Mastering AWS Lambda with Boto3 and Python: A Complete Step-by-Step Tutorial

AWS Lambda is one of the easiest ways to build serverless applications on AWS. You don’t need to set up or maintain servers — AWS does it for you. In this tutorial, we’ll start simple and build our way up. This tutorial walks you through everything — writing your first Lambda function, deploying it, automating it with Boto3, and working with advanced features like environment variables and triggers. Let’s get started.

What Exactly is AWS Lambda?

1. “AWS Lambda runs your code only when needed”

Normally, you rent or own a server. That server stays on all the time, even if nobody is using it.
With AWS Lambda, there’s no always-on server.

• Your code sits waiting in AWS.
• When something happens (like a request or an event), AWS starts your code, runs it, and then stops.
• You only pay for the time your code actually runs.

2. “You upload your code and define a trigger”

A trigger is simply an event that tells AWS Lambda: “run this code now.”

• Example triggers:
  • API Call → Someone visits your website’s API.
  • S3 Event → A file is uploaded to an S3 bucket.
  • CloudWatch Schedule → A timer that runs your code every hour.
    You just upload your code to Lambda, select a trigger, and AWS connects them.

3. “AWS handles the rest”

You don’t have to:

• Buy servers.
• Install operating systems.
• Patch or update software.
  AWS automatically does all of that. This is why Lambda is called a serverless service — the servers exist but AWS manages them for you.

4. Key Features (explained simply)

No server management
You don’t need to think about servers at all. AWS gives you an empty space where your code runs.

Pay per execution
Instead of paying for a server that sits idle, you only pay when your code actually runs.
Example:

• If your code runs for 200 milliseconds once a day, you pay for just those milliseconds.

Automatic scaling
If one user runs your code, AWS runs one instance.
If 10,000 users run it at the same time, AWS automatically runs thousands of instances in parallel — no setup needed from you.

In Short

• Upload code → Set trigger → Done.
• AWS runs your code on demand.
• No servers to manage, pay only when used, and automatic scaling to handle any number of requests.

Setting Up the Environment

Before writing any Python code or using Boto3, you need to prepare your system with the right tools.

Prerequisites

• AWS Account – Sign up at https://aws.amazon.com.
• Python 3.x – Installed on your computer.
• AWS CLI – Installed and configured with your AWS credentials.
• Boto3 Library – Installed in Python.

Step 1: Install the AWS CLI

Open your terminal or command prompt and run:

pip install awscli

pip install awscli

Step 2: Configure the AWS CLI

After installation, configure your credentials:

aws configure

aws configure

You’ll be prompted to enter:

• AWS Access Key ID
• AWS Secret Access Key
• Default Region (for example, us-east-1)
• Output format (for example, json)

This securely saves your credentials on your system so Boto3 and AWS CLI can use them.

Step 3: Install Boto3

Run:

pip install boto3

pip install boto3

Boto3 is the official Python SDK for AWS services, including Lambda.

Step 4: Verify Everything

• Check AWS CLI version: aws --version
• Check Python version: python --version
• Test importing Boto3 in Python: import boto3 print("Boto3 is ready!")

At this point, your environment is fully prepared to start creating and managing AWS Lambda functions with Python and Boto3.

Creating Your First AWS Lambda Function

You’re going to create a small program on AWS called a Lambda function. When you press a button, AWS will run it and give you the output.

Think of it as making a tiny program in the cloud that says “Hello from Lambda!”

Step by Step (Very Simple)

1. Sign in to AWS Console

• Go to https://aws.amazon.com/console/.
• Sign in with your AWS account.

2. Open Lambda Service

• In the search box at the top, type Lambda.
• Click Lambda.

3. Click “Create Function”

You’ll see a big button called Create function. Click it.

4. Choose “Author from Scratch”

On the page that appears:

• Select Author from scratch.
• Function name: type hello_lambda.
• Runtime: choose Python 3.x from the dropdown.
• Scroll down and click Create function at the bottom.

This creates an empty Lambda function for you.

5. Add the Code

On the next page, you’ll see a section called Code source or Code editor.
Delete anything that’s already there and paste this code:

def lambda_handler(event, context):
    return {
        'statusCode': 200,
        'body': 'Hello from Lambda!'
    }

def lambda_handler(event, context):
    return {
        'statusCode': 200,
        'body': 'Hello from Lambda!'
    }

This is a tiny Python function. AWS will run this whenever you “invoke” it.

