Serverless Platform Operator Kit (SPOK)

AWS SPOK, known as an opinionated solution bundle, offers a robust technical stack that emphasizes best practices. What this means for you is that launching production-level workloads on ECS can now be done much quicker. But it's not just about ECS and Fargate. SPOK also incorporates essential tools such as Terraform, GitHub, Amazon CloudWatch, and AWS Secrets Manager for establishing end-to-end solution. These services are the go-to choices for many customers already using ECS/Fargate, ensuring you're working with the tools that have proven their worth. Importantly, this bundle is designed to simplify your workflow, letting you focus more on what you do best.

SPOK V1

Getting started

To make it easy for you to get started with SPOK(Serverless Platform Operator Kit), here's a list of next steps. We are going to provision ECS Cluster with AWS Fargate, serverless compute engine for containers.

Deploy Your Pattern

Go to your pattern and run the following command:

terraform init
terraform validate
terraform plan #Here check the plan that Terraform outputs in case you want to change something.
terraform apply --auto-approve

---

Choose your integration available on terraform.tfvars

Please refer following readme file for integration with third party patterns.

• Datadog
• AquaSec
• Github Pipeline

Outputs

After the execution of the Terraform code you will get an output with needed IDs and values needed as input for the following terraform apply. You can use this infrastructure to run other examples, all you need is the cluster_name.

Check Deployment Status

On the ECS Console you will find the ecs-core cluster, in it there will be a service called ecsdemo-backend, this service has an Application Load Balancer associated, you can execute the command terraform output on the ecs-service path and it will show the URL of the ALB, once you open that URL in a web browser it will show the following sample page:

If you keep refreshing the page it will show a different IP address and Availability Zone to demonstrate the Load Balancing.

Cleanup

Run the following command if you want to delete all the resources created before. If you have created other blueprints and they use these infrastructure then destroy those blueprint resources first.

Go to your pattern and run the following command:

terraform destroy

Is possible that the deletion of the ECR Repository fails because there are images left on the repository, hence please delete manually the images and execute the command again.

Best Practices

Terraform uses persisted state data to keep track of the resources it manages. You can either integrate with HCP Terraform to store state data or define a backend block to store state in a remote object. This lets multiple people access the state data and work together on that collection of infrastructure resources.

• Dynamodb Table (A single DynamoDB table is used to lock multiple remote state files)
• S3 Bucket (Stores the state as a given key in a given bucket)

In case you don't have it, use this CLI to create the DynamoDB table it:

  aws dynamodb create-table --table-name TerraformLockStates \
  --attribute-definitions AttributeName=LockID,AttributeType=S \
  --key-schema AttributeName=LockID,KeyType=HASH \
  --billing-mode PAY_PER_REQUEST --region us-east-1

That CLI will create a DynamoDB table on us-east-1 region with name TerraformLockStates and pay per request billing mode, the table will have a Hash Key called LockID

Use this CLI to create the S3 bucket:

  account_id=$(aws sts get-caller-identity --query "Account" --output text) && \
  aws s3api create-bucket --bucket ${account_id}-terraform-states \
  --region us-east-1 --output text --no-cli-pager && aws s3api put-bucket-versioning \
  --bucket ${account_id}-terraform-states --versioning-configuration Status=Enabled

That CLI will crete a private S3 Bucket with name AccountID-terraform-states with versioning enabled by default.

Configure Backend

Go to your pattern and create the following file:

• backend.tf

Replace the bucket name and dynamodb_table, if your region is different than us-east-1 also change region

terraform {
  backend "s3" {
    bucket         = "${account_id}-terraform-states" <--- here
    ...
    region         = "us-east-1" <--- here if your region is different than us-east-1
    dynamodb_table = "TerraformLockStates" <--- here
  }
}

After configuration, please run the following command:

terraform init

Troubleshooting

Error: Error acquiring the state lock

Error message: 2 errors occurred:

• ResourceNotFoundException: Requested resource not found
• ResourceNotFoundException: Requested resource not found

Solution

Create DynamoDB Table on backend.tf to store state lock (terraform.backend.dynamodb_table).