mediumMultiple ChoiceObjective-mapped
PDE Practice Question: A streaming Dataflow pipeline ingests events from…
A streaming Dataflow pipeline ingests events from Cloud Pub/Sub and writes to BigQuery. The event schema evolves occasionally (new columns added). The pipeline fails when new columns appear. What is the best long-term solution?
⚠ Common exam trap
Google Cloud often tests the misconception that manual schema updates or external transformations are acceptable long-term solutions, when in fact the correct answer leverages a built-in BigQuery feature (schema auto-detection) to handle schema evolution dynamically without pipeline downtime.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Configure the BigQuery sink to use stored 'dynamic' schema by setting create_disposition to CREATE_NEVER and writing to a temporary table with schema auto-detection
It leverages BigQuery's schema auto-detection with a temporary table to handle schema evolution dynamically. By setting create_disposition to CREATE_NEVER, the pipeline writes to a table that already exists, while the temporary table with auto-detection allows the pipeline to infer new columns from the incoming data. This approach avoids pipeline failures when new columns appear, as the sink can adapt without manual intervention or pipeline restarts.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Configure the BigQuery sink to use stored 'dynamic' schema by setting create_disposition to CREATE_NEVER and writing to a temporary table with schema auto-detection
Why this is correct
Using schema auto-detection on a temporary table and then merging into the main table with wildcard tables or using BigQuery's schema flexibility can handle new columns.
- ✗
Stop the pipeline and update the BigQuery schema manually whenever a new column appears
Why it's wrong here
Manual intervention is error-prone and not scalable for frequent changes.
- ✗
Switch to Dataproc to process the data with Spark and write to BigQuery using the Avro format
Why it's wrong here
This adds complexity and does not inherently solve schema evolution without custom handling.
- ✗
Use a Cloud Function to transform the data and add null columns for missing fields
Why it's wrong here
This approach requires maintaining the schema mapping in the function and still fails if new columns are unknown.
Go deeper
Related to this question
About these practice questions
One of 890 original PDE practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This PDE practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the PDE exam.