DAG concept (Directed Acyclic Graph) in Apache Airflow - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding DAG Structure in Airflow

Which statement best describes the structure of a DAG in Apache Airflow?

A A DAG is a collection of tasks with dependencies that form a cycle allowing repeated execution paths.

B A DAG is a linear sequence of tasks executed one after another without any branching.

C A DAG is a set of tasks connected in a way that no cycles exist, ensuring a clear start and end.

D A DAG is a random graph where tasks can depend on each other in any order.

💡 Hint

Think about what 'acyclic' means in the context of graph structures.

A DAG (Directed Acyclic Graph) means tasks are connected with directed edges and no cycles exist, so the workflow has a clear direction and no loops.

💻 Command Output

intermediate

Output of DAG Task Execution Order

Given the following Airflow DAG snippet, what is the order of task execution?

Apache Airflow

from airflow import DAG
from airflow.operators.dummy import DummyOperator
from datetime import datetime

dag = DAG('example_dag', start_date=datetime(2024, 1, 1))

t1 = DummyOperator(task_id='task1', dag=dag)
t2 = DummyOperator(task_id='task2', dag=dag)
t3 = DummyOperator(task_id='task3', dag=dag)

t1 >> [t2, t3]

A task1 runs first, then task2 and task3 run in parallel.

B task3 runs first, then task1 and task2 run in parallel.

C task2 runs first, then task1 and task3 run in parallel.

D task1, task2, and task3 run all at the same time.

💡 Hint

Look at the dependency operator '>>' and what it means for task order.

The '>>' operator means task1 must finish before task2 and task3 start. Tasks 2 and 3 can run at the same time after task1.

❓ Configuration

advanced

Identifying Invalid DAG Configuration

Which of the following DAG definitions will cause an error due to cycle detection in Airflow?

Apache Airflow

from airflow import DAG
from airflow.operators.dummy import DummyOperator
from datetime import datetime

dag = DAG('cycle_dag', start_date=datetime(2024, 1, 1))

t1 = DummyOperator(task_id='task1', dag=dag)
t2 = DummyOperator(task_id='task2', dag=dag)
t3 = DummyOperator(task_id='task3', dag=dag)

# Dependencies to check

A [t1, t2] >> t3

B t1 >> t2 >> t3

C t1 >> [t2, t3]

D t1 >> t2 >> t3 >> t1

💡 Hint

Check if any task depends on itself directly or indirectly.

Option D creates a cycle because task1 depends on task3, which depends back on task1, causing a loop. Airflow does not allow cycles in DAGs.

❓ Troubleshoot

advanced

Troubleshooting a DAG with Missing Task Dependencies

You have a DAG where task3 runs before task2, but your code specifies t1 >> t2 >> t3. What is the most likely cause?

A The DAG file is not saved, so Airflow is running an old version without dependencies.

B Airflow allows tasks to run in any order regardless of dependencies.

C The start_date is set in the future, so tasks run out of order.

D The tasks are defined without a DAG object.

💡 Hint

Think about how Airflow reads DAG files and updates task dependencies.

If the DAG file is not saved after adding dependencies, Airflow will run the last saved version which may not have the correct order, causing tasks to run out of sequence.

🔀 Workflow

expert

Determining the Number of Tasks Executed in a Complex DAG

Consider a DAG with tasks: t1, t2, t3, t4, t5. The dependencies are:

• t1 >> t2
• t1 >> t3
• t2 >> t4
• t3 >> t4
• t4 >> t5

If task t3 fails and depends_on_past=False, how many tasks will run successfully in the next DAG run?

A 3 tasks will run successfully: t1, t2, and t5.

B 2 tasks will run successfully: t1 and t2.

C 4 tasks will run successfully: t1, t2, t3, and t4.

D Only 1 task will run successfully: t1.

💡 Hint

Consider which tasks depend on the failed task and how failure affects downstream tasks.

Task t3 fails, so t4 cannot run because it depends on both t2 and t3. t5 depends on t4, so it also won't run. Only t1 and t2 run successfully.