Feature activation

cds-data-pipeline separates three concerns:

Layer	What it does	Enabled by default?
Pipeline runtime	cds.connect.to('data-pipeline'), addPipeline, adapters	Yes — npm add cds-data-pipeline
Management API	Tracker schema + DataPipelineManagementService at /pipeline/	No — opt in via management.reuse.api or manual using
Pipeline Console	Pre-built UI5 FCL at /pipeline-console/	No — opt in via management.reuse.console or cds add pipeline-console

Reuse vs own (CAP reuse & compose)

Following CAP's reuse & compose pattern, management and UI can be activated in two ways:

Mode	When	How
Reuse (config)	Zero project files; local dev; CAP serves static UI	management.reuse.api, management.reuse.console
Own (manual)	BTP HTML5 repo, customization, explicit control	using … index.cds and/or cds add pipeline-console

Manual equivalents:

Reuse flag	Manual alternative
management.reuse.api	using from 'cds-data-pipeline/index.cds' in srv/
management.reuse.console	cds add pipeline-console (copies UI into app/pipeline-console/)

Do not combine reuse flags with their manual equivalent for the same concern.

Quick decision guide

Need pipeline engine only?
  └─ npm add cds-data-pipeline

Need management OData at /pipeline/ (reuse from package)?
  └─ "data-pipeline": { "impl": "cds-data-pipeline", "management": { "reuse": { "api": true } } }

Need Pipeline Console locally (reuse UI from package)?
  └─ "management": { "reuse": { "api": true, "console": true } }

Need console on BTP (own the UI project)?
  └─ cds add pipeline-console
  └─ cds add html5-repo && cds add mta
  └─ do NOT set management.reuse.console

Configuration

Minimal (engine only)

npm add cds-data-pipeline

Optional explicit wiring (mirrors cds-caching):

{
  "cds": {
    "requires": {
      "data-pipeline": {
        "impl": "cds-data-pipeline"
      }
    }
  }
}

Connect: await cds.connect.to('data-pipeline').

Local development (reuse API + console)

{
  "cds": {
    "requires": {
      "data-pipeline": {
        "impl": "cds-data-pipeline",
        "management": {
          "reuse": {
            "api": true,
            "console": true
          }
        }
      }
    }
  }
}

cds deploy && cds watch
# → http://localhost:4004/pipeline-console/index.html
# → http://localhost:4004/pipeline/Pipelines

Field	Purpose
management.reuse.api	Plugin injects tracker schema + management service via index.cds
management.reuse.console	Plugin serves pre-built UI at /pipeline-console/; implies reuse.api: true

API only (reuse API, no UI)

"management": {
  "reuse": { "api": true }
}

SAP BTP production

"data-pipeline": {
  "impl": "cds-data-pipeline",
  "management": { "reuse": { "api": true } }
}

cds add pipeline-console
cds add html5-repo
cds add mta
cd app/pipeline-console && npm install && npm run build:cf

Secure the management service in your CDS model. Do not use management.reuse.console on standard HTML5 repo + approuter deployments.

Feature matrix

Setup	Engine	Tracker + /pipeline/	Pipeline Console
npm add only	✓
+ manual using from 'cds-data-pipeline/db'	✓	partial (schema only)
+ management.reuse.api	✓	✓
+ management.reuse.console	✓	✓	✓
+ cds add pipeline-console	✓	✓*	✓

*When using cds add pipeline-console, the generator writes srv/pipeline-management.cds with using from 'cds-data-pipeline/index.cds' unless you already import the API manually.

Avoiding duplicate model loading

The plugin deduplicates automatic loading:

• management.reuse.api or management.reuse.console → injects index/ model root only
• Manual using from 'cds-data-pipeline/index.cds' → no separate injection when reuse flags are off

Do not combine

Combination	Problem
management.reuse.api + using … index.cds	Duplicate management service definitions
management.reuse.console + cds add pipeline-console	Duplicate UI route at /pipeline-console
using … index.cds in two .cds files	Duplicate service registration

Federation and materialization consumers that only need tracker tables (no /pipeline/ UI) may still use:

using from 'cds-data-pipeline/db';

That path is independent of management.reuse.api.

Securing the management API

annotate DataPipelineManagementService with @requires: 'authenticated-user';

Use the service name from your consumption model — do not repeat the using import in the annotate statement.

See also

• Pipeline Console — UI capabilities and BTP setup
• Management service — OData entities and actions
• Get started — end-to-end walkthrough