Creating custom operators in Apache Airflow - Visual Walkthrough

Process Flow - Creating custom operators

Define Custom Operator Class

↓

Inherit from BaseOperator

↓

Override execute() Method

↓

Use Custom Operator in DAG

↓

Airflow Scheduler Runs DAG

↓

Custom Operator execute() Runs

↓

Task Completes or Fails

This flow shows how to create a custom operator by defining a class, overriding its execute method, then using it in a DAG where Airflow runs it.

Execution Sample

Apache Airflow

from airflow.models import BaseOperator

class MyOperator(BaseOperator):
    def execute(self, context):
        print('Hello from MyOperator')

Defines a custom operator class that prints a message when executed.

Process Table

Step	Action	Evaluation	Result
1	Define MyOperator class inheriting BaseOperator	Class created	Ready to use in DAG
2	Override execute() method	Method defined	Prints message when run
3	Instantiate MyOperator in DAG	Task created	Task ready for scheduling
4	Airflow scheduler triggers task	Task starts	execute() method runs
5	execute() prints message	Output: Hello from MyOperator	Task completes successfully
6	Task status updated	Success recorded	DAG continues or ends

💡 Task completes after execute() method runs and prints message

Status Tracker

Variable	Start	After Step 2	After Step 3	After Step 5	Final
MyOperator class	Not defined	Defined with execute()	Instance created	execute() runs	Task done
execute() output	None	Defined	None	'Hello from MyOperator'	Printed

Key Moments - 2 Insights

Why do we override the execute() method in the custom operator?

Because the execute() method contains the code that runs when the task executes, as shown in execution_table step 4 and 5.

Can we use the custom operator without inheriting BaseOperator?

No, inheriting BaseOperator is required to integrate with Airflow's task system, as shown in execution_table step 1.

Visual Quiz - 3 Questions

Look at the execution table, what happens at step 4?

A The Airflow scheduler triggers the task and execute() runs

B The custom operator class is defined

C The task status is updated to success

D The execute() method is overridden

💡 Hint

Refer to execution_table row with Step 4 describing task start and execute() running

According to variable_tracker, what is the output of execute() after step 5?

A None

B 'Hello from MyOperator'

C Task instance

D Class definition

💡 Hint

Check execute() output column after Step 5 in variable_tracker

If we do not inherit BaseOperator, what will happen according to the key moments?

A The task will run normally

B execute() will run twice

C Airflow will not recognize the custom operator as a task

D The DAG will fail to parse

💡 Hint

See key_moments about inheritance requirement for Airflow integration

Concept Snapshot

Creating custom operators in Airflow:
- Define a class inheriting BaseOperator
- Override execute(self, context) method
- Put task logic inside execute()
- Use the operator in DAGs like any other task
- Airflow runs execute() when task runs
- Enables reusable, custom task behavior

Full Transcript

To create a custom operator in Airflow, you start by defining a new class that inherits from BaseOperator. This inheritance is necessary so Airflow can treat your class as a task. Next, you override the execute() method in your class. This method contains the code that runs when the task executes. For example, you can print a message inside execute(). After defining your custom operator, you use it in your DAG by creating an instance of it as a task. When the Airflow scheduler runs your DAG, it triggers your custom operator's execute() method. The task runs, prints the message, and completes successfully. This process allows you to create reusable, custom task logic integrated with Airflow's scheduling and execution system.