Fabric to Blindata catalog

This page explains where each Fabric object lands in the Blindata catalog, so you know what to look for after a run.

The rule of thumb:

• An item that stores data (a lakehouse, a warehouse, a database) becomes a system of its own.
• What it contains (tables, views, files) becomes physical entities, and their columns become physical fields.
• An item that processes or presents data (a pipeline, a notebook, a report) becomes a physical entity inside a shared product system, such as Microsoft Fabric Data Factory.
• Data moving between them becomes data flows. See Data lineage .

The workspace has no object of its own: its name prefixes the system names and groups the item entities inside product systems.

Systems

One system per data store

Fabric item	Blindata system	Technology
Lakehouse	<workspace> / <lakehouse>	Microsoft Fabric Lakehouse
Warehouse, warehouse snapshot	<workspace> / <warehouse>	Microsoft Fabric Warehouse
Mirrored database, mirrored warehouse	<workspace> / <item>	Microsoft Fabric Mirrored Database
Mirrored catalog	<workspace> / <item>	Microsoft Fabric Mirrored Catalog
Mirrored Snowflake database	<workspace> / <item>	Microsoft Fabric Snowflake Database
SQL database	<workspace> / <database>	Microsoft Fabric SQL Database
Cosmos DB database	<workspace> / <database>	Microsoft Fabric Cosmos DB
KQL database	<workspace> / <database>	Microsoft Fabric KQL Database
Datamart	<workspace> / <datamart>	Microsoft Fabric Datamart

You can change the name pattern, lowercase it, or pin an item to a system that already exists in Blindata with systemNaming .

One system per product

Product system (default name)	Contains
Microsoft Fabric Data Engineering	Notebooks, Spark job definitions
Microsoft Fabric Data Factory	Pipelines, Copy jobs, Dataflows Gen2 and their queries, Airflow and dbt jobs
Microsoft Fabric Real-Time Intelligence	Eventstreams, eventhouses
Microsoft Fabric Graph	Graph models
Power BI Datasets	Semantic models with their tables, columns and measures, Gen1 dataflows
Power BI Reports	Reports, dashboards and tiles, org apps

Product systems are created only when the run publishes at least one of their items. You can rename them or point them to existing systems under products .

Physical entities

Fabric object	System	Schema	Name	Type
Lakehouse, warehouse or database table	The item’s system	Table schema (dbo when the lakehouse has no schemas)	Table name	TABLE
View	The item’s system	View schema	View name	VIEW
Lakehouse folder or file	The lakehouse system	Files	Path under Files/, such as raw/2024/orders.csv	FOLDER, FILE
Shortcut	The lakehouse system	As the table or folder it exposes	As the table or folder it exposes	SHORTCUT
KQL table, materialized view	The KQL database system	none (KQL has no schemas)	Table or view name	TABLE, MATERIALIZED_VIEW
Pipeline, Copy job, Dataflow Gen2	Data Factory	Workspace name	Item name	DATA_PIPELINE, COPY_JOB, DATAFLOW_GEN2
Dataflow Gen2 query	Data Factory	<workspace>.<dataflow>	Query name	DATAFLOW_QUERY
Notebook, Spark job definition	Data Engineering	Workspace name	Item name	NOTEBOOK, SPARK_JOB_DEFINITION
Eventstream, eventhouse	Real-Time Intelligence	Workspace name	Item name	EVENTSTREAM, EVENTHOUSE
Graph model	Graph	Workspace name	Item name	GRAPH_MODEL
Semantic model	Power BI Datasets	Workspace name	Model name	DATASET
Semantic model table	Power BI Datasets	<workspace>.<model>	Table name	DATASET_TABLE
Dataflow Gen1, and its entities	Power BI Datasets	Workspace name, or <workspace>.<dataflow> for entities	Dataflow or entity name	DATAFLOW, DATAFLOW_ENTITY
Report, dashboard	Power BI Reports	Workspace name	Item name	REPORT, DASHBOARD
Org app	Power BI Reports	Workspace name	App <app id>	ORG_APP
Dashboard tile	Power BI Reports	<workspace>.<dashboard>	Tile title	TILE
Item listed as inventory only	Its product system	Workspace name	Item name	The Fabric item type

Physical fields

Fabric object	Blindata field	Type
Table or view column	Column name, position and comment	The T-SQL type from the SQL endpoint, such as decimal(10,2); otherwise the Spark type, such as double
KQL column	Column name	KQL type
Semantic model column	Column name, with format, category and sort-by column	Column data type
Measure	Measure name, with its DAX expression as description	MEASURE
Report page, visual	Page or visual name	REPORT_SECTION, or the visual kind

Stored procedures and functions of lakehouses, warehouses and SQL databases are published as system routines with their definition, so Blindata’s SQL lineage can analyze them. View definitions are published as query statements in a query bucket named <system> (Fabric SQL).

Properties

Each entity carries additional properties that tie it back to Fabric. The most useful ones:

Property	Meaning
fabricWorkspaceName, fabricWorkspaceId	Workspace the item belongs to
fabricItemId, fabricItemType	The Fabric item, and its type such as Lakehouse or Notebook
fabricPortalUrl	Link to open the item in the Fabric or Power BI portal
oneLakePath	Path of a table or file in OneLake
shortcutTarget	Where a shortcut points to
sqlEndpointHost	SQL endpoint of the item
fabricMetadataGap	Why some metadata is missing, for example COLUMNS_UNAVAILABLE, VIEW_DEFINITION_UNAVAILABLE, FILES_LISTING_TRUNCATED or ACCESS_DENIED:WORKSPACE_ROLE

Notebooks also list their attached lakehouses, language and environment and, with code analysis on, their kernel, parameters and analysis status.

The connector sets no tags and creates no tasks or logical entities.

Example

A workspace Sales Medallion with three lakehouses (Bronze, Silver, Gold) and a notebook produces four systems. Silver holds shortcuts to the Bronze data, so its entities point back to Bronze instead of copying it.

Re-runs and changes

You can run the job again as often as you like. Each run updates what is already in Blindata instead of adding a second copy.

What changes in Fabric	What happens in Blindata
A workspace or item is renamed	The existing system is renamed. Systems are matched to their Fabric item by ID, so no new system appears.
A table, column or other object changes	The entity or field is updated in place, with no duplicates. Data flows are updated in the same way.
An object is deleted	It stays in Blindata. It is removed only by the optional cleanup, which deletes content of a published system that has not been updated for a number of days. See cleanupOptions .
The application cannot read an item	The item is still published, with a fabricMetadataGap property that says why its content is missing, such as ACCESS_DENIED:WORKSPACE_ROLE.

A system’s description and technology are written only when the system is created. Edit them in Blindata afterwards and later runs keep your changes.

Updated on 13 Oct 2026