Examkin

Question

Difficulty: MediumCost-Optimized Database Design and Capacity Planning

A company is developing an event-driven system that writes audit logs from several AWS Lambda functions into a database. The log volume is extremely unpredictable; the database experiences hours of inactivity, but can suddenly receive spikes of up to 6,0006,000 write operations per second when upstream batch jobs trigger. The database must handle these sudden bursts immediately without throttling, and the solution must minimize operational costs. Which database capacity configuration is the most cost-effective for this workload?

  1. Configure an Amazon DynamoDB table using on-demand capacity mode.Answer
  2. B
    Configure an Amazon DynamoDB table using provisioned capacity mode with Auto Scaling enabled.
  3. C
    Configure an Amazon DynamoDB table using provisioned capacity mode statically provisioned for the maximum peak throughput of 6,0006,000 write capacity units (WCUs).
  4. D
    Migrate the application database to Amazon RDS with Multi-AZ enabled and use read replicas to handle the write bursts.

Answer

Configure an Amazon DynamoDB table using on-demand capacity mode.
Configuring the Amazon DynamoDB table with on-demand capacity mode is the most cost-effective choice. It automatically manages and scales the throughput capacity to meet the sudden spikes of 6,0006,000 write operations per second without any capacity planning or latency lag, preventing throttling. Since it charges only for the actual read/write requests consumed, no costs are incurred during the hours of inactivity.

Step-by-Step Solution

1
Analyze the workload characteristics
The workload has hours of complete inactivity (idle) and sudden, unpredictable spikes of up to 6,0006,000 write operations per second.
Understanding the traffic pattern is critical for choosing between provisioned and on-demand capacity modes.
2
Evaluate the scaling latency of capacity modes
Provisioned capacity with Auto Scaling is rejected due to the lag in scaling up, which would cause write throttling during sudden spikes.
Auto Scaling requires multiple minutes and metrics evaluations to adjust capacity, failing to meet the zero-throttling requirement for instant bursts.
3
Calculate cost-efficiency across options
Statically provisioning for 6,0006,000 WCUs is rejected due to high idle costs. On-demand mode costs 00 during idle periods and scales instantly to support the 6,0006,000 requests per second.
On-demand mode eliminates the cost of idle capacity while providing the necessary burst performance.

Key Concept

Selecting the cost-optimal capacity mode in Amazon DynamoDB based on workload predictability and burst characteristics.
Rate this question

Topics

Cost-Optimized Database Design and Capacity Planning