Question

Difficulty: MediumPlanning Database Options and Persistent Disk Types

A healthcare analytics application hosted on Compute Engine requires a managed database service to store high-throughput, real-time IoT device telemetry metrics. The application processes high volumes of non-relational time-series data with unpredictable write spikes and requires sub-millisecond write latency without needing complex relational joins. Which Google Cloud database service should you recommend to meet these performance requirements with minimal operational overhead?

  1. Cloud BigtableAnswer
  2. B
    Cloud SQL for PostgreSQL
  3. C
    Compute Engine Spot VMs with attached Local SSDs running self-managed Cassandra
  4. D
    Cloud Storage Coldline storage class bucket

Answer

Cloud Bigtable is the recommended service for high-throughput, low-latency, non-relational time-series telemetry metrics.
Cloud Bigtable is Google Cloud's fully managed NoSQL wide-column database service optimized for petabyte-scale, high-throughput analytics, time-series data, and real-time streaming metrics with microsecond-to-millisecond latency.

Step-by-Step Solution

1
Analyze the workload requirements
Identified key constraints: high-throughput time-series data, non-relational key-value access pattern, sub-millisecond write latency, and minimal operational overhead.
Matching technical requirements to the appropriate Google Cloud storage/database solution ensures optimal performance and cost-efficiency.
2
Evaluate database engine paradigms
Cloud Bigtable is built specifically for large-scale NoSQL key-value and time-series data under high read/write load.
Relational databases like Cloud SQL struggle to deliver scalable sub-millisecond write latency for massive IoT streaming metrics without severe performance degradation or high cost.

Key Concept

Selecting Cloud Bigtable for high-throughput, non-relational time-series and IoT telemetry workloads.
Rate this question