Lorem ipsum dolor sit . Proin gravida nibh vel vealiquete sollicitudin, lorem quis bibendum auctonisilin sequat. Nam nec tellus a odio tincidunt auctor ornare.

Stay Connected & Follow us

What are you looking for?

Simply enter your keyword and we will help you find what you need.

A journey to AWS Lambda integration testing with Python, Localstack, and Terraform

To start with unit testing for AWS Lambda, we can just choose a testing framework and start writing our beautiful unit tests. For testing AWS services we also can use some mock tools like moto. But what about integration testing? A solution may be to deploy with the Continuous Deployment tool and run some test code against real AWS services. But there are some problems:

  • It takes more time to deploy every time
  • Running test code against AWS takes more time
  • Increases the AWS bills

So one good solution could be using a local infrastructure that resembles the same as AWS real infrastructure. So here comes our friend Localstack to solve the problem. And today we will take a look at how we can use Localstack for integration testing of AWS Lambda.



In one of my medium posts, the way of deploying python codes at Lambda with python packages at the lambda layer was described. So here we will not describe that part here. Also as Lambda Layer is a part of the pro version of Localstack, we will avoid the lambda layer. The only change will be rather deploying to real infrastructure, we will deploy to Localstack. So let’s start with sharing the end result GitHub repository first πŸ˜„ 🀫

The skeleton

Let’s start with installing pipenv from here. Then we will install terraform from here.

# Create project directory 
mkdir lambda-testing-localstack
cd lambda-testing-localstack
# Create directory for lambda codes
mkdir lambda# Add file in lambda directory
touch lambda/ Create directory for terraform
mkdir terraform# Add Pipfile in project root directory
touch Pipfile# Create directory for test code
mkdir tests# Create test code python file
touch tests/ So the skeleton will look like this
β”œβ”€β”€ lambda
β”‚ └──
β”œβ”€β”€ Pipfile
β”œβ”€β”€ terraform
└── tests

Add requirements

Let’s add some python requirements and python version 3.8 in Pipfile

name = "pypi"
url = ""
verify_ssl = true

localstack = "=="
pipenv = "==2020.8.13"
# Below packages are required to start localstack from python code
h11 = "==0.10.0"
amazon_kclpy = "==2.0.1"
quart = "==0.13.0"
flask_cors = "==3.0.8"
moto = "==1.3.14"
moto-ext = ">="
crontab = "==0.22.9"
pyOpenSSL = "==19.1.0"

python_version = "3.8"

As Localstack will not be in with lambda codes, so it is added in dev-packages section. Run

pipenv install --dev

to initiate a virtual environment and install packages with pipenv

Add Lambda Code

We will add a very simple code. Our lambda will just get the webpage and return the page source as text.

# lambda/lambda_handler.pyimport logging
import urllib.request

LOGGER = logging.getLogger()

def lambda_handler(event, context):
with urllib.request.urlopen('') as response:
html ='utf-8')
return html

Add Files to deploy Lambda

Let’s add a terraform variable for the lambda function name. So from our test code, we will provide the lambda function name and then test it to make the testing more dynamic.

# terraform/vars.tfvariable "lambda_function_name" {
type = string
default = "test-lambda-function"

Now add terraform code for lambda at terraform/

# terraform/ Zip lambda function codes
data "archive_file" "lambda_zip_file" {
output_path = "${path.module}/lambda_zip/"
source_dir = "${path.module}/../lambda"
excludes = ["__pycache__", "*.pyc"]
type = "zip"
// IAM Policy document for lambda assume role
data "aws_iam_policy_document" "lambda_assume_role" {
version = "2012-10-17"

statement {
sid = "LambdaAssumeRole"
effect = "Allow"
actions = [
principals {
identifiers = [
type = "Service"
// Lambda IAM role
resource "aws_iam_role" "lambda_role" {
name = "test-lambda-role"
assume_role_policy = data.aws_iam_policy_document.lambda_assume_role.json

lifecycle {
create_before_destroy = true
// Lambda function terraform code
resource "aws_lambda_function" "lambda_function" {
function_name = var.lambda_function_name
filename = data.archive_file.lambda_zip_file.output_path
source_code_hash = data.archive_file.lambda_zip_file.output_base64sha256
handler = "handler.lambda_handler"
role = aws_iam_role.lambda_role.arn
runtime = "python3.8"

lifecycle {
create_before_destroy = true

Now we are going to add terraform/localstack.tfto which will tell terraform to use Localstack for deployment

