Schedule a user data function

The job scheduler in Microsoft Fabric enables you to run user data functions (UDFs) on a recurring schedule without requiring a pipeline, notebook, or external orchestration service. You can configure schedules directly on a user data function, specify recurrence patterns, pass function parameters, and monitor executions from the Fabric monitoring experience.

Prerequisites

Before creating a schedule, ensure that:

• You have a published user data function.
• You have Contributor or higher permissions on the workspace.
• The function is tested successfully by using manual invocation.
• You configure any required connections, secrets, or dependencies.

Use cases

Schedule user data functions to automate recurring business and operational tasks. Common scenarios include:

• Data quality validation – Run checks on datasets before reporting or downstream processing.
• Customer feedback processing – Analyze new feedback, generate sentiment scores, and store results.
• Business event generation – Detect changes in operational data and publish business events for downstream applications.
• Scheduled cleanup – Archive old records, remove temporary files, or perform routine maintenance tasks.
• Periodic data synchronization – Sync data from external systems into Fabric on a recurring schedule.
• Automated notifications – Generate alerts or notifications based on business rules and thresholds.

Open the scheduler

1. Open your Fabric workspace.
2. Open the user data function item.
3. Select Settings.
4. Select Schedule.

The Schedule page enables you to:

• Create schedules
• Configure failure notifications
• View scheduled executions
• Edit or disable schedules
• Manually run scheduled jobs

Create a schedule

1. On Schedule, select Add schedule.

2. Configure the schedule properties.

   The scheduler supports the following settings:

   Setting	Description
   Controller	Variable used to control execution from variable library
   Repeat	Defines the recurrence pattern
   Interval	Frequency interval
   Start date and time	When the schedule begins
   End date and time	When the schedule expires
   Time zone	Time zone for schedule execution
   Parameters	Input parameters passed to the function

   Select how often the function should run and provide the scheduling details. The scheduler supports recurring executions such as every few minutes, hourly, daily, weekly, or monthly. For example, to run a function every 15 minutes, configure:

3. Provide parameter for FunctionName. This parameter is required to define the function that you're calling.

   Parameter name	Type	Value
   FunctionName	String	name of the functions, for example hello_fabric.

4. Add more parameters for the function inputs. For example, for hello_fabric(name:str), add a parameter name of type string.

   Parameter name	Type
   name	String

   Important

   Parameter names must exactly match the user data function signature. Parameter names are case-sensitive.

5. Save the schedule.

6. Add more schedules for other functions in your user data function item.

Run a scheduled function manually

To test a scheduled function immediately, select Run. Fabric submits an execution without waiting for the next scheduled interval. Manual execution is useful for validating parameters, verifying connections, and testing function logic.

Monitor scheduled executions

You can monitor user data function executions from the Fabric monitoring experience.

1. Open Monitor Hub.
2. Select Activities.
3. Filter by User data functions.

Use the monitoring experience to troubleshoot failed executions and validate successful runs.

Configure failure notifications

You can configure email notifications for failed scheduled runs.

1. Open the Schedule page.
2. In Failure notifications, enter one or more email addresses.
3. Save the configuration.

Fabric sends email notifications whenever a scheduled execution fails.

Next steps

• Learn about user data functions.
• Learn about the job scheduler in Fabric.
• Monitor and troubleshoot Fabric workloads using the Monitor hub.