Playbooks: Design, Triggers, and Actions
Playbooks: Design, Triggers, and Actions covers X-Play playbooks in Prism Central. The exam tests whether you can design a safe, scoped automation flow and choose the right trigger and action types.
Design rules
- A playbook is a trigger plus a flow of tasks: decisions, actions, approvals, loops, and parallel branches.
- Use variables to pass trigger payload and task output; avoid hardcoding names, IDs, or secrets.
- Attach credentials and endpoints as reusable objects; do not embed passwords in scripts.
- Scope triggers narrowly, for example by alert type, category, or project, so unrelated events do not fire the playbook.
Trigger and action types
- Triggers include alert-based, scheduled, and manual. Alert triggers receive alert context; manual triggers are useful for testing.
- Actions include REST API calls, script execution, email notifications, and approval tasks. Decision tasks branch on values; loops repeat until a condition or limit.
Worked example
An alert trigger fires on a high-CPU alert. A decision checks that the VM's category is Env:Prod. If true, a script action collects diagnostics and an approval task asks the owner before a REST action scales the VM. If false, the playbook ends. The trigger payload supplies the VM UUID; the endpoint supplies the Prism Central connection.
Traps
- Testing a disabled playbook and seeing no run.
- Using a broad alert trigger that fires for every VM.
- Omitting an exit condition in a loop.
- Forgetting that approval tasks pause execution and need a responder.
- Assuming script actions run on the target VM; they run in the automation execution context.