Performance Improvement Strategy — AWS SysOps Administrator (SOA-C03)
Every Performance Fix Targets a Specific Bottleneck Layer
A scenario describes slow response times and offers four options: larger EC2 instance, ElastiCache for session data, CloudFront distribution, and an RDS read replica. Each addresses a different layer. The instinct is to choose the most comprehensive option. The exam rewards identifying which layer is actually constrained. If the scenario mentions repeated database queries for identical reference data, caching is the specific fix — adding compute does not address a data-layer bottleneck.
What This Pattern Tests
The exam describes a performance issue and tests whether you identify the bottleneck before optimizing. High CPU = scale out (more instances) or up (larger instance type). High memory = increase instance memory or add caching (ElastiCache). High disk IOPS = switch to provisioned IOPS SSD (io2) or instance store. Network bottleneck = use enhanced networking, placement groups, or a higher bandwidth instance type. The trap is adding more instances when the bottleneck is a single-threaded process that needs a faster CPU (scale up, not out) or adding cache when the bottleneck is write throughput (caching helps reads, not writes).
Decision Axis
Bottleneck identification before optimization selection. CPU, memory, disk, and network each have different solutions.
Associated Traps
More Top Traps on This Exam
Decision Rules
Whether gp3's independently configurable IOPS and throughput ceilings are sufficient for the stated numeric I/O profile, making io2 an unjustified cost premium rather than the required production-grade choice.
Domain Coverage
Difficulty Breakdown