recovery-controller-setup

Create one routing control per Region using route53-recovery-control-config create-routing-control against an existing cluster and control panel, then add an assertion safety rule requiring at least one active control. Wire the routing controls to Route53 health checks and failover DNS records so state changes actually affect traffic.

Enabling zonal autoshift is a two-step flow: first create a practice-run configuration with arc-zonal-shift create-practice-run-configuration including outcome alarms and blocked windows, then run update-zonal-autoshift-configuration with status ENABLED. Verify the application can handle losing one AZ of capacity first.

No. Readiness checks, recovery groups, cells, and resource sets are explicitly out of scope. The skill only covers routing controls with safety rules and zonal shift or zonal autoshift, and redirects readiness questions to ARC Region switch.

A safety rule prevents turning off the last active routing control. Turn on another routing control first, then turn off the one you want. List existing rules with route53-recovery-control-config list-safety-rules to see which rule is blocking the update.

Routing controls only affect traffic when wired to Route53 health checks and failover DNS records. Verify each routing control has an associated health check and that the hosted zone has failover records pointing to those health checks.

Yes. All operations work with the AWS CLI directly; the AWS MCP server's call_aws tool is recommended but optional. Short-lived IAM role credentials such as an instance profile or assumed role are recommended over long-lived access keys.