Google Cloud Solutions Builder

A solution framework to generate code with built-in structure and modules to accelerate your project setup.

TL;DR

Solutions Builder is a boilerplate tool for building repeatable solutions with the best practices in architecture on Google Cloud, including Cloud Run, GKE clusters, Test Automation, CI/CD, etc, with built-in modules including:

• Terraform boilerplate modules
• Modular microservice templates, deployable to Cloud Run or a Kubernetes cluster.
• Generating code from templates at remote Git repo (e.g. a private Git repo)
• Container image building and deploying with Skaffold
• Modules and templates based on Copier.
• CI/CD deployment templates.

Build a solution by adding modules

Solutions Builder is designed with modules and templates. It starts from a base project skeleton, then start adding modules into this solution folder.

For example, here are the steps when creating a new solution:

• Start a new solution with a base (skeleton) code from a "root template"
• Add a Cloud Run microservice with APIs
• Add a Load Balancer with Identity-Aware Proxy
• Add CI/CD for deployment

See Quick Start section to create a sample solution.

Concepts and Design Details

• See Overview for overall concepts and design details.

Prerequisite

Required dependencies:

• Python (3.10 or later)
• gcloud CLI (Latest)
• Terraform (v1.4.0 or later)
• Skaffold (v2.4.0 or later)

Optional dependencies for GKE cluster deployment:

• Kustomize (v5.0.1 or later)

Install dependencies with package managers:

• Mac:

  brew install gcloud terraform skaffold
  

• Windows:

  choco install gcloud terraform skaffold
  

• Others

  • Google Cloud Shell - See docs/guides/cloudshell.md

Install Solutions Builder CLI

pip install solutions-builder

Quick Start - Create a new solution

This quick start steps will create the following:

• Create a new solution folder.
• Create a new GCP project and initialize Terraform infrastructure.
• Add a FastAPI-based microservice with a sample API.
• Deploy the service to Cloud Run.

Generate a new solution in my-solution-folder folder.

sb new my-solution-folder

This will prompt options and variables:

🎤 What is your project name? (Spaces are allowed.)
   my awesome solution name
🎤 What is your Google Cloud project ID?
   my-project-id
🎤 Create GCP project 'my-project-id'? (yes/no)
   yes
🎤 Which Google Cloud region?
   us-central1
🎤 Default deploy method? (cloudrun or gke)
   Cloud Run

(...Leave the rest as default)

Once answered, it will generate the code from the template:

Copying from template
    create  .
    create  .pylintrc
    create  .copier-answers.yml
    ...

Answer questions when adding terraform_base module:

Adding module 'terraform_base'...
🎤 Use GCS Bucket for Terraform backend?
   Yes
🎤 What is your GCP billing account?
   12345-12345-12345
🎤 Use GCS Bucket for Terraform backend?
   Yes

Copying from template
 identical  .
    create  terraform
    create  terraform/stages
    ...


Complete. New solution folder created at ./my-solution-folder.

• At this point, you'll see a solution folder created at ./my-solution-folder, and a sb.yaml file with all configuration about this solution folder.
• See overview.md for more details about sb.yaml.

Next, add a Sample Cloud Run microservice using blank_service module template:

cd my-solution-folder
sb add component sample_service -t blank_service

• Alternatively, use a template from a remote Git repo like:

   sb add component sample_service -t git@github.com:GoogleCloudPlatform/solutions-builder.git/modules/blank_service
  

And answer the following:

This will add component 'sample_service' to 'components' folder. Continue? [Y/n]:
🎤 Resource name (lower case, alphanumeric characters, '-')?
   sample-service
🎤 Which Google Cloud region?
   us-central1
🎤 Add a sample API?
   Yes
🎤 Does this component require the Common image?
   No
🎤 Default deploy method? (cloudrun or gke)
   Cloud Run

Apply all Terraform stages (in terraform/stages folder)

sb terraform apply --all --yes

The command above will run the following:

• Init and apply all Terraform stages. (terraform init and terraform apply))
• Build the sample_service container image and published to Artifact Registry. (using Skaffold)
• Deploy the sample_service container to Cloud Run. (using Terraform)

Optionally, manually build and deploy sample_service using sb deploy.

sb deploy -m sample_service

Once deployed, you can test the Cloud Run instance at: https://console.cloud.google.com/run/detail/us-central1/sample-service

CLI Usage

Check out CLI_USAGE.md for more CLI usages and tutorials.

Additional Guides

• cloudshell.md - Step-by-step guidance to run Solutions Builder on Google Cloud Shell.
• cloudrun.md - The guidance if you want to deploy microservice to Cloud Run.
• gke.md - The overall development guidance on Google Kubernetes Engine.

Development and contributions

• CONTRIBUTIONS.md - How to contribute this project, code submission process.
• DEVELOPMENT.md - Guidances on developing new templates, modules, and components.

Troubleshooting

• TROUBLESHOOTING.md - Development guide and code submission process.

FAQ

• Who are the target audience/users for this Solutions Builder?

  • A: Any engineering team to start a new solution development project.

• Can I use this template for non-Google or multi-Cloud environments?

  • A: We design this Solutions Builder to work 100% out of the box with Google Cloud products. However, you could customize the solution to meet your needs on multi-Cloud environment. See Why Google Cloud for details.