Examkin

Question

Difficulty: HardPlanning Database Options and Persistent Disk Types

An enterprise organization is migrating an online transaction processing (OLTP) core application to Google Cloud. The system requires a relational database engine with strict multi-row ACID compliance, sub-10 millisecond write latencies, and persistent data retention across virtual machine restart events. The database requires 30,000 sustained random read/write IOPS for a 1 TB dataset. The architecture team mandates achieving these requirements with minimal operational management overhead. Which database and block storage architecture should you recommend?

  1. A
    Deploy a Cloud Bigtable cluster provisioned with SSD storage nodes to host the transactional database.
  2. B
    Deploy a self-managed relational database engine on a Compute Engine instance using attached Local SSD volumes for primary data files.
  3. C
    Deploy a self-managed relational database on a Compute Engine instance using a Standard Persistent Disk (pd-standard) volume configured with automated lifecycle tiering to Cloud Storage Archive.
  4. Deploy a Cloud SQL instance with High Availability enabled, utilizing a Persistent Disk SSD (pd-ssd) storage volume.Answer

Answer

Deploying a Cloud SQL instance with High Availability and Persistent Disk SSD (pd-ssd) storage provides a fully managed relational engine, persistent storage across host restarts, low latencies, and necessary IOPS performance.
The solution recommending Cloud SQL with High Availability and Persistent Disk SSD (pd-ssd) meets all criteria. Cloud SQL offers a fully managed relational database engine (MySQL, PostgreSQL, or SQL Server) with ACID compliance and automatic maintenance/failover, reducing management overhead. Persistent Disk SSD provides durable non-volatile storage that retains data through VM restarts and delivers high performance (30 IOPS per GB30\text{ IOPS per GB} for reads and writes, achieving the 30,000 IOPS30,000\text{ IOPS} requirement at 1 TB1\text{ TB}).

Step-by-Step Solution

1
Evaluate workload database requirements
Relational schema and multi-row ACID compliance eliminate NoSQL options like Cloud Bigtable.
Cloud Bigtable lacks multi-row ACID transaction guarantees required for relational OLTP engines.
2
Evaluate storage persistence and operational overhead constraints
Local SSDs are ephemeral and lose data upon instance shutdown. Managed Cloud SQL reduces operational overhead compared to self-managed Compute Engine databases.
Primary storage for persistent transactional databases must survive VM restarts and host maintenance.
3
Calculate Persistent Disk IOPS performance
A 1 TB Persistent Disk SSD (pd-ssd) yields 30,000 random IOPS (1,000 GB×30 IOPS/GB=30,000 IOPS1,000\text{ GB} \times 30\text{ IOPS/GB} = 30,000\text{ IOPS}), meeting latency and throughput targets.
pd-standard disks provide significantly lower IOPS per GB and are intended for sequential I/O.

Key Concept

Selecting GCP Managed Databases and Persistent Block Storage Types based on ACID requirements, persistence, IOPS, and operational overhead.
Rate this question

Topics

Planning Database Options and Persistent Disk Types

More questions from this topic