6. Save/Deploy the Code

Click the Deploy button above the editor to save your code.

7. Test the Function

• Click Test at the top of the page.
• A box will pop up asking for a test event name. Just type something like test1 and click Save.
• Click Test again.

AWS will run your Lambda function and show the result at the bottom of the page:

{
  "statusCode": 200,
  "body": "Hello from Lambda!"
}

{
  "statusCode": 200,
  "body": "Hello from Lambda!"
}

That’s AWS running your code and giving you the response.

What You Just Did

• You created a Lambda function in AWS.
• You wrote simple Python code inside it.
• You clicked Test to run it.
• AWS showed you the output of your code.

You now have a working Lambda function in your account.

Writing Lambda locally with Python

We are going to write our Lambda function on our own computer first, then upload it to AWS Lambda.

Why write locally?

When you use the AWS Console, you’re typing directly in the browser.
But in real projects, you want your code saved on your computer. This allows version control (Git), testing locally, and easier editing.

Create a Python file

On your computer, open any folder.
Create a new file named lambda_function.py.

Inside it, paste this code:

def lambda_handler(event, context):
    name = event.get('name', 'World')  # If 'name' not provided, default is 'World'
    return {
        'statusCode': 200,
        'body': f'Hello {name}!'
    }

def lambda_handler(event, context):
    name = event.get('name', 'World')  # If 'name' not provided, default is 'World'
    return {
        'statusCode': 200,
        'body': f'Hello {name}!'
    }

This is your Lambda function.

• event = input AWS sends (like JSON data).
• context = info about the runtime.
• The function returns a statusCode and body.

Compress it into a ZIP

AWS Lambda wants your code uploaded as a .zip file.

In your terminal (Command Prompt, PowerShell, or Linux terminal), run:

zip function.zip lambda_function.py

zip function.zip lambda_function.py

Now you have function.zip containing your code.
(This step literally zips your file.)

Upload the ZIP to AWS

You can upload it in two ways:

Option A – AWS Console:

• Go to AWS Console → Lambda → Create Function.
• Name it hello_lambda_cli.
• Runtime: Python 3.9.
• Under “Code”, choose Upload from .zip file.
• Upload function.zip.
• Choose an execution role (it gives AWS permissions).
• Save.

Option B – AWS CLI (command line):

Run this command in your terminal:

aws lambda create-function \
  --function-name hello_lambda_cli \
  --runtime python3.9 \
  --role arn:aws:iam::<your-account-id>:role/<lambda-execution-role> \
  --handler lambda_function.lambda_handler \
  --zip-file fileb://function.zip

aws lambda create-function \
  --function-name hello_lambda_cli \
  --runtime python3.9 \
  --role arn:aws:iam::<your-account-id>:role/<lambda-execution-role> \
  --handler lambda_function.lambda_handler \
  --zip-file fileb://function.zip

Here:

• --function-name = name of your Lambda.
• --runtime = Python version.
• --role = IAM role ARN (permissions for Lambda to run).
• --handler = entry point to your code.
• --zip-file = your zipped code file.

Test it

After uploading, go to AWS Lambda Console → your function → Test.
Input:

{
  "name": "Emmimal"
}

{
  "name": "Emmimal"
}

Output will be:

{
  "statusCode": 200,
  "body": "Hello Emmimal!"
}

{
  "statusCode": 200,
  "body": "Hello Emmimal!"
}

Introducing Boto3

What is Boto3?

• Boto3 is a Python library created by AWS.
• It lets your Python code talk directly to AWS services (like Lambda, S3, DynamoDB).
• Instead of clicking buttons in the AWS Console, you can write Python code to do the same tasks.

Example:

• Console way: Click “Upload” → Upload ZIP.
• Boto3 way: lambda_client.create_function(...) from Python.

This is great for automation and large-scale projects.

Installing Boto3

In your terminal (Command Prompt, PowerShell, or Linux shell):

pip install boto3

pip install boto3

If it already says “Requirement already satisfied” — you’re good.

