Examkin

Question

Difficulty: HardPlanning Database Options and Persistent Disk Types

An enterprise organization is migrating a core reporting system to Google Cloud Compute Engine. The system architecture has two distinct storage requirements:

1. A backend transactional database that requires full SQL ACID compliance, automated daily backups, high availability (HA) with regional failover across two zones, and a dataset size scaling up to 3 TB.
2. A temporary high-throughput scratch space for worker virtual machines (VMs) performing batch analytics sorting, which requires maximum IOPS and microsecond read/write latency, while tolerating complete data loss when the VM stops or restarts.

Which storage and database configuration should a cloud engineer select to meet these requirements with minimal operational overhead?

  1. Deploy Cloud SQL for PostgreSQL with a High Availability (regional) configuration for the transactional database, and attach Local SSDs to the worker Compute Engine VMs for scratch space.Answer
  2. B
    Deploy Cloud Bigtable for the transactional database, and attach Standard Persistent Disks (pd-standard) to the worker Compute Engine VMs for scratch space.
  3. C
    Deploy a PostgreSQL database on Spot Compute Engine VMs using extreme persistent disks (pd-extreme), and store worker scratch space files in Coldline Cloud Storage buckets.
  4. D
    Deploy Cloud SQL for PostgreSQL with a Single Zone configuration, and configure worker VMs to write scratch space directly to Archive Cloud Storage buckets using object lifecycle rules.

Answer

Deploy Cloud SQL for PostgreSQL with a High Availability (regional) configuration for the transactional database, and attach Local SSDs to the worker Compute Engine VMs for scratch space.
Cloud SQL for PostgreSQL configured for High Availability delivers a fully managed relational database with cross-zone failover, automated backups, and full ACID compliance for datasets under 64 TB. Attaching Local SSDs to worker Compute Engine instances provides ultra-low microsecond latency and high IOPS performance directly attached to the host physical server, ideal for ephemeral scratch space that can tolerate data loss upon VM termination.

Step-by-Step Solution

1
Analyze the backend database requirements
Requirements call for a relational database with SQL ACID support, regional high availability, automated backups, and 3 TB capacity.
Cloud SQL supports PostgreSQL relational engines, up to 64 TB storage, automated maintenance/backups, and regional failover pairs across zones with minimal operational burden.
2
Analyze the worker VM scratch space requirements
Requirements call for maximum IOPS, microsecond latency block storage, with tolerance for disk loss on VM stop/restart.
Local SSD disks are physically attached to the server hosting the VM instance, delivering sub-millisecond latencies and high IOPS suitable for scratch disk usage.
3
Combine the optimal storage solutions
Pair Cloud SQL PostgreSQL (HA) with Local SSD block storage attached to the worker compute nodes.
This combination satisfies both functional criteria while optimizing for performance, operational simplicity, and cloud design best practices.

Key Concept

Selecting GCP Managed Databases and Block Storage Types based on Workload Characteristics
Estimated Time:2m 0s
Rate this question

Topics

Planning Database Options and Persistent Disk Types

More questions from this topic