Workload Rightsizing And Capacity Optimization — AWS Developer (DVA-C02)
Lambda Memory Controls CPU. And Therefore Duration and Cost.
Lambda memory and timeout look like independent knobs, but they are not. Memory allocation determines CPU allocation, which determines how fast the function runs, which determines billable duration. A function processing large payloads that times out at 3 seconds may complete in under 1 second at a higher memory setting, cutting both duration charges and retry overhead despite the higher per-GB-second rate. Raising the timeout alone does not improve performance. Memory is the only Lambda knob that increases CPU throughput, and CPU throughput is what reduces duration. Lambda Power Tuning, an open-source Step Functions state machine, automates the memory-versus-cost tradeoff by running the function across a range of memory configurations.
What This Pattern Tests
The exam presents an over-provisioned or under-provisioned workload and tests whether you identify the rightsizing opportunity. AWS Compute Optimizer analyzes CPU, memory, and network metrics to recommend instance type changes. CloudWatch metrics show utilization history. The trap is recommending Reserved Instances for an over-provisioned instance — you save 40% on a $200/month instance ($80 savings) when rightsizing first saves 75% ($150 savings) and then you reserve the smaller instance. Always rightsize before applying pricing optimizations.
Decision Axis
Current utilization vs. provisioned capacity. Rightsize first (reduce waste), then optimize pricing (reduce rate).
Associated Traps
More Top Traps on This Exam
Decision Rules
Determine whether increasing memory allocation reduces net GB-second billing cost for a CPU-bound Lambda function by cutting duration proportionally more than it raises the per-unit memory rate.
Domain Coverage
Difficulty Breakdown