Examkin

Question

Difficulty: Very hardPlanning Database Options and Persistent Disk Types

A financial technology firm is deploying a real-time trading processing engine on Google Cloud consisting of two core workload components. Component 1 is a trading account management database requiring full SQL compliance, multi-table ACID transactions, automated regional failover, and high write performance. Component 2 is an ephemeral risk simulation pipeline performing transient parallel calculations requiring ultra-low sub-millisecond I/O latency, where scratchpad data can be re-generated if an instance terminates. Which TWO storage and database configurations should the solution architect implement to meet these requirements with minimal operational overhead? (Select TWO.)

  1. Provision a Cloud SQL instance configured with High Availability (regional failover) using SSD Persistent Disk (pd-ssd) for the trading account management database.Answer
  2. Attach Local SSD storage to the Compute Engine virtual machines executing the transient risk simulation pipeline.Answer
  3. C
    Deploy a Cloud Bigtable cluster for the trading account management database to ensure relational SQL schema support and multi-table ACID transaction guarantees.
  4. D
    Mount Coldline Cloud Storage buckets to the simulation nodes to store transient scratchpad data for the risk simulation pipeline.

Answer

The architect should deploy a High Availability Cloud SQL instance backed by SSD Persistent Disks for the relational account database, and attach Local SSDs to Compute Engine instances for the high-speed scratchpad risk simulation workload.
Cloud SQL configured with High Availability delivers an automated failover solution with full relational SQL and ACID transactional guarantees using SSD Persistent Disk storage. For transient high-performance scratchpad data, Local SSD delivers sub-millisecond latency and high IOPS directly attached to Compute Engine instances.

Step-by-Step Solution

1
Analyze requirement for Component 1 (trading account database)
Identified needs: relational SQL, multi-table ACID compliance, regional automated failover, and high write throughput.
Cloud SQL with HA configuration in a regional deployment meets relational and regional failover needs, while pd-ssd provides required IOPS performance.
2
Analyze requirement for Component 2 (risk simulation pipeline)
Identified needs: sub-millisecond latency, extremely high IOPS, transient/ephemeral data retention.
Local SSD provides maximum IOPS and microsecond latency directly on the host instance, perfectly matching ephemeral scratchpad requirements.
3
Evaluate and eliminate non-compliant distractors
Cloud Bigtable lacks relational SQL/ACID features, and Coldline Cloud Storage introduces massive latency and per-operation retrieval costs for frequent transient data access.
Matching workload patterns against Google Cloud storage characteristics prevents architectural anti-patterns.

Key Concept

Selecting GCP Managed Databases and Storage Types based on Relational/ACID needs, latency, availability, and durability constraints.
Rate this question

Topics

Planning Database Options and Persistent Disk Types

More questions from this topic