Terraform module for Iris DR posture across four operational areas:
MultiRegionDeployment: warm-standby DR VPC + optional DR EKS.DataReplication: S3, ECR, and optional cross-region RDS replica.FailoverProcedures: Route53 failover routing, health checks, alerting, and runbooks.DRTesting: scheduled drill playbook and objective alarms (RPO/RTO).
What It Creates#
- Secondary-region network and security baseline.
- Optional secondary EKS control/data plane.
- Cross-region data replication primitives.
- DNS failover records (
PRIMARY/SECONDARY) tied to health checks. - SSM runbooks for failover/failback and periodic DR drills.
- CloudWatch alarms and dashboard for DR objective tracking.
Example Usage#
hcl
module "iris_dr" {
source = "./infra/terraform/iris/dr"
project_name = "iris"
environment = "prod"
primary_region = "us-east-1"
dr_region = "us-west-2"
source_s3_buckets = ["iris-prod-conversations-123456789012", "iris-prod-backups-123456789012"]
primary_rds_arn = "arn:aws:rds:us-east-1:123456789012:db:iris-prod-postgres"
service_domain_name = "api.iris.example.com"
hosted_zone_id = "Z123456ABCDEFG"
primary_service_fqdn = "primary-alb-123.us-east-1.elb.amazonaws.com"
dr_service_fqdn = "dr-alb-456.us-west-2.elb.amazonaws.com"
failover_notification_emails = ["oncall@example.com"]
}
Operational Notes#
- Keep DR node group small (
desired=1,min=0) for cost-efficient warm standby. - DR drill automation expects at least one SSM-managed runner instance tagged
by:
- key:
Role - value:
dr-runner
- key:
- Publish
Iris/DRcustom metrics from drill workflows:RecoveredRPOHoursRecoveredRTOHoursDRTestFailureCountDRTestSuccessCount
Validation#
bash
terraform -chdir=infra/terraform/iris/dr init
terraform -chdir=infra/terraform/iris/dr validate
terraform -chdir=infra/terraform/iris/dr plan