Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
61 changes: 61 additions & 0 deletions infrastructure/ao/demos.yml
Original file line number Diff line number Diff line change
Expand Up @@ -172,3 +172,64 @@ ao_demos:
credentials:
- machine
limit: aws_gitea

# Zero-Day Response (PHP2Shell) — KNOW+CONTAIN owns p2s-* JTs; FIX+PROVE reuses them.
# Upstream content: ansible-tmm/aap-orchestrator-demos zero-day-response/
- slug: zero-day-response

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Merge order: ao_demos_ref is still main, and zero-day-response/ is 404 on ansible-tmm/aap-orchestrator-demos main today (only on PR #12).

Please keep this PR open until that lands — otherwise the next APD Install that seeds enabled: true entries will try to import zero-day-response/ao/flow-a-know-contain.json (and create JTs against playbooks) that do not exist yet.

Author already called this out in the PR body; treating it as a hard gate before merge.

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Agreed, this stays open until ansible-tmm/aap-orchestrator-demos#12 is merged to main. #12 had fallen behind main with a _config.yml conflict; I merged main into it (68acfc4) and it's mergeable again.

enabled: true
workflow: zero-day-response/ao/flow-a-know-contain.json
requires_cloud_stack: false
requires_mattermost: false
job_templates:
- name: p2s-estate-configure
playbook: zero-day-response/playbooks/estate_configure.yml
credentials:
- machine
- name: p2s-estate-provision
playbook: zero-day-response/playbooks/infra_provision_aws.yml
credentials:
- aws
- machine
- name: p2s-independent-check
playbook: zero-day-response/playbooks/p2s_independent_check.yml
credentials:
- machine
- name: p2s-know-scan
playbook: zero-day-response/playbooks/p2s_scan.yml
credentials:
- machine
- name: p2s-know-lookback
playbook: zero-day-response/playbooks/p2s_lookback.yml
credentials:
- machine
- name: p2s-ai-output-check
playbook: zero-day-response/playbooks/p2s_ai_check.yml
credentials: []
- name: p2s-contain-spec-check
playbook: zero-day-response/playbooks/p2s_spec_check.yml
credentials: []
- name: p2s-contain-controls
playbook: zero-day-response/playbooks/p2s_controls.yml
credentials:
- machine
- name: p2s-fix-availability
playbook: zero-day-response/playbooks/p2s_fix_availability.yml
credentials: []
- name: p2s-fix-patch
playbook: zero-day-response/playbooks/p2s_patch.yml
credentials:
- machine
- name: p2s-fix-standdown
playbook: zero-day-response/playbooks/p2s_standdown.yml
credentials:
- machine
- name: p2s-prove-case-record
playbook: zero-day-response/playbooks/p2s_case_record.yml
credentials: []

- slug: zero-day-response-fix
enabled: true
workflow: zero-day-response/ao/flow-b-fix-prove.json
requires_cloud_stack: false
requires_mattermost: false
job_templates: []

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same pattern as multi-os (job_templates: [] while the sibling slug owns the JTs) — good. Flow B refs (p2s-fix-*, shared scan/AI/case-record) are covered by the list above.

Minor: the Install docs bullet only names zero-day-response; both slugs are enabled: true, so seed will also import zero-day-response-fix. Worth naming both in automation-orchestrator.md so presenters know two AO workflows show up.

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done in 9de5b02: the Install summary lists both zero-day-response and zero-day-response-fix, and a short "Zero-Day Response — two workflows" section names both AO workflows as presenters see them.

11 changes: 8 additions & 3 deletions infrastructure/docs/automation-orchestrator.md
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@ Still simple — shared lab services (and cloud VMs only when a demo needs them)

1. Ensure **APD Multi-demo setup** (or Single demo setup with cloud + infrastructure) created the Cloud AWS patch/snapshot job templates and the OpenFlake / Gitea / Mattermost credential types.
2. Launch **Infrastructure AWS - Deploy Shared Services** ([details](./shared-services.md)). One workflow: VPC + keypair, then Mattermost + Gitea in parallel, plus OpenFlake on OpenShift; when all three succeed it publishes a password-gated S3 credentials portal for this lab. That covers chat (disk-utilization), SCM/Actions (inventory-service + act_runner), and ITSM (OpenFlake) credentials — you do not need separate Provision Mattermost / Provision Gitea / OpenFlake jobs unless you prefer to run them alone. Copy the portal URL and password (`Ansible123!`) from the **Shared Services Portal** nested job.
3. Only if you need disk-utilization or Multi-OS Cloud Patching hosts: also run **Deploy Cloud Stack in AWS** (aws-rhel9, Windows, reports, and so on). Skip for CI/CD Pipeline or Task Step Example alone.
3. Only if you need disk-utilization or Multi-OS Cloud Patching hosts: also run **Deploy Cloud Stack in AWS** (aws-rhel9, Windows, reports, and so on). Skip for CI/CD Pipeline, Zero-Day Response, or Task Step Example alone.
4. Launch **Infrastructure Automation Orchestrator Install**. Soft-failed / retried tasks in the middle are normal on a cold cluster; the job should still finish green (often about 10 minutes). Soft warnings for missing optional deps do not fail Install.
5. Use the **seed summary** at the end of Install for AO URL / admin password (and Mattermost / OpenFlake / Gitea login details when Shared Services ran).
6. For CI/CD: Install creates an AO webhook service account, authorizes it on lightwell-ci-cd, publishes the workflow, and writes Gitea Actions secrets (AO_WEBHOOK_URL + OAuth client). Confirm that in the seed summary, then open the seeded workflow.
Expand All @@ -40,6 +40,7 @@ Still simple — shared lab services (and cloud VMs only when a demo needs them)
| Disk utilization | [Disk Utilization & Remediation](https://ansible-tmm.github.io/aap-orchestrator-demos/demos/disk-utilization/) | aws-rhel9 + Mattermost |
| Multi-OS Cloud Patching | [Multi-OS Cloud Patching](https://ansible-tmm.github.io/aap-orchestrator-demos/demos/multi-os-cloud-patching/) | Full cloud stack + existing Cloud AWS patch/snapshot JTs |
| CI/CD Pipeline | [CI/CD Pipeline](https://ansible-tmm.github.io/aap-orchestrator-demos/demos/cicd-pipeline/) | [Gitea](./gitea.md) (inventory-service) + OpenFlake + AO_WEBHOOK_URL. Re-run helper: **CI/CD Demo - Reset Change Request** (default CHG0034907). |
| Zero-Day Response | [Zero-Day Response](https://ansible-tmm.github.io/aap-orchestrator-demos/demos/zero-day-response/) | Its own six-instance AWS estate (no cloud stack, no Mattermost): **p2s-estate-provision**, update **AWS Inventory**, then **p2s-estate-configure** |