# terraform/localstack.tfprovider "aws" {
region = "eu-west-1"
access_key = "fakekey"
secret_key = "fakekey"
skip_credentials_validation = true
skip_metadata_api_check = true
skip_requesting_account_id = true
s3_force_path_style = true

endpoints {
lambda = "http://localhost:4566"
iam = "http://localhost:4566"

Deploy to Localstack with Terraform

Now time to test deployment at LocalStack

# Start localstack service only for Lambda and IAM
SERVICES=lambda,iam localstack start
# In another terminal
# Terraform initialization
terraform init# Plan to build
terraform plan# Deploy all
terraform apply -auto-approve# Destroy all because we will make it automated as we
# are going to implement automated test
terraform destroy -auto-approve# Kill the localstack service using Ctrl + C in 1st terminal

Automate Test Code

Now time to automate the lambda testing. Our strategy:

  • Deploy terraform to LocalStack from test code e.g python code
  • Execute lambda using boto3
  • Check if we get the webpage as text
  • Destroy infrastructure with terraform to LocalStack using test code

Now add terraform helper function at tests/ which will responsible for creating infrastructure at test start and destroy after the test

# tests/terraform_helper.pyimport subprocess

import os

TERRAFORM_DIR_PATH = os.path.dirname(os.path.realpath(__file__)) + "/../terraform/"

def terraform_init():
"""Terraform init command"""
tf_init = ["terraform", "init", TERRAFORM_DIR_PATH]

def create_resources():
"""Create a tf resource."""
proc = subprocess.Popen("terraform apply -auto-approve " + TERRAFORM_DIR_PATH, shell=True)

def destroy_resources():
"""Destroy all tf resources.

This method will destroy any resources it can find in the state file,
and delete all resources from the state file.
tf_destroy = [

tf_refresh = [

def terraform_start():
""" teardown and create resources at the beginning of feature test """
return create_resources()

Let’s add test code and we are going to use python’s unit test module to test our code

# tests/test_lambda.pyimport os
import unittest

import boto3
from import infra

from tests import terraform_helper

class AWSLambdaTest(unittest.TestCase):
localstack_endpoint = 'http://localhost:4566'
lambda_function_name = 'dynamic-test-lambda-function'

def set_tf_var(self):
os.environ["TF_VAR_lambda_function_name"] = self.lambda_function_name

def setUp(self):
# Start localstack
infra.start_infra(apis=['lambda', 'iam', 'cloudwatch'], asynchronous=True)

def test_lambda_response(self):
client = boto3.client(
response = client.invoke(
assert response['StatusCode'] == 200
assert response['Payload']
html = response['Payload'].read().decode('utf-8')
# Check if "Example Domain" text exists in
assert 'Example Domain' in html

def tearDown(self):
# Stop localstack

if __name__ == '__main__':

Finally, the project will look like this

β”œβ”€β”€ lambda
β”‚ └──
β”œβ”€β”€ Pipfile
β”œβ”€β”€ Pipfile.lock
β”œβ”€β”€ terraform
β”‚ β”œβ”€β”€
β”‚ β”œβ”€β”€
β”‚ └──
└── tests

Time to test the Test

Now come to the moment of playing 🎻

As we implemented the tests with the unit test module, we can run the test using

pipenv shell
python -m unittest

After a lot of logs we can see final output like this

Ran 1 test in 36.943sOK


Localstack is growing very faster. So if it works today may break tomorrow if the required packages are not fixed. So please try to fix all the packages version in Pipfile. Sometimes it may require some python packages to start various services like DynamoDB or StepFuntions. Add those at Pipfile accordingly.

Final codes

Again mentioning the GitHub repository:

author avatar
Mahabubur Rahaman Melon
No Comments

Sorry, the comment form is closed at this time.