How to fix Error: Error creating ELB: DuplicateLoadBalancerName in Terraform

First, verify that the load balancer actually exists in AWS. Open the AWS console, navigate to EC2 > Load Balancers, and search for the load balancer name mentioned in the error.

aws elb describe-load-balancers --load-balancer-names my-elb-name

If the load balancer exists in AWS, proceed to the next steps. If it doesn't exist, the issue may be different.

Verify whether the load balancer is tracked in your Terraform state:

terraform state list | grep aws_elb
terraform state show aws_elb.my_elb_name

If the resource is NOT in the state but EXISTS in AWS, the load balancer is orphaned and you need to import it.

If the load balancer exists in AWS but not in Terraform state, import it using the load balancer name:

terraform import aws_elb.my_elb_name my-elb-name

Or in Terraform v1.5.0+, use an import block in your configuration:

import {
  to = aws_elb.my_elb_name
  id = "my-elb-name"
}

Then run terraform apply to finalize the import.

If you don't want to keep the existing load balancer, delete it from AWS and re-run Terraform:

aws elb delete-load-balancer --load-balancer-name my-elb-name
terraform apply

Be cautious with this approach if the load balancer is serving production traffic.

If you're modifying the load balancer configuration and getting replacement errors, ensure Terraform destroys the old resource before creating the new one. Add a lifecycle block to your load balancer resource:

resource "aws_elb" "example" {
  name = "my-elb"
  # ... other configuration ...

  lifecycle {
    create_before_destroy = false
  }
}

This ensures the old load balancer is destroyed first before the new one is created, preventing name collisions.

If running multiple environments or CI/CD pipelines, use unique names for each load balancer to prevent conflicts:

resource "aws_elb" "example" {
  name = "my-elb-${var.environment}"
  # Or use a timestamp/random suffix
  # name = "my-elb-${random_string.lb_suffix.result}"
}

This prevents naming collisions when multiple deployments run simultaneously.