Component Access, Events, and Metrics

Infrastructure components do more than create AWS resources. Their Access, Handlers, Metrics, and Envs controls define how services use those resources and how StationOps observes them.

Access Grants

Use the component's Access tab to select the service and grant only the operations it needs. Examples include reading or writing objects, sending or receiving messages, and reading or updating records.

• Grant access to the workload that performs the operation, not every service in the project.
• Keep read and write permissions separate where the editor offers them.
• Recheck grants when a service is renamed, replaced, or excluded from an environment.

Access grants create IAM permissions. They do not automatically tell the application the resource name or endpoint.

Pass Connection Values to Services

Use the component's environment-variable controls to choose a variable name and the services that should receive it. StationOps can inject derived values such as a database connection string, queue URL, bucket name, table name, cache endpoint, Kafka bootstrap servers, or Bedrock agent identifiers.

See Environment Variables for sensitive values and environment overrides.

Event Handlers

Components that expose a Handlers tab can connect events to compatible service targets. The available events depend on the source component. A handler is a runtime relationship; confirm the destination target supports the event source and that failures can be retried safely.

For streams, topic names can be entered as custom events. Keep event names stable and treat consumer code as idempotent because managed event sources may retry delivery.

Metrics and Monitors

Select the useful built-in metrics for a component instead of collecting every available signal. For services, custom metrics and metric monitors can add project-specific thresholds and severity levels. Operations dashboard metrics define a metric, comparison, threshold, and evaluation period for alarms shown in operational views.

Start with symptoms users experience—errors, latency, availability, queue age, connection pressure, or capacity—then add resource saturation metrics that help explain the symptom.

Review Before Deployment

1. Confirm access grants and passed variables point to the intended services.
2. Check handler targets in every environment.
3. Remove stale grants and monitors.
4. Review the generated configuration diff.
5. After deployment, verify the application can connect and the expected metrics appear.