Examkin

Soru

Zorluk: Çok zorPlanning Database Options and Persistent Disk Types

An enterprise is migrating an online transaction processing system to Google Cloud. The architecture consists of two primary components with distinct storage requirements:

1. A transactional relational database requiring full ANSI SQL support, multi-region global strong consistency, and automatic horizontal write scaling without manual database sharding.
2. A real-time telemetry processing pipeline running on Compute Engine instances that requires a local block storage volume for temporary scratch files, capable of delivering over 100,000 IOPS at sub-millisecond latency.

Which combination of Google Cloud database and block storage services satisfies all operational and performance requirements with minimal administrative overhead?

  1. Cloud Spanner for the transactional relational database, and Local SSD for the telemetry scratch storage.Cevap
  2. B
    Cloud SQL with cross-region read replicas for the transactional relational database, and SSD Persistent Disk (pd-ssd) for the telemetry scratch storage.
  3. C
    Cloud Bigtable for the transactional relational database, and Standard Persistent Disk (pd-standard) for the telemetry scratch storage.
  4. D
    Self-managed PostgreSQL instances on Spot Compute Engine VMs for the transactional database, and Cloud Storage Coldline for the telemetry scratch storage.

Cevap

Cloud Spanner for the transactional relational database, and Local SSD for the telemetry scratch storage.
The correct answer combines Cloud Spanner and Local SSD. Cloud Spanner is Google Cloud's fully managed relational database designed specifically for workloads requiring enterprise ANSI SQL compliance, global multi-region ACID transactions, and horizontal write scaling without sharding. Local SSDs provide high-performance ephemeral block storage physically attached to the VM host instance, delivering high throughput and sub-millisecond latency for temporary processing scratch space.

Adım Adım Çözüm

1
Analyze the relational database requirement for global strong consistency, ANSI SQL compliance, and horizontal write scaling.
Identify Cloud Spanner as the unique managed GCP relational service offering global ACID transactions and seamless horizontal scaling.
Cloud SQL provides managed relational databases but is limited to vertical scaling for writes per instance, while Cloud Bigtable is a NoSQL store lacking SQL/ACID multi-row transaction capabilities.
2
Analyze the block storage requirements for high IOPS (>100,000 IOPS) and ultra-low sub-millisecond latency for temporary scratch data on Compute Engine.
Select Local SSD as the appropriate storage option.
Local SSDs are physically attached to the host machine hosting the VM, providing microsecond-level latency and extreme IOPS performance suited specifically for non-persistent scratch space.
3
Synthesize the service choices against operational overhead constraints.
Combine Cloud Spanner and Local SSD.
Both selected solutions are fully managed or native performance features that minimize management overhead while satisfying high-scale throughput and relational integrity constraints.

Anahtar Kavram

Choosing Cloud Spanner for globally scalable relational SQL workloads and Local SSD for ultra-low latency VM scratch space.
Tahmini Süre:2m 0s
Bu soruyu puanla

Konular

Planning Database Options and Persistent Disk Types

Bu konudan daha fazla soru