Apache Airflow for ML Workflows Quiz

Answers and explanations appear as you take the quiz.

1. In Apache Airflow, what is the core abstraction used to define a workflow?
   • A YAML manifest
   • A SQL view
   • A DAG (Directed Acyclic Graph) written in Python
   • A Jupyter notebook
2. What role does Airflow most commonly play in an ML pipeline?
   • It orchestrates and schedules tasks that run on other systems
   • It trains deep learning models directly on GPUs
   • It stores feature vectors in memory
   • It serves model predictions over HTTP
3. Which Airflow construct is used to wait for an external condition, such as a file arriving in storage?
   • A variable
   • A connection pool
   • A macro
   • A sensor
4. What does an Airflow 'backfill' allow you to do?
   • Delete old DAG runs permanently
   • Run a DAG over a range of past dates to reprocess historical data
   • Encrypt the metadata database
   • Disable the scheduler temporarily
5. Which Airflow component determines when eligible task instances should be scheduled?
   • The scheduler
   • The graph view in the web UI
   • A model’s feature encoder
   • The artifact file itself
6. What is 'data-aware scheduling' introduced in newer Airflow versions?
   • Encrypting all data passed between tasks
   • Automatically labeling training data
   • Reducing the number of GPUs per task
   • Triggering DAGs when a dataset is updated rather than on a fixed clock
7. An Airflow cleanup task must run after upstream tasks finish even if one fails. Which setting controls this behavior?
   • The number of model layers
   • The cleanup task’s trigger rule
   • The feature normalization method
   • The UI theme selected for the DAG
8. A team backfills three months of ML feature data in Airflow. What makes repeating an interrupted task safer?
   • Appending duplicate outputs each time it retries
   • Ignoring the run’s historical date
   • An idempotent operation with the correct historical inputs
   • Changing the training labels on every retry