Fail playbook when Ironic node reaches deploy failed state

[eurijon]

Stop silently accepting 'deploy failed' as a successful deployment outcome.
Add failed_when so the task fails immediately when the node reaches that state, while still exiting the retry loop early.

Summary by CodeRabbit

• Bug Fixes
  • Improved failure detection during hardware node provisioning: the wait step now aborts immediately if a node enters a "deploy failed" state, preventing further retries.
  • Preserves the existing until/retries/delay retry loop while providing faster failure feedback and clearer error behavior during provisioning.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 7dff4195-3544-4279-9dfd-a6a26a2fd482

📥 Commits

Reviewing files that changed from the base of the PR and between 5349652 and bc97dc2.

📒 Files selected for processing (1)

• playbooks/tasks/configure_hardware_ironic_node.yaml

🚧 Files skipped from review as they are similar to previous changes (1)

• playbooks/tasks/configure_hardware_ironic_node.yaml

---

Walkthrough

A failure condition was added to the "Wait for node to reach active state" task in the Ironic node configuration playbook: the task now fails immediately if the node provisioning state becomes "deploy failed", while existing retry/until behavior remains unchanged.

Changes

Cohort / File(s)	Summary
Ironic Node Provisioning Task playbooks/tasks/configure_hardware_ironic_node.yaml	Added a failed_when condition to the wait task to fail immediately when the node's provisioning state equals "deploy failed". Existing until/retries/delay settings remain unchanged.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately and concisely describes the main change: adding a failed_when condition to fail the playbook when an Ironic node reaches deploy failed state.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch ironic-deploy-failed-handling

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@playbooks/tasks/configure_hardware_ironic_node.yaml`:
- Line 192: The current per-host loop uses the failed_when condition checking
node_state.json.provision_state == "deploy failed", which causes the loop to
bail out and prevent subsequent cleanup tasks like wait_for_deployment.yaml from
running; modify the playbook so the loop body remains the same (keep the
failed_when check) but wrap the entire per-host loop in an Ansible block with an
always: section (or use block/rescue/always) so that cleanup tasks (e.g.,
wait_for_deployment.yaml and the Ironic detach/virtual media cleanup tasks) are
executed regardless of a deploy failure; ensure references to the failed_when
condition, node_state.json.provision_state, and wait_for_deployment.yaml are
preserved so the semantics and logging remain unchanged.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: dcae7010-c833-4213-a8d8-39c9bcf96b21

📥 Commits

Reviewing files that changed from the base of the PR and between 997e4ce and 2c65a27.

📒 Files selected for processing (1)

• playbooks/tasks/configure_hardware_ironic_node.yaml

[eurijon]

@agonzalezrh this change as it's now it's failing the playbook when the node goes to deploy failed, meaning that it exists from the agent_hosts loop and doesn't execute playbooks/tasks/wait_for_deployment.yaml, so playbooks/tasks/configure_hardware_ironic_detach_vmedia.yaml and playbooks/tasks/configure_hardware_ironic_cleanup.yaml are not executed. Is that ok? Should we do a proper cleanup if there is some issue with a node?