---
title: "OneLake Catalog vs a Platform-Independent Data Catalog | Dawiso Blog"
description: "OneLake catalog vs a platform-independent data catalog: what each governs, where the OneLake catalog stops, and how to choose after Purview Data Governance."
url: "https://www.dawiso.com/blog-post/onelake-catalog-vs-platform-independent-catalog"
language: "en"
---

# When the OneLake Catalog Is Enough, and When You Need a Platform-Independent Data Catalog

Samuel Nagy

VP of Strategic Growth

November 16, 2026

The OneLake catalog is the built-in catalog of Microsoft Fabric, and now that Microsoft is retiring Purview Data Governance it is Microsoft's answer to data governance. It is enough when your data and your governance work live in Fabric. When Fabric runs next to Databricks, Snowflake, SAP or on-premises systems, or when governance means a business glossary, data products and cross-platform lineage, you also need a data catalog that does not belong to any one platform.

TL;DR

The **OneLake catalog** governs **Microsoft Fabric items** well: discovery, endorsement, tags, sensitivity labels, lineage and access, included in Fabric capacity. A **platform-independent data catalog** connects to Fabric **and** to Databricks, Snowflake, SAP and on-premises systems, and holds the **business glossary, data products and lineage** that span them. If almost all your data is in Fabric, the OneLake catalog may be enough. If it is not, or if AI agents need one set of definitions across platforms, use both and let one tool own the definitions.

## What is the OneLake catalog?

**The [OneLake catalog](https://www.dawiso.com/glossary/onelake-catalog) is the catalog built into [Microsoft Fabric](https://www.dawiso.com/glossary/microsoft-fabric), where users discover, govern and secure the Fabric items they own or can access.** According to [Microsoft Learn](https://learn.microsoft.com/en-us/fabric/governance/onelake-catalog-overview), it has three tabs: Explore for finding lakehouses, warehouses, semantic models, reports and other items, Govern for governance health and recommended actions, and Secure for workspace roles and OneLake security roles. It is also embedded in Teams, Excel, Copilot Studio and Foundry.

Since November 12, 2026 it carries more weight. Microsoft [announced](https://techcommunity.microsoft.com/blog/microsoft-security-blog/a-next-step-in-the-evolution-of-the-microsoft-data-governance-strategy/4572227) that Purview Data Governance will be retired, with support ending November 12, 2027, and called the OneLake catalog the single foundation for governing data. Our article on [the Purview Data Governance retirement](https://www.dawiso.com/blog-post/microsoft-purview-data-governance-retirement) covers the dates and what changes.

## What is a platform-independent data catalog?

**A platform-independent [data catalog](https://www.dawiso.com/glossary/data-catalog) is a catalog that is not part of any one data platform and reads metadata from the platforms it connects to.** It connects to Fabric, Databricks, Snowflake, SAP, Oracle, BI tools and on-premises databases through connectors, usually read-only, and builds one inventory with one [business glossary](https://www.dawiso.com/glossary/business-glossary), one set of owners and [lineage](https://www.dawiso.com/glossary/data-lineage) that crosses platform boundaries.

The difference is where governance content lives. In a platform catalog, a definition, an owner or a data product is stored inside the platform and describes that platform's assets. In a platform-independent catalog it is stored once and linked to assets wherever they are, so the definition of "net revenue" is the same whether the number comes from a Fabric semantic model or a Databricks table.

## OneLake catalog vs a platform-independent data catalog

**The two differ mainly in scope and in which governance objects they manage.** The comparison below uses Microsoft's documentation as of November 2026.

|  | OneLake catalog | Platform-independent data catalog |
| --- | --- | --- |
| Scope | Items in the Fabric tenant, including data shortcut or mirrored into OneLake | Every connected platform, Fabric included, through connectors |
| Other platforms | Through OneLake shortcuts, mirroring and supported connectors; Microsoft notes that data connectivity is not enterprise metadata scanning | Scanned in place, in their own structure |
| Governance objects | Domains, endorsement, tags, descriptions, sensitivity labels; a business glossary, data products as governed objects, critical data elements and data quality rules are not described | Typically a business glossary, domains, data products, owners and stewards, with approval workflows |
| Lineage | Shown in the item view for Fabric items | Across platforms, from source systems through transformation to BI |
| Access control | Manages workspace roles and OneLake security roles, including row-level and column-level security | Usually documents access; enforcement stays in each platform |
| AI and developer access | Catalog search through the Catalog Search REST API, the Fabric MCP and the Fabric CLI, partly in preview | Depends on the product; some expose governed definitions over MCP |
| Pricing | Included in Fabric capacity | Separate license |
| Roadmap set by | Microsoft's Fabric strategy | The catalog vendor, independent of any one platform |

Neither column is better in general. The OneLake catalog is closer to the data and can enforce access, which a catalog outside Fabric typically does not. A platform-independent catalog sees further and holds the business layer. The question is which of these your governance program needs.

