Examkin

Question

Difficulty: HardPlanning Database Options and Persistent Disk Types

A logistics enterprise is designing a high-throughput fleet tracking platform on Google Cloud to record continuous telemetry and diagnostic metrics from 50,00050,000 active delivery vehicles. The ingest layer requires consistent sub-10 millisecond write latencies, must dynamically scale capacity up to hundreds of terabytes without database schema maintenance or manual sharding, and relies solely on single-row key lookups by vehicle ID and timestamp. Which database architecture should you recommend to meet these requirements with minimal operational overhead?

  1. Provision Cloud Bigtable instances using SSD storage for high-throughput time-series key-value telemetry data.Answer
  2. B
    Deploy a Cloud SQL for PostgreSQL instance configured with regional High Availability and standard persistent disks (pd-standard).
  3. C
    Deploy Compute Engine instances running on Spot VMs attached to Local SSD scratch disks hosting a self-managed NoSQL cluster.
  4. D
    Stream incoming vehicle telemetry files directly into Cloud Storage buckets configured with the Archive storage class.

Answer

Provision Cloud Bigtable instances using SSD storage for high-throughput time-series key-value telemetry data.
Cloud Bigtable is a fully managed NoSQL wide-column database optimized for high-volume time-series metrics and telemetry data. It offers sub-10 millisecond latency and scales seamlessly to petabytes of key-value data with low operational overhead.

Step-by-Step Solution

1
Analyze access pattern and functional requirements
Workload involves high-volume time-series writes, key-value single-row queries (vehicle ID + timestamp), sub-10 ms latency, scale to 100s of TBs, and no relational joins.
Determines whether relational (Cloud SQL/Spanner) or NoSQL (Cloud Bigtable/Firestore) service fits best.
2
Evaluate GCP managed database options
Cloud Bigtable is engineered for large-scale time-series and key-value workloads requiring ultra-low latency and seamless horizontal scaling.
Cloud Bigtable provides seamless scaling and high throughput for non-relational telemetry data with minimal operational management.
3
Assess storage media and operational constraints
Bigtable with SSD storage provides predictable sub-10 ms performance compared to self-managed VM options on pre-emptible infrastructure or inappropriate storage tiers.
SSD storage clusters in Bigtable ensure optimal read/write IOPS and low latency under continuous stream loading.

Key Concept

Cloud Bigtable is the optimal GCP database choice for high-throughput, low-latency NoSQL time-series data at multi-terabyte scale.
Rate this question

Topics

Planning Database Options and Persistent Disk Types

More questions from this topic