Serverless Framework

Introduction

In this series, we’ll focus on Serverless Framework, one of the fastest ways to build and deploy serverless applications.

If you’ve already worked with Terraform, you know how great it is for infrastructure automation. So why another tool?
Because Serverless Framework is built specifically for deploying serverless apps — it abstracts away the heavy lifting of infrastructure management and lets you deploy complex setups in just a few lines of configuration.

Just like Terraform, Serverless is cloud-agnostic — it supports AWS, GCP, and Azure out of the box.

👉 Official website

---

Setting Up

Refer to the official setup guide for installation instructions.

If you’re on Windows, consider using WSL (Windows Subsystem for Linux) — it makes development smoother and closer to a native Linux experience, especially when working with AWS or Python.

Once Serverless is installed, it can use your existing AWS CLI credentials automatically.
If you already have AWS CLI configured with a default profile, you don’t need extra setup.

However, if you haven’t configured AWS yet, here’s how:

1. Go to IAM → Users in AWS Console

2. Create a new user named serverless

3. Attach the AdministratorFullAccess policy (for simplicity — use least privilege in production)

4. Create an Access Key and Secret Key

5. Configure credentials locally:

serverless --version
serverless config credentials --provider aws --key <yourKey> --secret <yourSecret>

---

Test Setup

To confirm everything works, let’s deploy a simple Lambda function.

Create a Test Project

serverless create --template aws-python --path new_project --name testFunc

This will generate a project structure like:

new_project/
├── handler.py
└── serverless.yml

Minimal serverless.yml

Remove comments to simplify it:

service: testFunc
frameworkVersion: '3'

provider:
  name: aws
  runtime: python3.8

functions:
  hello:
    handler: handler.hello

Deploy

Now deploy the function:

sls deploy

Once deployed, you’ll see a confirmation similar to this:

Deploying testFunc to AWS...
Service Information
stage: dev
region: us-east-1
stack: testFunc-dev

To remove everything later:

sls remove

🎉 Congratulations! You just deployed your first serverless function.

---

Invoking the Function

Testing locally or through the CLI is straightforward:

# Invoke and view output
sls invoke -f hello --log

# Stream recent logs
sls logs -f hello --tail

Sample Output

$ sls invoke -f hello --log
{
  "statusCode": 200,
  "body": "{\"message\": \"Go Serverless v1.0! Your function executed successfully!\", \"input\": {}}"
}
--------------------------------------------------------------------
START
END Duration: 1.33 ms (init: 130.95 ms) Memory Used: 38 MB

---

Using External Libraries

Most real-world Lambdas depend on external libraries (like pandas or requests).
Serverless makes it easy to package and deploy them in two ways:

Option 1 — Include Dependencies as a Zip

Bundle dependencies directly into your deployment zip.

Option 2 — Use Lambda Layers

Upload dependencies as a Lambda Layer to share across multiple functions.
Learn more: AWS Blog – Lambda Layers

---

Using the serverless-python-requirements Plugin

Install the plugin:

sls plugin install -n serverless-python-requirements

Then update serverless.yml:

Using Zip Packaging

plugins:
  - serverless-python-requirements

custom:
  pythonRequirements:
    dockerizePip: true
    zip: true

For zipped dependencies, include this at the top of your Python code:

import unzip_requirements

Using Lambda Layers

plugins:
  - serverless-python-requirements

custom:
  pythonRequirements:
    dockerizePip: true
    layer:
      name: PandasLayer

Define dependencies in requirements.txt, for example:

pandas==1.0.1

Serverless will automatically build and deploy them as a Lambda Layer.

---

Summary

The Serverless Framework offers a fast, cloud-agnostic way to create and manage serverless applications.
It simplifies deployment, testing, and packaging, letting developers focus on code — not infrastructure.

👉 Check out the GitHub repository for examples:
github.com/k-pulkit/source-demo-serverless-framework