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Prep sequence looks right (provision → AWS Inventory sync → configure), and the six-instance count matches nodes-aws.yml.

Please also call out that this depends on ansible-tmm/aap-orchestrator-demos#12 being on main (or that Install must use a ref that contains zero-day-response/), and that two AO workflows are seeded (KNOW/CONTAIN + FIX/PROVE). The GitHub Pages link will 404 until that demos PR’s site content is published too.

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

9de5b02 adds the two-workflow note and says ao_demos_ref must point at a ref that contains zero-day-response/. I kept the PR number and the 404 out of the doc itself: this PR merges only after #12, when the folder and the page are already on main, so that note would be stale on arrival. The merge order stays in the PR description.


Install rewrites CI/CD close/rollback ServiceNow notes to trigger-based text so an empty AI structured output cannot block the AAP step (no Controller job = expression failed before launch). Seed also forces agentic note nodes to empty tool selections plus `response_schema.required` (LiteMaaS/qwen otherwise often returns `finish_reason: tool_calls` with empty `result.content` while the node still shows completed).

Expand All @@ -60,6 +61,10 @@ The seeded workflow pre-fills inventory limits for the cloud stack. Set **`aws_r
}
```

### Zero-Day Response — two workflows

Install seeds two workflows from two `demos.yml` entries: **Zero-Day Response - PHP2Shell: KNOW + CONTAIN** (`zero-day-response`, which also creates the `p2s-*` job templates) and **Zero-Day Response - PHP2Shell: FIX + PROVE** (`zero-day-response-fix`, which reuses them). Run KNOW + CONTAIN first. Both come from the `zero-day-response/` folder of [aap-orchestrator-demos](https://github.com/ansible-tmm/aap-orchestrator-demos), so `ao_demos_ref` in [`demos.yml`](../ao/demos.yml) must point at a ref that contains it.

## The three AO templates

| Template | When to use it | Playbook |
Expand All @@ -83,8 +88,8 @@ The seeded workflow pre-fills inventory limits for the cloud stack. Set **`aws_r

What **Install** does when it seeds curated demos:

- Reads [`demos.yml`](../ao/demos.yml) and seeds entries with `enabled: true` from [aap-orchestrator-demos](https://github.com/ansible-tmm/aap-orchestrator-demos) (**disk-utilization**, **multi-os-cloud-patching**, **cicd-pipeline**) plus local workflows such as **Task Step Example**.
- Creates the **AAP Orchestrator Demos** project and any demo-owned job templates (disk-utilization / CI/CD playbooks). Multi-OS reuses existing **Cloud | AWS |*** templates from cloud setup — seed only imports the workflow.
- Reads [`demos.yml`](../ao/demos.yml) and seeds entries with `enabled: true` from [aap-orchestrator-demos](https://github.com/ansible-tmm/aap-orchestrator-demos) (**disk-utilization**, **multi-os-cloud-patching**, **cicd-pipeline**, **zero-day-response**, **zero-day-response-fix**) plus local workflows such as **Task Step Example**.
- Creates the **AAP Orchestrator Demos** project and any demo-owned job templates (disk-utilization / CI/CD playbooks and the zero-day-response `p2s-*` templates). Multi-OS reuses existing **Cloud | AWS |*** templates from cloud setup — seed only imports the workflow.
- Wires AO-native **AAP** and **LiteMaaS** (LLM Provider) integrations; allow-lists the LiteMaaS hostname for AO workers.
- Rewrites each workflow’s AAP org/credential placeholders → APD, and fills `llm_model_id` on agentic nodes when LiteMaaS is wired, then imports into AO.
- When present, attaches the **Mattermost** credential and **Product Demos EE** to **Notify Chatroom**.
Expand Down
Loading