rust-lambda-starter

This is a template for starting serverless projects with AWS Lambda, Rust, and OpenTofu.

Components

The project contains:

• a Cargo project with an empty library and a hello_world binary representing an example Lambda function;
• a Python script for building all the Lambda funtions present in the ./src/bins directory;
• a base infrastructure written in OpenTofu with IAM, Lambda, and API Gateway resources;
• a CI pipeline that triggers on any PR against the main branch or any commit to the main branch;
• a CD pipeline that triggers on any push to the main branch.

Usage

Getting the project

To get a local copy of the template, either use pont:

pont build --name <your-project-name> --from https://github.com/soupdevsolutions/rust-lambda-starter

or clone the repository and change the occurences of the template name:

git clone https://github.com/soupdevsolutions/rust-lambda-starter ./<your-project-name>
cd <your-project-name>

# rename the occurences using bash or your IDE tooling
TEMPLATE_NAME_KEBAB_CASE=rust-lambda-starter
TEMPLATE_NAME_SNAKE_CASE=rust_lambda_starter
PROJECT_NAME=<your-project-name>

find . \( -name \*.rs -o -name \*.tf -o -name \*.toml -o -name \*.md -o -name \*.lock \) -exec sed -i '' "s#$TEMPLATE_NAME_KEBAB_CASE#$PROJECT_NAME#g" {} \;
find . \( -name \*.rs -o -name \*.tf -o -name \*.toml -o -name \*.md -o -name \*.lock \) -exec sed -i '' "s#$TEMPLATE_NAME_SNAKE_CASE#$PROJECT_NAME#g" {} \;

Adding functionality

To create new Lambda functions, you can copy the ./src/bins/hello_world.rs code to a new file in the same directory and modify the function_handler body with your desired functionality.
Any new functions should be added as binaries in the Cargo.toml file, as follows:

[[bin]]
name = "<function_name>"
path = "src/bins/<function_name>.rs"

Any common code can be added to the library and split into modules just like in any other Cargo project.

For the infrastructure code, duplicate the code from the ./infrastructure/api.tf, ./infrastructure/lambdas.tf, and ./infrastructure/iam.tf files, modifying the resource names and any fields referring to resource names. Any additional resources can be added to new files (e.g. ./infrastrucutre/s3.tf).

Deployment

Before being able to deploy anything, you will need to create the state bucket manually. Use the bucket name from ./infrastructure/main.tf or change it, if it's taken.

The recommended way for deploying the project is to use GitHub Actions; however, you can also do it from your local machine.

For deploying with GitHub Actions, push your project to a new repository, go to the repository's Settings -> Secrets and variables -> Actions, and add secrets for your IAM user credentials:

• AWS_ACCESS_KEY_ID -> the access key id
• AWS_SECRET_ACCESS_KEY -> the secret access key Any push to the main branch will trigger both the CI and CD pipelines.

If you choose to deploy from your local machine, make sure to follow the guide for configuring the AWS CLI, so OpenTofu can deploy to your AWS account. Then, you can run all the steps from the CD pipeline locally:

# init OpenTofu (only needs to be done once locally)
tofu -chdir=infrastructure init

# build the Lambda functions
python3 ./scripts/build.py

# plan and apply the changes
tofu -chdir=infrastructure plan
tofu -chdir=infrastructure apply