PDE Practice Test 33 — 15 Questions

Question 1

Which TWO statements about designing a data processing pipeline on Google Cloud are correct? (Choose 2.)

Accepted Answer

Cloud Data Fusion allows you to build and manage data pipelines visually without writing code.. Option D is correct because Cloud Data Fusion is a fully managed, code-free ETL/ELT service built on the open-source CDAP framework, providing a visual drag-and-drop pipeline designer and a broad library of connectors and transformations, so pipelines can be built and managed without writing code. Option E is correct because Dataflow, based on Apache Beam, uses a unified programming model in which the same pipeline code can run in batch or streaming mode, with the runner handling windowing, triggers, and watermarks for both. Option A is wrong because Pub/Sub does not guarantee global ordering across all subscribers; ordering is only provided per ordering key within a region when message ordering is explicitly enabled. Option B is wrong because Cloud Bigtable is a wide-column NoSQL database optimized for high-throughput, low-latency reads and writes at scale, not for data warehousing or SQL analytics. Option C is wrong because Dataproc is a managed Apache Hadoop/Spark service for running clusters and jobs, whereas BigQuery is the fully managed data warehousing and analytics service.

Answer

Pub/Sub guarantees message ordering across all subscribers globally.

Answer

Cloud Bigtable is ideal for data warehousing and SQL analytics.

Answer

Dataproc is the best choice for fully managed data warehousing and analytics.

Question 2

Your team has implemented a CI/CD pipeline using Cloud Composer (Apache Airflow) to retrain a model every day. The pipeline reads new data from BigQuery, trains a model using Vertex AI Training, evaluates it, and if the accuracy improves, deploys it to a Vertex AI Endpoint. For the past week, the pipeline has been running successfully but no new model has been deployed because the evaluation accuracy never exceeds the previous model's accuracy. The training data volume has been consistent. You suspect that the model is not learning from the new data. What should you do?

Accepted Answer

Examine the training data for any data quality issues such as missing values or label leakage.. When a model trained on consistent data volume fails to improve accuracy over a week, the most likely root cause is a data quality problem rather than a training capacity problem. Missing values, label leakage, or corrupted features can prevent the model from learning meaningful patterns, causing accuracy to plateau or degrade. Examining the training data for these issues is the correct diagnostic step before changing hyperparameters or deployment strategy.

Answer

Deploy the new model anyway and run an A/B test in production to see if it performs better online.

Answer

Increase the training budget or number of training steps to allow the model to converge better.

Answer

Change the evaluation metric to a different one that may show improvement, such as F1 score instead of accuracy.

Question 3

You are designing a streaming pipeline that needs to handle sudden spikes in traffic without losing data. The pipeline uses Pub/Sub and Dataflow. Which configuration ensures data is not lost if Dataflow falls behind?

Accepted Answer

Use Pub/Sub with a pull subscription and set the message retention duration to 7 days. A Pub/Sub pull subscription with a 7-day message retention duration ensures that if Dataflow falls behind, messages are retained in the subscription backlog for up to 7 days, preventing data loss. Dataflow pulls messages and acknowledges them only after processing, so unacknowledged messages remain available for redelivery. This configuration decouples ingestion from processing and provides a buffer for spikes.

Answer

Use Cloud Pub/Sub Lite with a smaller retention period

Answer

Use Pub/Sub with a push subscription and increase the acknowledgment deadline

Answer

Use Pub/Sub with exactly-once delivery and Dataflow with at-least-once processing

Question 4

A data engineer needs to restrict access to BigQuery datasets such that only data from approved VPC networks can query them. They also need to audit data access. Which two security controls should they implement? (Choose two.)

Accepted Answer

Cloud Audit Logs. VPC Service Controls (E) is correct because it creates a service perimeter around BigQuery that restricts access to only approved VPC networks, blocking requests from outside the perimeter and mitigating data exfiltration. Cloud Audit Logs (A) is correct because it records BigQuery data access and administrative activity (Data Access audit logs must be explicitly enabled), providing the required audit trail. CMEK (B) only controls encryption key ownership, not network-based access or auditing. DLP (C) discovers and classifies sensitive data but does not restrict network access or provide access auditing. IAM roles (D) grant permissions to identities but cannot enforce VPC network origin restrictions on BigQuery.

Answer

Customer-managed encryption keys (CMEK)

Answer

Data Loss Prevention (DLP)

Answer

IAM roles

Question 5

A machine learning engineer needs to deploy a custom TensorFlow model for online predictions with low latency. The model is already trained and saved in SavedModel format. Which Vertex AI service should they use?

Accepted Answer

Vertex AI Prediction. Vertex AI Prediction allows you to deploy custom models (including TensorFlow SavedModel) to an endpoint for online predictions. It supports autoscaling and low-latency serving.

Answer

Vertex AI Workbench

Answer

Vertex AI Feature Store

Answer

Vertex AI AutoML

Question 6

A financial services company uses Cloud Bigtable to store trade data. They are experiencing hot-spotting on a single node, causing high latency. The row key format is [trade_id]#[timestamp]. Which row key design change would BEST distribute writes across tablets?

Accepted Answer

Use a hashed prefix of the trade_id, e.g., [hash(trade_id)]#[trade_id]#[timestamp]. Adding a hashed prefix of the trade_id ensures that writes are evenly distributed across all Bigtable tablets. Bigtable partitions data by row key lexicographic order; without a hash, sequential trade IDs or timestamps cause all recent writes to land on a single tablet, creating a hotspot. The hash spreads the write load uniformly, regardless of the underlying key pattern.

Answer

Use a single row key of [timestamp]

Answer

Increase the number of Bigtable nodes to 20

Answer

Change row key to [timestamp]#[trade_id]

Question 7

You are operating a streaming data pipeline that uses Cloud Pub/Sub and Dataflow. The data source sometimes emits events that are delayed by several minutes due to network issues. Your pipeline must produce accurate aggregations (e.g., counts per minute) even for late data, but you also need to avoid waiting for a long time before emitting results. Which approach should you use?

Accepted Answer

Use event-time processing with allowed lateness and a trigger that fires early to provide speculative results.. It uses event-time processing to handle late data via allowed lateness, combined with early triggers to emit speculative results before the window closes. This balances accuracy for delayed events with low latency for downstream consumers, which is a common requirement in streaming pipelines using Cloud Pub/Sub and Dataflow.

Answer

Use processing-time windows and ignore the event timestamps entirely.

Answer

Use global windows and hold all data for 24 hours before processing to ensure completeness.

Answer

Use event-time processing and discard any data that arrives after the window ends.

Question 8

Your MLOps pipeline uses Vertex AI Pipelines. You want to ensure that model training uses a consistent environment with specific Python package versions. Which approach best achieves this?

Accepted Answer

Build a custom container image with all dependencies and use it in the training step. Building a custom container image with all dependencies ensures a fully deterministic and reproducible environment for model training. Vertex AI Pipelines executes each step as a container, so by pre-installing specific Python package versions into a custom image, you eliminate any risk of version drift or network issues during package installation at runtime. This approach aligns with MLOps best practices for environment consistency and is the most reliable method when exact package versions are critical.

Answer

Include a requirements.txt file in the pipeline step and let Vertex AI install them

Answer

Use a pre-built deep learning container from Deep Learning Containers and install packages at runtime

Answer

Specify the Python version and package versions in the training job configuration

Question 9

A company is evaluating BigQuery for a data warehouse migration. They have a mix of reporting queries and ad-hoc analytical queries. They want to control query costs and prevent runaway queries. Which THREE strategies should they implement?

Accepted Answer

Set a custom quota for concurrent queries. Option B is correct because setting a custom quota for concurrent queries (via BigQuery custom quotas in Cloud Console/IAM) caps the number of simultaneously running queries per project or user, preventing a flood of ad-hoc queries from exhausting slots and causing runaway costs. Option C is correct because partitioning (e.g., by ingestion/date column) and clustering (e.g., by frequently filtered columns) prune the data scanned, directly reducing bytes processed and therefore on-demand query cost. Option E is correct because BigQuery reservations with flex slots (or committed slots) let predictable reporting workloads run on dedicated capacity with fixed pricing, isolating them from unpredictable ad-hoc on-demand costs and giving cost predictability. Option A is not correct because authorized views control data access/visibility, not query cost or runaway query prevention. Option D is not correct because materialized views can accelerate specific recurring queries but do not by themselves control costs or prevent runaway queries across a mixed workload.

Answer

Grant authorized view access to limit data visibility

Answer

Create materialized views for all reporting queries

Question 10

A retail company needs to generate product recommendations for millions of users every few hours. The model is a small scikit-learn model. Which prediction method should be used to minimize infrastructure cost while meeting the latency requirements?

Accepted Answer

Use a Vertex AI batch prediction job that reads from BigQuery and writes results back to BigQuery or Cloud Storage.. Batch prediction is the most cost-effective approach for generating recommendations for millions of users every few hours. Vertex AI batch prediction jobs process large datasets in parallel without maintaining always-on infrastructure, and they can read from BigQuery and write results directly to BigQuery or Cloud Storage, minimizing compute costs while meeting the latency requirement of 'every few hours' (not real-time).

Answer

Use Cloud Run to host the model and invoke it for each user request.

Answer

Export the model as a container and run on Google Kubernetes Engine with cluster autoscaling.

Answer

Deploy the model to a Vertex AI endpoint with a single replica for online predictions.

Question 11

You are loading a CSV file into BigQuery. The file has a column `price` that contains values like `$1,234.56`. You want to load it into a NUMERIC column. What should you do?

Accepted Answer

Preprocess the CSV file to remove the dollar signs and commas before loading, or use a transformation in a data pipeline.. BigQuery requires numeric columns to contain plain numeric values without formatting characters. To load a CSV with values like `$1,234.56` into a NUMERIC column, you must remove the dollar signs and commas, either by preprocessing the file or using a data transformation step. Options that suggest loading as-is or using unrelated options will fail.

Answer

Use the `--null_marker` option to treat the dollar sign as null.

Answer

Specify the schema with `price` as a NUMERIC and set the `--allow_quoted_newlines` option.

Answer

Load the data as a STRING and then use a CAST to NUMERIC in a query to create a new table.

Question 12

You are setting up Dataplex data quality rules for a BigQuery table. You want to define rules that check for non-null values in key columns and also validate that a column's values fall within a certain range. Which TWO rule types must you use? (Choose 2)

Accepted Answer

Row rule (e.g., not null). Dataplex data quality rules include row rules (for null checks) and column rules (for range or value checks). Table rules apply to the entire table; custom SQL can be used but row and column rules are the standard.

Answer

Table rule (e.g., row count)

Answer

Partition rule

Answer

Custom SQL rule

Question 13

Your company is building a real-time fraud detection system using Google Cloud. Transactions are streamed into Pub/Sub, and you need to process them with low latency (under 100ms per event) and aggregate data over sliding windows. Which Google Cloud service is best suited for this processing logic?

Accepted Answer

Dataflow. Dataflow is the best choice because it provides a unified stream and batch processing model with native support for Pub/Sub, exactly-once semantics, and low-latency sliding window aggregations. Its autoscaling and millisecond-level checkpointing enable sub-100ms per event processing, which is critical for real-time fraud detection.

Answer

BigQuery streaming inserts with scheduled queries

Answer

Dataproc with Spark Streaming

Answer

Cloud Functions

Question 14

You need to create a time-series forecast for inventory demand using BigQuery ML. The data includes daily sales for 5 years. Which model type should you use?

Accepted Answer

ARIMA_PLUS. BigQuery ML supports ARIMA_PLUS for time-series forecasting. Linear regression, k-means, and matrix factorization are not appropriate for time-series forecasting.

Answer

K-means

Answer

Linear regression

Answer

Matrix factorization

Question 15

You are using Cloud Workflows to orchestrate a series of API calls. You need to handle errors and retries. Which THREE features of Cloud Workflows can you use? (Choose THREE.)

Accepted Answer

Use try/except blocks to catch and handle errors.. Option A is correct because Cloud Workflows supports try/except blocks, allowing you to catch errors raised by a step and execute fallback logic or handle the failure gracefully. Option C is correct because Cloud Workflows supports conditional branching with if/else constructs, letting you evaluate step results and choose different execution paths based on those outcomes. Option D is correct because Cloud Workflows lets you attach a retry policy to a step, specifying parameters such as max_retries, initial_delay, max_delay, and multiplier to automatically retry failed calls. Option B is not correct because Cloud Load Balancing is a traffic-distribution service for workloads, not a Cloud Workflows error-handling or retry feature. Option E is not correct because logging is handled by Cloud Logging integration and is not a configurable per-step error-handling or retry feature of Cloud Workflows.

Answer

Integrate with Cloud Load Balancing for high availability.

Answer

Enable automatic logging for each step.