Shortcuts in Microsoft OneLake allow you to unify your data across domains, clouds, and accounts by creating a single virtual data lake for your entire enterprise. All Fabric experiences and analytical engines can directly connect to your existing data sources such as Azure, Amazon Web Services (AWS), and OneLake through a unified namespace. OneLake manages all permissions and credentials, so you don't need to separately configure each Fabric experience to connect to each data source. Additionally, you can use shortcuts to eliminate edge copies of data and reduce process latency associated with data copies and staging.
What are shortcuts?
Shortcuts are objects in OneLake that point to other storage locations. The location can be internal or external to OneLake. The location that a shortcut points to is known as the target path of the shortcut. The location where the shortcut appears is known as the shortcut path. Shortcuts appear as folders in OneLake and any experience or service that has access to OneLake can use them. Shortcuts behave like symbolic links. They're an independent object from the target. If you delete a shortcut, the target remains unaffected. If you move, rename, or delete a target path, the shortcut can break.
Where can I create shortcuts?
You can create shortcuts in lakehouses and Kusto Query Language (KQL) databases. Furthermore, the shortcuts you create within these items can point to other OneLake locations, Azure Data Lake Storage (ADLS) Gen2, Amazon S3 storage accounts, or Dataverse.
When creating shortcuts in a lakehouse, you must understand the folder structure of the item. Lakehouses are composed of two top level folders: the Tables folder and the Files folder. The Tables folder represents the managed portion of the lakehouse, while the Files folder is the unmanaged portion of the lakehouse. In the Tables folder, you can only create shortcuts at the top level. Shortcuts aren't supported in other subdirectories of the Tables folder. If the target of the shortcut contains data in the Delta\Parquet format, the lakehouse automatically synchronizes the metadata and recognizes the folder as a table. In the Files folder, there are no restrictions on where you can create shortcuts. You can create them at any level of the folder hierarchy. Table discovery doesn't happen in the Files folder.
When you create a shortcut in a KQL database, it appears in the Shortcuts folder of the database. The KQL database treats shortcuts like external tables. To query the shortcut, use the
external_table function of the Kusto Query Language.
Where can I access shortcuts?
Any Fabric or non-Fabric service that can access data in OneLake can use shortcuts. Shortcuts are transparent to any service accessing data through the OneLake API. Shortcuts just appear as another folder in the lake. Spark, SQL, Real-Time Analytics, and Analysis Services can all use shortcuts when querying data.
Spark notebooks and Spark jobs can use shortcuts that you create in OneLake. Relative file paths can be used to directly read data from shortcuts. Additionally, if you create a shortcut in the Tables section of the lakehouse and it is in the Delta format, you can read it as a managed table using Spark SQL syntax.
df = spark.read.format("delta").load("Tables/MyShortcut") display(df)
df = spark.sql("SELECT * FROM MyLakehouse.MyShortcut LIMIT 1000") display(df)
The Delta format doesn't support tables with space characters in the name. Any shortcut containing a space in the name won't be discovered as a Delta table in the lakehouse.
You can also read shortcuts in the Tables section of a lakehouse through the SQL analytics endpoint for the lakehouse. You can access the SQL analytics endpoint through the mode selector of the lakehouse or through SQL Server Management Studio (SSMS).
SELECT TOP (100) * FROM [MyLakehouse].[dbo].[MyShortcut]
Shortcuts in KQL databases are recognized as external tables. To query the shortcut, use the
external_table function of the Kusto Query Language.
external_table('MyShortcut') | take 100
KQL databases don't currently support data in the Delta format. Tables in a KQL database only export to OneLake as Parquet files. Shortcuts in KQL databases that contain Delta-formatted data in the target aren't recognized as tables.
You can create semantic models for lakehouses containing shortcuts in the Tables section of the lakehouse. When the semantic model runs in Direct Lake mode, Analysis Services can read data directly from the shortcut.
Applications and services outside of Fabric can also access shortcuts through the OneLake API. OneLake supports a subset of the ADLS Gen2 and Blob storage APIs. To learn more about the OneLake API, see OneLake access with APIs.
Types of shortcuts
OneLake shortcuts support multiple filesystem data sources. These include internal OneLake locations, Azure Data Lake Storage (ADLS) Gen2, Amazon S3, and Dataverse.
Internal OneLake shortcuts
Internal OneLake shortcuts allow you to reference data within existing Fabric items. These items include lakehouses, KQL databases and data warehouses. The shortcut can point to a folder location within the same item, across items within the same workspace or even across items in different workspaces. When you create a shortcut across items, the item types don't need to match. For instance, you can create a shortcut in a lakehouse that points to data in a data warehouse.
When a user accesses data through a shortcut to another OneLake location, the identity of the calling user is used to authorize access to the data in the target path of the shortcut*. This user must have permissions in the target location to read the data.
When accessing shortcuts through Power BI semantic models or T-SQL, the calling user’s identity is not passed through to the shortcut target. The calling item owner’s identity is passed instead, delegating access to the calling user.
Shortcuts can also be created to ADLS Gen2 storage accounts. When you create shortcuts to ADLS, the target path can point to any folder within the hierarchical namespace. At a minimum, the target path must include a container name.
ADLS shortcuts must point to the DFS endpoint for the storage account.
Access to storage account endpoint can't be blocked by storage firewall or VNET.
ADLS shortcuts use a delegated authorization model. In this model, the shortcut creator specifies a credential for the ADLS shortcut and all access to that shortcut is authorized using that credential. The supported delegated types are Organizational account, Account Key, Shared Access Signature (SAS), and Service Principal.
