Skip to main content

Step 2 — Configure Each Service

Clicking any service node opens the Node Configuration panel. Every panel reflects the actual AWS service model.

What is in each Node Configuration panel​

Display label — the name shown on the canvas node. Keep these meaningful for easy reading of simulation metrics.

Engineering Notes — an explanation of the service's role and performance characteristics at scale. For example, the Lambda panel explains cold start behaviour, the DynamoDB panel covers partition key design and hot partition risk. These notes are particularly useful if you are newer to a service.

Service-specific properties — the full configuration surface for that service. For example:

  • Lambda: memory, timeout, runtime, reserved concurrency, provisioned concurrency
  • DynamoDB: capacity mode (on-demand vs. provisioned), DAX, partition key, GSI configuration
  • API Gateway: throttling limits, authorisation, caching configuration, stage settings
  • CloudFront: cache mode (aggressive, balanced, minimum), edge compute options

Limits, Bottlenecks & Pro Tips — built-in guidance on service quotas and known architectural pitfalls at scale. Read these before simulating at high RPS.

Service Endpoints — regional endpoint configuration.

Service Quotas — applicable soft and hard limits for the chosen configuration.

Configuration best practices​

  • Set your target region consistently across all services before running the first simulation. Mismatched regions will not be flagged as errors but will produce cost estimates that do not reflect real-world multi-region routing.
  • For Lambda, set memory and runtime before simulating. These directly affect cold start duration and cost. See the AWS Service Reference for Lambda-specific guidance.
  • For DynamoDB, decide on capacity mode (on-demand vs. provisioned) and your partition key design before simulating. A poorly chosen partition key will produce hot partition throttling that looks like a DynamoDB capacity problem but is actually a design problem. See the AWS Service Reference for DynamoDB-specific guidance.
  • Configuration panels are pre-populated with AWS-accurate defaults. Check these against your intended configuration — the defaults are suitable starting points but may not reflect your production requirements.