How to Mark a Task as Success in Airflow

In Airflow, you can mark a task as success by calling task_instance.set_state("success"). This is typically done inside a task's Python callable or via Airflow CLI commands.

Syntax

To mark a task as success programmatically, use the set_state method on the task instance object with the argument "success". This changes the task's state in the Airflow metadata database.

• task_instance: The current task instance object.
• set_state(state): Method to set the task state, where state is a string like "success".

python

task_instance.set_state("success")

Example

This example shows how to mark a task as success inside a PythonOperator's callable function by accessing the task instance from the context and setting its state.

python

from airflow import DAG
from airflow.operators.python import PythonOperator
from airflow.utils.dates import days_ago

def mark_success(**context):
    ti = context['ti']
    ti.set_state("success")

with DAG(dag_id="example_mark_success", start_date=days_ago(1), schedule_interval=None) as dag:
    task = PythonOperator(
        task_id="mark_task_success",
        python_callable=mark_success
    )

Output

Task instance state is set to success in Airflow metadata after task runs.

Common Pitfalls

Common mistakes when marking a task as success include:

• Trying to set the state outside of a running task context, which causes errors because task_instance is not available.
• Not importing or passing the task_instance correctly in the callable.
• Using deprecated methods or manipulating the database directly instead of using set_state.

Always use the Airflow context and APIs to change task states safely.

python

def wrong_mark_success():
    # This will fail because no task_instance context is available
    task_instance.set_state("success")

# Correct way inside a task callable

def correct_mark_success(**context):
    ti = context['ti']
    ti.set_state("success")

Quick Reference

Summary tips to mark a task as success in Airflow:

• Use task_instance.set_state("success") inside task code.
• Access task_instance via the context['ti'] parameter.
• Do not manipulate task states outside of Airflow's API.
• Use Airflow CLI airflow tasks mark_success for manual marking.

Key Takeaways

Use task_instance.set_state("success") inside task code to mark success.

Access the task instance via context parameter 'ti' in PythonOperator.

Avoid changing task states outside Airflow APIs to prevent errors.

Airflow CLI can also mark tasks as success manually.

Always run state changes within a valid task context.