- Organizational account - must have Storage Blob Data Reader, Storage Blob Data Contributor, or Storage Blob Data Owner role on storage account
- Shared Access Signature (SAS) - must include at least the following permissions: Read, List, and Execute
- Service Principal - must have Storage Blob Data Reader, Storage Blob Data Contributor, or Storage Blob Data Owner role on storage account
You can also create shortcuts to Amazon S3 accounts. When you create shortcuts to Amazon S3, the target path must contain a bucket name at a minimum. S3 doesn't natively support hierarchical namespaces but you can use prefixes to mimic a directory structure. You can include prefixes in the shortcut path to further narrow the scope of data accessible through the shortcut. When you access data through an S3 shortcut, prefixes are represented as folders.
S3 shortcuts must point to the https endpoint for the S3 bucket.
You do not need to disable the S3 Block Public Access setting for your S3 account for the S3 shortcut to function.
Access to the S3 endpoint must not be blocked by a storage firewall or Virtual Private Cloud.
S3 shortcuts use a delegated authorization model. In this model, the shortcut creator specifies a credential for the S3 shortcut and all access to that shortcut is authorized using that credential. The supported delegated credential is a Key and Secret for an IAM user.
The IAM user must have the following permissions on the bucket that the shortcut is pointing to.
S3 shortcuts are read-only. They don't support write operations regardless of the permissions for the IAM user.
Dataverse direct integration with Microsoft Fabric enables organizations to extend their Dynamics 365 enterprise applications and business processes into Fabric. The View in Microsoft Fabric feature, which is built into the PowerApps maker portal, makes all your Dynamics 365 data available for analysis in Microsoft Fabric. For more information, see Dataverse direct integration with Microsoft Fabric.
Dataverse shortcuts can't be created through the Fabric UX. They must be created through the PowerApps maker portal.
Dataverse shortcuts use a delegated authorization model. In this model, the shortcut creator specifies a credential for the Dataverse shortcut and all access to that shortcut is authorized using that credential. The supported delegated credential type is Organizational account (OAuth2). The organizational account must have permissions to access data in Dataverse Managed Lake.
Service Principals are currently not supported for Dataverse shortcut authorization.
How shortcuts utilize cloud connections
ADLS and S3 shortcut authorization is delegated by using cloud connections. When you create a new ADLS or S3 shortcut, you either create a new connection or select an existing connection for the data source. Setting a connection for a shortcut is a bind operation. Only users with permission on the connection can perform the bind operation. If you don't have permissions on the connection, you can't create new shortcuts using that connection.
A combination of the permissions in the shortcut path and the target path governs the permissions for shortcuts. When a user accesses a shortcut, the most restrictive permission of the two locations is applied. Therefore, a user that has read/write permissions in the lakehouse but only read permissions in the shortcut target can't write to the shortcut target path. Likewise, a user that only has read permissions in the lakehouse but read/write in the shortcut target also can't write to the shortcut target path.
The following table shows the shortcut-related permissions for each workspace role. For more information, see Workspace roles.
|Create a shortcut||Yes1||Yes1||Yes1||-|
|Read file/folder content of shortcut||Yes2||Yes2||Yes2||-|
|Write to shortcut target location||Yes3||Yes3||Yes3||-|
|Read data from shortcuts in table section of the lakehouse via TDS endpoint||Yes||Yes||Yes||Yes|
1 Users must have a role that provides write permission the shortcut location and at least read permission in the target location.
2 Users must have a role that provides read permission both in the shortcut location and the target location.
3 Users must have a role that provides write permission both in the shortcut location and the target location.
How do shortcuts handle deletions?
Shortcuts don't perform cascading deletes. When you perform a delete operation on a shortcut, you only delete the shortcut object. The data in the shortcut target remains unchanged. However, if you perform a delete operation on a file or folder within a shortcut, and you have permissions in the shortcut target to perform the delete operation, the files and/or folders are deleted in the target. The following example illustrates this point.
User A has a lakehouse with the following path in it:
MyShortcut is a shortcut that points to an ADLS Gen2 account that contains the Foo\Bar directories.
Deleting a shortcut object
User A performs a delete operation on the following path:
In this case, MyShortcut is deleted from the lakehouse. Shortcuts don't perform cascading deletes, therefore the files and directories in the ADLS Gen2 account Foo\Bar remain unaffected.
Deleting content referenced by a shortcut
User A performs a delete operation on the following path:
In this case, if User A has write permissions in the ADLS Gen2 account, the Bar directory is deleted from the ADLS Gen2 account.
Workspace lineage view
When creating shortcuts between multiple Fabric items within a workspace, you can visualize the shortcut relationships through the workspace lineage view. Select the Lineage view button ( ) in the upper right corner of the Workspace explorer.
The lineage view is scoped to a single workspace. Shortcuts to locations outside the selected workspace won't appear.
Limitations and considerations
- The maximum number of shortcuts per Fabric item is 10,000. In this context, the term item refers to: apps, lakehouses, warehouses, reports, and more.
- The maximum number of shortcuts in a single OneLake path is 10.
- The maximum number of direct shortcuts to shortcut links is 5.
- ADLS and S3 shortcut target paths can't contain any reserved characters from RCF 3986 section 2.2.
- OneLake shortcut target paths can't contain "%" characters.
- Shortcuts don't support non-Latin characters.
- Copy Blob API not supported for ADLS or S3 shortcuts.
- Copy function doesn't work on shortcuts that directly point to ADLS containers. It's recommended to create ADLS shortcuts to a directory that is at least one level below a container.
- OneLake shortcuts pointing to ADLS or S3 shortcuts isn't supported.
- Additional shortcuts can't be created inside ADLS or S3 shortcuts.
- Lineage for shortcuts to Data Warehouses and Semantic Models is not currently available.