## When is the OneLake catalog enough?

**The OneLake catalog is enough when Fabric is effectively your whole data estate and governance means making Fabric items findable, trusted and secure.** That is typically the case when:

- **Data lives in Fabric and Power BI**, with other sources loaded into OneLake rather than used where they are.
- **Endorsement, tags, descriptions and sensitivity labels** cover what the business needs to trust a report or a lakehouse.
- **Fabric admins own governance**, and there is no separate data office running a glossary or data products program.
- **Budget is the deciding factor**, since the OneLake catalog is included in Fabric capacity.

For these teams, adding a second catalog would mean maintaining the same descriptions twice. Microsoft's [Fabric blog](https://blog.fabric.microsoft.com/en-us/blog/onelake-catalog-the-trusted-catalog-for-organizations-worldwide) reports that the OneLake catalog was adopted by over 230,000 organizations in its first year, and for many of them it is the right scope.

## When do you need a platform-independent data catalog?

**You need a platform-independent data catalog when governance has to cover more than Fabric, or more than the objects the OneLake catalog manages.** The common signals:

- **Fabric runs next to Databricks, Snowflake, SAP or on-premises systems**, and not all of that data will be shortcut or mirrored into OneLake.
- **Lineage has to start at the source system.** Impact analysis that stops at the edge of Fabric does not answer what breaks when an SAP or Oracle table changes.
- **Governance is built on a business glossary and [data products](https://www.dawiso.com/glossary/data-product)**, with owners, approval states and a lifecycle. As of November 2026, Microsoft's OneLake catalog documentation does not describe these objects, nor [critical data elements](https://www.dawiso.com/glossary/critical-data-elements) or data quality rules.
- **Regulators expect an audit trail** of definitions, ownership and approvals that survives a change of platform.
- **AI agents work across platforms** and need one set of definitions, owners and lineage, regardless of which platform holds the data they query.

Microsoft itself says some organizations have needs that reach beyond what is in Fabric today, and that Fabric integrates with third-party governance vendors for them. The [Govern documentation](https://learn.microsoft.com/en-us/fabric/governance/onelake-catalog-govern) adds concrete limits: insights do not cover subitems such as tables, the estate report charts exclude third-party workload items, and cross-tenant scenarios are not supported.

## How the OneLake catalog and a data catalog work together

**In a multi-platform estate, the two catalogs do different jobs, and the main design decision is which one owns business definitions.** A pattern that works:

1. **Fabric users keep the OneLake catalog** for daily discovery, endorsement and access management inside Fabric. Nothing about their workflow changes.
2. **The platform-independent catalog reads Fabric metadata** as one source among several, together with Databricks, Snowflake and the source systems.
3. **Business definitions live in one place.** The glossary, data products, owners and approval states are kept in the platform-independent catalog and linked to Fabric items, Databricks tables and Snowflake views alike.
4. **Security stays where it is enforced.** Sensitivity labels and data loss prevention stay in Microsoft Purview, OneLake security roles in Fabric, and grants in each platform. The catalog documents them and does not replace them.
5. **AI agents read the governed layer.** Agents can still search Fabric items through the Fabric MCP, and they get definitions, owners and cross-platform lineage from the catalog that holds them.

The same split works for Databricks and Snowflake. Unity Catalog and Snowflake Horizon govern their own platforms in depth, and a platform-independent catalog connects them, as our comparisons of [Unity Catalog and Microsoft Purview](https://www.dawiso.com/glossary/unity-catalog-vs-microsoft-purview) and of [Databricks and Microsoft Fabric](https://www.dawiso.com/glossary/databricks-vs-microsoft-fabric) describe.

## Why platform independence matters after Purview Data Governance

**Governance content is built slowly and outlives the tools it is stored in, so where it is stored is a long-term decision.** This is the second time Microsoft has asked catalog customers to move. Azure Data Catalog users were moved to Purview, a different service, and Microsoft's [migration guide](https://learn.microsoft.com/en-us/previous-versions/azure/data-catalog/data-catalog-migration-to-microsoft-purview) noted there was no in-place upgrade. Purview Data Governance customers now face a move to Fabric, where Microsoft's own [transition guidance](https://techcommunity.microsoft.com/blog/analyticsonazure/from-purview-data-governance-to-the-fabric-onelake-catalog/4573422) recommends exporting and preserving glossary terms and ownership. Each move was reasonable from Microsoft's point of view. For customers, each one is a project.

A platform-independent catalog does not remove migrations, since connectors and platforms keep changing. It does mean that a change in one vendor's platform strategy does not force a move of the business layer. That matters more as AI agents start to depend on the same definitions: a glossary that has to be rebuilt every few years is a weak foundation for answers people are expected to trust.

