Professional Data EngineerDesigning data processing systemsHard

A global ride-sharing company processes millions of sensor data points per second from vehicles, including GPS coordinates, accelerometer readings, and engine diagnostics. This data is critical for real-time fleet management, predictive maintenance, and driver safety analytics. The company requires a database that can handle extremely high write throughput, efficient time-series queries, and scale globally to petabytes of data with minimal latency. Which Google Cloud database is best suited for this demanding workload?

  1. ACloud Bigtable
  2. BFirestore
  3. CBigQuery
  4. DCloud SQL
Show answer & explanation

Correct answer: A. Cloud Bigtable

Cloud Bigtable is specifically designed for very large analytical and operational workloads that require high throughput and low latency, making it ideal for time-series data like sensor readings. Its wide-column NoSQL model excels at handling petabytes of data with consistent performance for both writes and time-series-based reads, perfectly fitting the requirements for real-time fleet management and predictive maintenance.

Why the other options are wrong

  • B. Firestore is a document database suitable for web/mobile applications, but not built for the petabyte-scale, high-throughput time-series data processing described.
  • C. BigQuery is an analytical data warehouse, excellent for complex queries over large datasets, but not designed for the extremely high-frequency, low-latency writes and operational reads of real-time time-series data.
  • D. Cloud SQL is a relational database not optimized for the extreme write throughput and petabyte-scale time-series data access required.

Cloud Bigtable for Time-Series Data

Cloud Bigtable is a petabyte-scale, low-latency, wide-column NoSQL database optimized for large analytical and operational workloads, especially well-suited for time-series data, IoT, and high-throughput applications.

  • Handles millions of writes/second.
  • Consistent low-latency performance.
  • Ideal for time-series, IoT, and operational analytics.
  • Scales to petabytes of data.

Memory trick: Bigtable: The 'Big Clock' for your time-series data.

More Designing data processing systems questions