Using Boto3 in Python

In your Python script, at the top:

import boto3

import boto3

This lets you use AWS services via Python code.

Then you create a client for a service. For Lambda:

lambda_client = boto3.client('lambda')

lambda_client = boto3.client('lambda')

Now lambda_client can create, update, delete, or invoke Lambda functions programmatically.

Why this matters

Up until now, you:

• Created Lambda using Console.
• Uploaded code manually.

With Boto3, you can:

• Upload new versions automatically.
• Invoke (call) your Lambda functions from Python.
• List or delete functions.
• Automate deployments.

Managing AWS Lambda with Python (Boto3) – Step by Step

We will learn how to create, run, update, and delete Lambda functions using Python.

Step 0: Prerequisites

1. Python 3.x installed on your computer.
2. AWS CLI installed and configured with your credentials. aws configure Enter: AWS Access Key, Secret Key, Region, output format.
3. Boto3 installed: pip install boto3

Step 1: Write Your Lambda Code Locally

Create a file called lambda_function.py:

def lambda_handler(event, context):
    name = event.get('name', 'World')  # If no name given, default is 'World'
    return {
        'statusCode': 200,
        'body': f'Hello {name}!'
    }

def lambda_handler(event, context):
    name = event.get('name', 'World')  # If no name given, default is 'World'
    return {
        'statusCode': 200,
        'body': f'Hello {name}!'
    }

Explanation:

• lambda_handler is the function AWS will run.
• event contains input data.
• context contains AWS runtime info (we don’t use it here).
• It returns a message like "Hello Emmimal!".

Step 2: Zip the Code

Lambda needs the code in a .zip file:

zip function.zip lambda_function.py

zip function.zip lambda_function.py

This creates function.zip that contains your code.

Step 3: Create a Lambda Function Using Boto3

Create a new Python file, for example manage_lambda.py:

import boto3

# Connect to AWS Lambda
lambda_client = boto3.client('lambda')

# Read the zipped Lambda code
with open('function.zip', 'rb') as f:
    zipped_code = f.read()

# Create the Lambda function
response = lambda_client.create_function(
    FunctionName='hello_lambda_boto3',                  # Name of your Lambda
    Runtime='python3.9',                                # Python version
    Role='arn:aws:iam::<your-account-id>:role/<lambda-execution-role>',  # IAM role
    Handler='lambda_function.lambda_handler',           # File.function_name
    Code={'ZipFile': zipped_code},                      # Upload zipped code
    Timeout=15,                                         # Max time Lambda can run
    MemorySize=128,                                     # Memory in MB
    Publish=True                                        # Make it live immediately
)

print("Lambda Created:")
print(response)

import boto3

# Connect to AWS Lambda
lambda_client = boto3.client('lambda')

# Read the zipped Lambda code
with open('function.zip', 'rb') as f:
    zipped_code = f.read()

# Create the Lambda function
response = lambda_client.create_function(
    FunctionName='hello_lambda_boto3',                  # Name of your Lambda
    Runtime='python3.9',                                # Python version
    Role='arn:aws:iam::<your-account-id>:role/<lambda-execution-role>',  # IAM role
    Handler='lambda_function.lambda_handler',           # File.function_name
    Code={'ZipFile': zipped_code},                      # Upload zipped code
    Timeout=15,                                         # Max time Lambda can run
    MemorySize=128,                                     # Memory in MB
    Publish=True                                        # Make it live immediately
)

print("Lambda Created:")
print(response)

What this does:

• Creates a new Lambda function on AWS.
• Uses your zipped code.
• Sets runtime, memory, timeout, and IAM role.

Step 4: Invoke (Run) the Lambda Function

Add this to the same file or a new file:

import boto3

lambda_client = boto3.client('lambda')

# Call the Lambda function
response = lambda_client.invoke(
    FunctionName='hello_lambda_boto3',      # Lambda name
    InvocationType='RequestResponse',       # Wait for the response
    Payload=b'{"name":"Emmimal"}'           # Input data
)

# Read and print the output
print(response['Payload'].read().decode('utf-8'))

import boto3

lambda_client = boto3.client('lambda')

# Call the Lambda function
response = lambda_client.invoke(
    FunctionName='hello_lambda_boto3',      # Lambda name
    InvocationType='RequestResponse',       # Wait for the response
    Payload=b'{"name":"Emmimal"}'           # Input data
)

