Python Microsoft Fabric API - Jobs

run_job, get_job and cancel_job, the job types for notebooks, pipelines and report refreshes, and the statuses.

Running a notebook, running a pipeline and refreshing a report's data are all jobs in Fabric, started, checked and cancelled with the same three methods. Every one takes the workspace ID and the ID of the item the job belongs to.

Job types

Item	Job type	What it does
Notebook	`RunNotebook`	Runs the notebook once
Pipeline	`Pipeline`	Runs the pipeline once
A report's dataset	`DefaultSemanticModelRefresh`	Refreshes the report's data from the tables it reads

Statuses

get_job reports one of these in status:

Status	Meaning
`NotStarted`	Accepted, not yet running
`InProgress`	Running
`Completed`	Ended without error
`Failed`	Ended with an error, details in `failureReason`
`Cancelled`	Stopped with `cancel_job` or from Fabric
`Deduped`	Not started, because the same job was already running

The first two mean the job is still going, everything else means it has ended.

run_job

conn.run_job(workspace_id, item_id, job_type, payload=None)

Starts a job and returns as soon as Fabric has accepted it, without waiting for it to run.

Parameter	Type	Meaning
`workspace_id`	str	The workspace the item is in
`item_id`	str	The notebook, pipeline or dataset
`job_type`	str	One of the job types above
`payload`	dict	Parameters for the job, or none

Returns the job's ID as a string, for get_job and cancel_job.

Parameters travel under executionData. A notebook takes them for its parameter cell, each with a value and a type, a pipeline takes them as plain values:

# -*- coding: utf-8 -*-

# Zato
from zato.server.service import Service

class RunNotebook(Service):

    def handle(self):

        workspace_id = '11111111-2222-3333-4444-555555555555'
        notebook_id = 'aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee'

        conn = self.microsoft.fabric['My Fabric']

        payload = {
            'executionData': {
                'parameters': {
                    'location': {'value': 'Riverside', 'type': 'string'},
                    'days_ahead': {'value': '1', 'type': 'int'},
                }
            }
        }

        job_id = conn.run_job(workspace_id, notebook_id, 'RunNotebook', payload)

        self.response.payload = {'job_id': job_id}

payload = {
    'executionData': {
        'parameters': {
            'source_table': 'admissions',
            'target_table': 'occupancy',
        }
    }
}

job_id = conn.run_job(workspace_id, pipeline_id, 'Pipeline', payload)

A report refresh takes no parameters:

job_id = conn.run_job(workspace_id, dataset_id, 'DefaultSemanticModelRefresh')

get_job

conn.get_job(workspace_id, item_id, job_id)

Returns the job's current state.

Parameter	Type	Meaning
`workspace_id`	str	The workspace the item is in
`item_id`	str	The item the job belongs to
`job_id`	str	The ID `run_job` returned

Returns a dict with id, itemId, jobType, invokeType, status, startTimeUtc, endTimeUtc and, for a failed job, failureReason.

# -*- coding: utf-8 -*-

# Zato
from zato.server.service import Service

class GetJob(Service):

    def handle(self):

        workspace_id = '11111111-2222-3333-4444-555555555555'
        item_id = 'aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee'
        job_id = 'dddddddd-eeee-ffff-0000-111111111111'

        conn = self.microsoft.fabric['My Fabric']

        job = conn.get_job(workspace_id, item_id, job_id)

        self.response.payload = {'status': job['status'], 'started': job['startTimeUtc'], 'ended': job['endTimeUtc']}

{"status": "Completed", "started": "2026-03-12T02:30:04Z", "ended": "2026-03-12T02:34:11Z"}

cancel_job

conn.cancel_job(workspace_id, item_id, job_id)

Asks Fabric to stop a running job. Returns nothing, and the job's status becomes Cancelled once it has stopped, which get_job shows.

Parameter	Type	Meaning
`workspace_id`	str	The workspace the item is in
`item_id`	str	The item the job belongs to
`job_id`	str	The ID `run_job` returned

# -*- coding: utf-8 -*-

# Zato
from zato.server.service import Service

class CancelJob(Service):

    def handle(self):

        workspace_id = '11111111-2222-3333-4444-555555555555'
        item_id = 'aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee'
        job_id = 'dddddddd-eeee-ffff-0000-111111111111'

        conn = self.microsoft.fabric['My Fabric']

        conn.cancel_job(workspace_id, item_id, job_id)

Page	What it covers
Picking up results when a notebook finishes	run_job and get_job across two scheduled services
Running a pipeline and refreshing a report	All three methods in one service, with a timeout
Workspaces	Finding the item IDs the jobs need