• Terraform >=v0.12.24
• provider.aws >=v2.57.0

# Clone the repo
git clone https://github.com/netoht/lambda-terraform.git
cd lambda-terraform/

Implement your lambda function in lambda.py.

# Packaging your lambda function
./lambda-package.sh

This action will create a zip inside ./tf/target/lambda.zip

The Terraform needs to save states about resources, then we need tell to him about that. The backend.tf file ask to save state in s3 and provider.tf inform about cloud provider, in this case AWS.

The backend-config/dev configure the variables in which the environment it will usage to save states.

# AWS Region of the bucket
region="sa-east-1"
# The bucket name to save state
bucket="myproject-dev"
# The project key of the terraform state
key="tfstate/myproject-dev"

For each environment you need create a variables-{ENV}.tfvar like this variables-dev.tfvar:

env="dev"
aws_region="sa-east-1"
lambda_name="minimal_lambda_function"
lambda_handler="lambda.handler"
lambda_runtime="python3.6"
lambda_filename="target/lambda.zip"
tags={"Project": "test"}

Here was informed the environment and information about lambda function that should be used to create the resources and it will used inside lambda.tf. In the lambda.tf has the resources that will be created, the aws_lambda_function and the aws_iam_role.

Terraform needs to initialize (terraform init) the state and apply (terraform apply) the changes. The run-terraform.sh script was created to help us.

$ cd tf/
# ./run-terraform.sh -e <Environment>
./run-terraform.sh -e dev

...
Do you want to perform these actions?
  Terraform will perform the actions described above.
  Only 'yes' will be accepted to approve.

  Enter a value: yes

Execute the command below to test the lambda function:

aws lambda invoke --invocation-type RequestResponse --function-name minimal_lambda_function_dev --region sa-east-1 --log-type Tail --payload '{"key1":"value1", "key2":"value2", "key3":"value3"}' target/outputfile.txt | jq .LogResult | sed 's/"//g' | base64 --decode

You should see something like this:

START RequestId: e02625f4-c922-4be1-be9f-15d04897a3ac Version: $LATEST
Received my event: {
  "key1": "value1",
  "key2": "value2",
  "key3": "value3"
}
END RequestId: e02625f4-c922-4be1-be9f-15d04897a3ac
REPORT RequestId: e02625f4-c922-4be1-be9f-15d04897a3ac  Duration: 0.41 ms       Billed Duration: 100 ms Memory Size: 128 MB     Max Memory Used: 42 MB

• Resource: aws_lambda_function
• How do I allow my Lambda execution role to access my Amazon S3 bucket?