## How Dawiso works with Microsoft Fabric

**Dawiso is a platform-independent data catalog with a [Microsoft Fabric connector](https://www.dawiso.com/connectors/microsoft-fabric), so Fabric is one governed source next to the rest of the estate.** Based on the [Dawiso documentation](https://help.dawiso.com/actual/microsoft-fabric/):

- **The Fabric connector** reads workspaces, lakehouses, warehouses, SQL databases, semantic models, reports, pipelines and notebooks through the Power BI Admin API and Fabric API, with an Entra ID service principal. It reads metadata only; the data stays in OneLake.
- **Advanced lineage**: with Contributor access, view and procedure definitions are read as well, which Dawiso uses for advanced lineage.
- **The rest of the estate**: connectors for Databricks, Snowflake, dbt, SAP HANA, Oracle, Tableau and more, in one catalog. See [all connectors](https://www.dawiso.com/connectors).
- **The business layer**: a [business glossary](https://www.dawiso.com/product/business-glossary) with domains, terms, indicators and rules, approval workflows, and [data products](https://www.dawiso.com/product/data-products) with a lifecycle from announcement and draft through production to retired.
- **AI agents**: the [Dawiso MCP Server](https://www.dawiso.com/product/mcp) gives MCP-compatible agents the governed definitions, owners and lineage.

Bertel O. Steen runs Microsoft Fabric and Databricks side by side in Dawiso, with more than 18,000 Fabric objects cataloged. [Read the Bertel O. Steen story](https://www.dawiso.com/case-study/bertel-o-steen).

## FAQ

**Is the OneLake catalog a data catalog?**

Yes, for Microsoft Fabric. The OneLake catalog lets people discover, govern and secure Fabric items such as lakehouses, warehouses, semantic models and reports, with owners, endorsement, tags, sensitivity labels and lineage. Its scope is the Fabric tenant: data from other platforms is covered when it is brought into OneLake through shortcuts or mirroring.

**Can the OneLake catalog govern Databricks or Snowflake data?**

Partly. Microsoft says the OneLake catalog governs data across Fabric, Azure and third-party sources through shortcuts, mirroring and supported connectors. Data from Databricks or Snowflake is therefore visible in the OneLake catalog as Fabric items once it is shortcut or mirrored into OneLake. The governance, lineage and ownership inside the original platform stay in that platform's own catalog.

**Do I need both the OneLake catalog and another data catalog?**

Many organizations with more than one data platform will use both. Fabric users keep working in the OneLake catalog for discovery, endorsement and access, while the platform-independent catalog holds the business glossary, data products and lineage across Fabric and the other platforms. Decide which tool owns business definitions, so they are not maintained twice.

**What is the difference between the OneLake catalog and Unity Catalog?**

Both are catalogs built into a data platform. The OneLake catalog covers Microsoft Fabric items, and Unity Catalog covers data and AI assets in Databricks. Each governs its own platform in depth and other platforms only through federation, shortcuts or mirroring. A platform-independent catalog connects to both and holds the definitions and lineage that span them.

## Sources

- Microsoft Learn - [OneLake catalog overview](https://learn.microsoft.com/en-us/fabric/governance/onelake-catalog-overview)
- Microsoft Learn - [Discover and explore Fabric items in the OneLake catalog](https://learn.microsoft.com/en-us/fabric/governance/onelake-catalog-explore)
- Microsoft Learn - [Govern your data in the OneLake catalog](https://learn.microsoft.com/en-us/fabric/governance/onelake-catalog-govern)
- Microsoft Fabric Blog - [OneLake catalog: The trusted catalog for organizations worldwide](https://blog.fabric.microsoft.com/en-us/blog/onelake-catalog-the-trusted-catalog-for-organizations-worldwide) (February 9, 2026)
- Microsoft Analytics on Azure Blog - [From Purview Data Governance to the Fabric OneLake Catalog](https://techcommunity.microsoft.com/blog/analyticsonazure/from-purview-data-governance-to-the-fabric-onelake-catalog/4573422) (November 13, 2026)
- Microsoft Security Community Blog - [A next step in the evolution of the Microsoft Data Governance strategy](https://techcommunity.microsoft.com/blog/microsoft-security-blog/a-next-step-in-the-evolution-of-the-microsoft-data-governance-strategy/4572227) (November 12, 2026)
- Microsoft Learn - [Migrate from Azure Data Catalog to Microsoft Purview](https://learn.microsoft.com/en-us/previous-versions/azure/data-catalog/data-catalog-migration-to-microsoft-purview)
- Dawiso Help - [Microsoft Fabric](https://help.dawiso.com/actual/microsoft-fabric/), [Business Glossary](https://help.dawiso.com/actual/business-glossary/) and [Data product lifecycle](https://help.dawiso.com/actual/data-product-lifecycle/)

[See it in action](https://www.dawiso.com/product/data-catalog)

One data catalog for Fabric and the rest of your stack

See how the Dawiso Data Catalog brings Microsoft Fabric, Databricks, Snowflake and more into one governed catalog.

### Table of contents

What is the OneLake catalog?What is a platform-independent data catalog?OneLake catalog vs a platform-independent data catalogWhen is the OneLake catalog enough?When do you need a platform-independent data catalog?How the OneLake catalog and a data catalog work togetherWhy platform independence matters after Purview Data GovernanceHow Dawiso works with Microsoft FabricFAQSources

## More like this

Keep reading and take a deeper dive into our most recent content on metadata management and beyond:

### [Microsoft Purview Data Governance Is Retiring](https://www.dawiso.com/blog-post/microsoft-purview-data-governance-retirement)

Read more

### [Best Data Catalog Tools](https://www.dawiso.com/blog-post/best-data-catalog-tools-2026)

Read more
