Verifying the Service

Retrieving Terraform Outputs

After terraform apply completes, run terraform output to retrieve the deployment endpoints and resource identifiers:

terraform output

Example output:

aws_region = "ap-southeast-2"
bedrock_max_tokens = 1024
bedrock_model_id = "au.anthropic.claude-sonnet-4-6"
cluster_name = "ecs-express-mode-demo-glm6pm-cluster"
container_image = "ghcr.io/platformfuzz/bedrock-image-analyzer:latest"
health_check_path = "/health"
health_check_url = "https://ec-04256d75ab0d4ba394bc90a351df08b0.ecs.ap-southeast-2.on.aws/health"
public_subnet_ids = [
  "subnet-01068991d9e01116f",
  "subnet-0b496507884f466e7",
]
service_access_type = "PUBLIC"
service_name = "ecs-express-mode-demo-glm6pm-web"
service_url = "https://ec-04256d75ab0d4ba394bc90a351df08b0.ecs.ap-southeast-2.on.aws"
vpc_id = "vpc-0ad6532587641ddce"
web_ui_url = "https://ec-04256d75ab0d4ba394bc90a351df08b0.ecs.ap-southeast-2.on.aws/docs"

Output	Description
aws_region	AWS region where resources are deployed
service_url	HTTPS endpoint Application URL The unique Express-provisioned HTTPS URL on *.ecs.<region>.on.aws used for all normal application traffic. for the web service on *.ecs.<region>.on.aws
service_access_type	Ingress access type — PUBLIC (internet-facing) or PRIVATE (VPC-internal only)
health_check_path	HTTP path the Express Mode load balancer uses for target health checks
health_check_url	Full URL the load balancer polls for target health checks
web_ui_url	Swagger UI for the image analysis API — open in a browser; the root URL returns 404
cluster_name	Name of the ECS cluster hosting the web service
service_name	Name of the ECS Express Mode web service
container_image	Container image URI used by the web service
bedrock_model_id	Bedrock model or inference profile ID configured on the container
bedrock_max_tokens	Maximum Bedrock response tokens configured on the container
vpc_id	VPC ID created for ECS Express networking
public_subnet_ids	Public subnet IDs used by the web service and ALB

Tip

Use service_url for API access and web_ui_url to open the Swagger UI in a browser. The root URL returns 404 by design — the demo serves its UI at /docs.

Verifying Health with curl

Confirm the service is responding by curling the health check URL:

curl -s -o /dev/null -w "%{http_code}" "$(terraform output -raw health_check_url)"

You should receive an HTTP 200 response indicating the container is healthy and serving traffic through the Shared ALB Shared ALB
The Application Load Balancer shared by up to 25 Express services in the same VPC using Host header routing rules. .

Note

If you receive a 503 response, the service may still be starting. Wait 30–60 seconds for the target group to register the tasks as healthy, then retry.

Confirming Tasks Are Running

Verify the ECS service has reached a steady state with desired count matching running count. Use the AWS CLI:

aws ecs describe-services \
  --region "$(terraform output -raw aws_region)" \
  --cluster "$(terraform output -raw cluster_name)" \
  --services "$(terraform output -raw service_name)" \
  --query "services[0].{desired:desiredCount,running:runningCount,pending:pendingCount}" \
  --output table

The output should show desired equal to running and pending at zero. This confirms the Express Gateway Service Express Gateway Service
The aws_ecs_express_gateway_service Terraform resource that provisions an ECS service with managed ALB, auto scaling, and simplified configuration for HTTP/HTTPS web applications and APIs. has launched successfully and all tasks passed the ALB health check.

Tip

You can also verify this in the ECS console by navigating to your cluster, selecting the service, and checking the Tasks tab. The number of running tasks should match the desired count configured in the scaling target.

Endpoint Outputs

The demo exposes three URLs derived from the same Express-provisioned endpoint:

Output	Example path	Purpose
service_url	/	Base HTTPS endpoint for API requests
health_check_url	/health	URL the ALB polls to determine target health
web_ui_url	/docs	Swagger UI — use this to interact with the API in a browser

Health Status “Unknown” Is Expected

When viewing the ECS Tasks tab in the console, you may notice the Health status column shows Unknown for all tasks. This is normal and expected.

Express Mode does not define a container-level healthCheck in the task definition. Instead, operational health is determined entirely by the ALB target group health check configured via health_check_path. As long as the target group reports targets as healthy, the service is functioning correctly — the Application URL Application URL
The unique Express-provisioned HTTPS URL on *.ecs.<region>.on.aws used for all normal application traffic. remains accessible.

Note

The “Health status: Unknown” indicator means ECS has no container-level health check to report against — it does not indicate a problem. Check the EC2 Target Groups console to confirm actual health. If targets show healthy, your deployment is working as intended.