, dynamic inventory)
- Execution order (hosts → roles → tasks → handlers)
- serial, strategy, forks
- check_mode, diff, –step
🔥 Practice:
ansible-playbook site.yml -vvv –check –diff
If you can’t explain why a task ran, you’re not advanced yet.
🗓️ Days 3–4: Variables & Precedence (MOST FAIL HERE)
Critical
- Variable precedence (memorize the order)
- group_vars, host_vars
- set_fact vs register
- vars_files, vars_prompt
🔥 Drill:
Override the same variable from 5 places and predict the final value.
🗓️ Days 5–6: Roles Like a Pro
Advanced role design
- defaults vs vars
- Role dependencies
- Reusable roles (no hardcoded values)
- include_role vs import_role
📁 Proper layout:
roles/
└── nginx/
├── defaults/
├── vars/
├── tasks/
├── handlers/
├── templates/
🎯 Interview killer question:
Why are variables in vars/ dangerous?
🗓️ Days 7–8: Error Handling & Control Flow
This is where senior engineers shine.
Must master
- failed_when
- changed_when
- ignore_errors
- block, rescue, always
- any_errors_fatal
- max_fail_percentage
🔥 Real scenario:
block:
– name: Patch server
yum:
name: kernel
state: latest
rescue:
– name: Rollback snapshot
command: revert_vm_snapshot
always:
– debug: msg=”Patch attempt completed”
🗓️ Days 9–10: Jinja2 & Templates (Power Tool)
If your templates are weak, your Ansible is weak.
Advanced Jinja
- Filters: map, selectattr, json_query
- Loops + conditionals
- default(), combine
- Dynamic config generation
🔥 Exercise:
Generate nginx.conf dynamically from a YAML dictionary.
🗓️ Days 11–12: AWX / Automation Controller
If you want enterprise credibility, this is mandatory.
Learn:
- Job Templates vs Workflows
- Credentials & RBAC
- Surveys
- Webhooks (Git-based automation)
- Rollback workflows
Understand how AWX executes playbooks vs CLI.
Also know how this differs in Red Hat Ansible Automation Platform.
🗓️ Days 13–14: Production Failure Scenarios
This separates operators from engineers.
Practice answering:
- Why did handlers not run?
- Why did playbook succeed but change didn’t apply?
- Inventory mismatch issues
- Parallel execution race conditions
- SSH / privilege escalation failures
🔥 Example: Task shows ok but service not restarted → handler never notified