Terraform Generator

BESSER offers a code generator designed to facilitate the multi-cloud deployment of your applications. This generator takes as input a deployment architecture model specified with B-UML. Within this model, you can define the specific characteristics of your cloud infrastructure environment and the containerized applications to be deployed.

To create this deployment model using B-UML, we recomend to use the grammar for deployment models provided by BESSER.

Basic Usage Example

from besser.generators.terraform import TerraformGenerator

terraform_generator = TerraformGenerator(deployment_model=deployment_model)
terraform_generator.generate()

Although B-UML enables the modeling of public clusters for different cloud providers and on-premises, currently the code generator supports the creation of Terraform files for AWS and Google Cloud Platform (GCP) only.

In addition to giving the code generator your B-UML deployment model, you’ll also need to provide a configuration file specific to your chosen cloud provider. This file includes provider details (sometimes sensitive data), like the login credentials necessary for Terraform deployment.

Configuration file for AWS

Your configuration file should have the following format:

access_key = ""  # Enter AWS IAM access key
secret_key = ""  # Enter AWS IAM secret key

Configuration file for GCP

Your configuration file should have the following format:

project = "neon-nexus-422908"  # Your GCP project

Note

The location of your configuration file must be specified in the input deployment model as a parameter of each public cluster defined.

Generated Folder Structure

The generator will create a folder with the following structure:

<provider_name>_<cluster_name>/

Inside this folder, you will find the necessary Terraform files for deploying your application on the specified cloud provider.

For example, if deploying to AWS with a cluster named “my-cluster”, the folder structure would look like:

aws_my-cluster/
├── eks.tf
├── iam-config.tf
├── igw.tf
├── nat.tf
├── nodes.tf
├── provider.tf
├── route.tf
├── setup.bat
├── subnet.tf
└── vpc.tf

For GCP with a cluster named “my-cluster”, the folder structure would be:

gcp_my-cluster/
├── api.tf
├── app.tf
├── cluster.tf
├── setup.bat
└── version.tf

Multiple Clusters and Network Configuration

The B-UML deployment model supports multiple clusters for multiple cloud providers. For public clusters, there is an attribute called net_config. If net_config is set to True (which is the default), the generator will create the minimum network and subnetwork configurations required for each cloud provider. This way, you won’t have to worry about modeling the networks and subnets as they will be configured automatically.

Launching Your Terraform Code

To launch your Terraform code, the generator also creates a setup file that initializes and applies the Terraform configuration. Here is an example of the generated setup.bat:

To run the setup script, use the following command in your terminal:

setup.bat

This script will initialize and apply your Terraform configuration, deploying your resources to the specified cloud provider.