> ## Documentation Index
> Fetch the complete documentation index at: https://docs.dqlabs.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Azure Data Factory Overview

> What the Azure Data Factory connector covers, what Prizm observes, and how it authenticates.

The Azure Data Factory (ADF) connector gives Prizm read-only observability over your ADF pipelines. Prizm reads pipeline metadata, run history, and failures through the Azure REST APIs. It never creates, edits, triggers, or deletes anything in ADF.

## Why Connect ADF to Prizm?

**Pipeline Catalog** Prizm discovers the pipelines, activities, data flows, notebooks, and stored procedures in your data factories, along with their triggers and schedules, and keeps the catalog current as pipelines are added, changed, or removed.

**Run Monitoring** Prizm polls ADF for new pipeline and activity runs, so status, duration, retries, and failures are tracked without waiting for someone to open the ADF monitoring pane.

**Failure Detection** When a run fails, Prizm records the error message, error code, and the activity that failed, and raises an alert.

**Pipeline Metrics** Success rate, failure count, retry count, execution duration, last run, and pipeline status are available per pipeline alongside the rest of your Prizm pipeline sources.

## How ADF Maps to Prizm

| ADF | Prizm |
| :- | :- |
| Pipeline | Job |
| Activity | Asset under the job |
| Data flow | Task |
| Notebook activity | Notebook |
| Stored procedure activity | Stored procedure |
| Pipeline run | Job run |

Triggers and schedules are stored as schedule information on the job. Datasets and linked services are read for context but aren't cataloged as assets of their own.

## Authentication

ADF connections authenticate with a Microsoft Entra ID (Azure AD) identity. Choose one of three authentication types:

| Type | What you provide |
| :- | :- |
| Service Principal | Tenant ID, Client ID, and Client Secret |
| Certificate | Tenant ID, Client ID, a PEM certificate and private key, and an optional certificate password |
| Managed Identity | Optional Client ID of a user-assigned managed identity (leave blank for a system-assigned identity) |

Turning on **Use Vault** for a service principal or certificate connection replaces the credential fields with a **Select Vault** dropdown and a **Key** field. See [Setup](https://docs.dqlabs.ai/sources/adf/setup.md) for the full walkthrough and the Azure role the identity needs.

## Current Scope

The connector is read-only and focuses on pipeline metadata, run history, failures, and metrics. Near-real-time event delivery through Azure Event Grid, circuit breaker data quality gates, and lineage-based root cause and impact analysis are separate capabilities, and their availability depends on your Prizm release. They aren't covered here.

## Next Steps

> [card] [Setup](https://docs.dqlabs.ai/sources/adf/setup.md)
>
> Connect Azure Data Factory to Prizm — Azure permissions, authentication, and scope.

> [card] [What We Collect](https://docs.dqlabs.ai/sources/adf/what-we-collect.md)
>
> The pipeline metadata, run history, and metrics Prizm extracts.

> [card] [FAQ](https://docs.dqlabs.ai/sources/adf/faq.md)
>
> Common questions about the Azure Data Factory connector.

> [card] [Troubleshooting](https://docs.dqlabs.ai/help/troubleshooting.md)
>
> Diagnose connection, permission, and performance errors.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.