DocumentationTutorialsUse Cases & Blueprints

Deployment instructions

Last update: September 11, 2026

CREATED FOR:

  • Developer
  • User
WARNING
This is only required if you’re working through the labs at your own pace. If you’re in a live training course or event, your sandbox has already been deployed for you.

The AJO Architectural Foundations lab pack is deployed to your sandbox using the DEP CLI, a command-line tool that creates the schemas, datasets, dataflows, and sample data you’ll use throughout the labs — covering both the profile store and the AJO relational store used for orchestrated campaigns.

What gets deployed

Profile track

  • Identity namespaces (customerID, planID, productID)
  • Standard XDM field groups, descriptors, and schemas enabled for profile
  • Catalog datasets enabled for profile
  • Merge policies and audiences
  • Profile data for three sample datasets: Depeche Mode (traits + events), Stranger Things (traits), and Decisioning (traits)

Relational track

  • The customerID identity namespace
  • 11 relational XDM schemas with primary key, foreign key, and version descriptors
  • 11 datasets enabled for AJO orchestrated campaigns
  • 11 dataflows loading data from the Data Landing Zone
NOTE
End-to-end deployment takes about 2 hours 23 minutes. The profile and relational tracks run in parallel, and most of the time is unattended wait time that the CLI enforces automatically.

Prerequisites

  • License Entitlements. Administrative privileges for an IMS Org with Real-Time CDP (w/streaming segmentation) and Adobe Journey Optimizer (w/Orchestrated Campaigns)
  • Access Rights. An Experience Platform role with all permissions on the target sandbox including the API credential you created from Developer Console setup.
  • Developer Console credentials. A project that includes both Adobe Experience Platform API and Adobe Journey Optimizer APIs. If you don’t have these yet, follow Developer Console setup first
  • A sandbox. Empty, of type dev and in a “Ready” state for at least 120 minutes before you start deployment
  • Node.js. Any recent LTS version, on Windows or Mac

1. Install the CLI

  1. Clone or download the dep-cli repository
  2. From the dep-cli directory, run npm install
  3. Start the CLI with npm start
NOTE
Node.js is required before you run the commands above. If you don’t have Node.js installed yet, see the wiki’s Node.js Setuppage first. For full install details, including screenshots and how to update an existing install, see the Installationwiki page

2. Configure your environment file

The CLI deploys to whichever sandbox your environment file points at, so this has to be set up correctly before you run anything.

  1. Copy envFiles/sample-env.json and give it a new name, e.g. my-env.json

  2. Open the file and fill in the following fields using the values from Developer Console setup:

    table 0-row-2 1-row-2 2-row-2 3-row-2 4-row-2 5-row-2
    Field Value
    API_KEY Client ID
    CLIENT_SECRET Client Secret
    IMS_ORG Organization ID
    SCOPES Must include both Experience Platform API and Adobe Journey Optimizer API scopes
    (e.g. cjm.suppression_service.client.delete, cjm.suppression_service.client.all, openid, session, AdobeID, read_organizations, additional_info.projectedProductContext)
    SANDBOX_NAME The sandbox you’re targeting — must be empty and of type dev

  3. Save and close the file

NOTE
You’re prompted for this file’s name each time you run a CLI command, so you can reuse it across every step below.

3. Run the AJO Architectural Foundations menu

From the main menu, select AJO arch foundations. There are six steps split across two tracks.

Profile track (run in order)

WARNING
The sandbox must have been in a “Ready” state for at least 60 minutes before you run Step 1.
Step
What it does
Before you run it
1. Create profile base
Deploys identity namespaces, schemas, datasets, merge policies, and audiences
Sandbox “Ready” for 60+ minutes
2. Load profile data
Creates dataflows and streams Depeche Mode, Stranger Things, and Decisioning profile data
Wait 60+ minutes after Step 1
3. Check profile health
Validates that all profile data loaded correctly
Wait 15+ minutes after Step 2

Step 1 takes about 2 minutes, Step 2 about 6 minutes.

NOTE
Step 2 is safe to re-run if something fails — it overwrites existing traits and skips duplicate events.

Relational track

WARNING
The sandbox must have been in a “Ready” state for at least 120 minutes before you run Step 4 or Step 6.
Step
What it does
Before you run it
4. Create relational base
Creates the customerID namespace and relational schemas/descriptors/datasets
Sandbox “Ready” for 120+ minutes
5. Load relational data
Uploads 11 CSV files and creates the dataflows that load them
Runs right after Step 4 — no manual wait needed
6. Deploy relational base & data
Combines Steps 4 and 5 into a single ~5-minute run
Sandbox “Ready” for 120+ minutes
NOTE
Use Step 6 instead of running Steps 4 and 5 separately — it does the same thing in one pass with the propagation wait handled for you.
NOTE
All wait times above are checked automatically by the CLI. If you run a step too early, it blocks and tells you how long to wait.

Troubleshooting

WARNING
Profile health check fails with missing events. Some profile data hasn’t finished propagating yet. Wait another 15 minutes and re-run Check profile health. If it still fails, re-run Load profile data, wait 15 minutes, and check again.
WARNING
Relational data load fails partway through. Each API call retries up to 3 times. If it still fails, cleanup removes the source connections, target connections, and dataflows it created so you can re-run Step 5 (or Step 6) cleanly. Mapping sets can’t be deleted through the API and may be left behind — this doesn’t affect redeployment.

Something else looks wrong. As a last resort, you can reset the sandbox from the CLI’s Sandbox management menu and redeploy from Step 1.

CAUTION
Resetting a sandbox is destructive. The CLI asks you to type the sandbox name to confirm before proceeding.
recommendation-more-help
blueprints-learn-help-blueprints