Answer-first summary for fast verification
Answer: Vehicles write data directly to Google Cloud Pub/Sub
The correct answer is B (Vehicles write data directly to Google Cloud Pub/Sub) based on the high data volume requirements and community consensus. The data ingestion rate of 20 million records per second (40 TB/hour) exceeds BigQuery's streaming insertion quotas (1 GB/second without deduplication), making option C unsuitable. Pub/Sub is designed for high-throughput, scalable message ingestion and acts as a buffer to handle spikes, decoupling data collection from processing. Option A (direct to GCS) is inefficient for real-time streaming due to high GET request costs and lack of message queuing. Option D (existing FTP system) cannot scale to handle the increased volume. Community discussion strongly supports B, with 60% of votes and high-upvoted comments citing Pub/Sub's scalability and BigQuery quota limitations.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
TerramEarth intends to connect its entire fleet of 20 million field vehicles to the cloud. This will generate a data volume of 20 million records per second, with each record being 600 bytes, resulting in an ingestion rate of 40 TB per hour.
How should the data ingestion pipeline be designed?
A
Vehicles write data directly to GCS
B
Vehicles write data directly to Google Cloud Pub/Sub
C
Vehicles stream data directly to Google BigQuery
D
Vehicles continue to write data using the existing system (FTP)
No comments yet.