Implementation:Apache Airflow Execute Workload

Knowledge Sources	Apache Airflow
Domains	Execution, Task_SDK
Last Updated	2026-02-08 00:00 GMT

Overview

Concrete tool for executing task workloads in isolated environments provided by the Task SDK.

Description

The execute_workload function is the entry point for task execution in isolated environments. It receives an ExecuteTask workload containing the task identity, DAG file path, bundle information, JWT token, and log path. The supervise function in the Task SDK manages the actual execution lifecycle including signal handling, heartbeats, and result reporting via the Execution API.

Usage

This is called internally by executors when dispatching tasks. Not typically invoked directly by users.

Code Reference

Source Location

Repository: Apache Airflow
File: task-sdk/src/airflow/sdk/execution_time/execute_workload.py
File: task-sdk/src/airflow/sdk/execution_time/supervisor.py

Signature

def execute_workload(workload: ExecuteTask) -> None:
    """Execute a task workload in the current process."""
    # workload.ti: TaskInstance — task to execute
    # workload.dag_rel_path: str — DAG file path
    # workload.bundle_info — bundle metadata
    # workload.token: str — JWT auth token
    # workload.log_path: str — log output path
    ...

Import

from airflow.sdk.execution_time.execute_workload import execute_workload

I/O Contract

Inputs

Name	Type	Required	Description
workload	ExecuteTask	Yes	Contains task identity, DAG path, JWT token
DAG file	Python file	Yes	DAG source file accessible at dag_rel_path

Outputs

Name	Type	Description
Task result	State	Success or failure state via Execution API
XCom values	JSON	Task return values pushed to API
Logs	Text	Task execution logs written to log_path

Usage Examples

Internal Executor Dispatch

# Internal to executor — not called directly by users
from airflow.sdk.execution_time.execute_workload import execute_workload

workload = ExecuteTask(
    ti=task_instance,
    dag_rel_path="dags/my_dag.py",
    token="jwt-token-here",
    log_path="/logs/my_task.log",
)
execute_workload(workload)

Related Pages

Implements Principle

Principle:Apache_Airflow_Task_Execution_TEI

Requires Environment

Environment:Apache_Airflow_Python_Runtime_Environment

Uses Heuristic

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment