Data & Analytics Practice Questions
Warehousing, streaming, pipelines and cost control. 8 original questions with full explanations.
Question 1 of 8
0 answered · 8 remaining
Data & Analyticsmedium
Messages that repeatedly fail processing are being retried indefinitely, blocking the subscription. What should be configured?
Select an answer to see the explanation.
These questions are written by GCP Prep to test understanding of publicly documented concepts. They are not real exam questions.
What this test covers
One concept per question. If you can explain each of these without hesitating, you are in good shape on Data & Analytics.
- Dead-letter topics isolate poison messages so the stream keeps moving
- Partitioning is what actually reduces bytes scanned in BigQuery
- Late data handling is governed by allowed lateness and triggers
- Clustering sorts data within partitions so filters on clustering columns skip blocks
- Columnar storage means you pay for the columns you read
- At-least-once delivery makes idempotent consumers a requirement, not an optimisation
- Authorised views let someone query derived results without access to the source
- Streaming ingestion through a message queue is the standard near-real-time analytics path
Other topics
Related Certifications
Related Cloud Topics
- BigQueryA serverless analytics warehouse: partitioning, clustering, the cost model, and how to make queries fast and cheap.
- Dataflow and Pipeline ProcessingBatch and stream processing in one model: windowing, watermarks, late data and the trade-offs that make streaming hard.
- Pub/Sub and MessagingAsynchronous messaging that decouples systems: topics, subscriptions, delivery guarantees, ordering and dead-letter handling.