# Read and print the output
print(response['Payload'].read().decode('utf-8'))

Output:

{"statusCode": 200, "body": "Hello Emmimal!"}

{"statusCode": 200, "body": "Hello Emmimal!"}

Step 5: Update Lambda Code

If you change lambda_function.py, zip it again, then run:

import boto3

lambda_client = boto3.client('lambda')

with open('function.zip', 'rb') as f:
    zipped_code = f.read()

lambda_client.update_function_code(
    FunctionName='hello_lambda_boto3',
    ZipFile=zipped_code,
    Publish=True
)

print("Lambda Code Updated")

import boto3

lambda_client = boto3.client('lambda')

with open('function.zip', 'rb') as f:
    zipped_code = f.read()

lambda_client.update_function_code(
    FunctionName='hello_lambda_boto3',
    ZipFile=zipped_code,
    Publish=True
)

print("Lambda Code Updated")

Explanation:

• This replaces the old Lambda code with the new code.
• Keeps the same name, memory, and settings.

Step 6: Delete Lambda Function

import boto3

lambda_client = boto3.client('lambda')

lambda_client.delete_function(FunctionName='hello_lambda_boto3')

print("Lambda Deleted")

import boto3

lambda_client = boto3.client('lambda')

lambda_client.delete_function(FunctionName='hello_lambda_boto3')

print("Lambda Deleted")

Explanation:

• This permanently deletes your Lambda function from AWS.

Summary

Environment Variables in Lambda

What Are Environment Variables?

Environment variables are key-value pairs you can store in your Lambda function.

Use them for:

• API keys or secrets
• Configurations like “dev” or “production” stage
• Other settings you don’t want to hard-code in your code

Setting Environment Variables Using Boto3

You can set them with update_function_configuration():

import boto3

lambda_client = boto3.client('lambda')

lambda_client.update_function_configuration(
    FunctionName='hello_lambda_boto3',
    Environment={
        'Variables': {
            'STAGE': 'dev',
            'API_KEY': '12345'
        }
    }
)

print("Environment variables set!")

import boto3

lambda_client = boto3.client('lambda')

lambda_client.update_function_configuration(
    FunctionName='hello_lambda_boto3',
    Environment={
        'Variables': {
            'STAGE': 'dev',
            'API_KEY': '12345'
        }
    }
)

print("Environment variables set!")

Explanation:

• STAGE → environment name, e.g., “dev” or “prod”
• API_KEY → secret key or any configuration value
• AWS stores these securely with the Lambda function

Access Environment Variables Inside Lambda

Inside your lambda_function.py, use Python’s os.environ:

import os

def lambda_handler(event, context):
    stage = os.environ.get('STAGE', 'default')   # Get the STAGE variable
    api_key = os.environ.get('API_KEY', 'none')  # Get API_KEY variable
    return {
        'stage': stage,
        'api_key': api_key
    }

import os

def lambda_handler(event, context):
    stage = os.environ.get('STAGE', 'default')   # Get the STAGE variable
    api_key = os.environ.get('API_KEY', 'none')  # Get API_KEY variable
    return {
        'stage': stage,
        'api_key': api_key
    }

Explanation:

• os.environ.get('KEY', 'default') reads the variable
• If the variable doesn’t exist, it returns the default value

Test Example

If you set:

STAGE = dev  
API_KEY = 12345

STAGE = dev  
API_KEY = 12345

And run your Lambda function, the output will be:

{
  "stage": "dev",
  "api_key": "12345"
}

{
  "stage": "dev",
  "api_key": "12345"
}

Why This Is Useful:

• You don’t hard-code secrets in your code.
• You can change environment variables without updating the Lambda code.
• Makes Lambda functions flexible and safe.

Permissions and IAM Roles for AWS Lambda

Why Lambda Needs Permissions

Lambda functions run inside AWS, but they cannot automatically access other AWS services like S3, DynamoDB, or SNS.

• If your Lambda tries to read an S3 bucket without proper permissions, it will fail.
• Permissions are controlled using IAM roles.

What is an IAM Role?

An IAM role is like a special ID card for your Lambda function.
It tells AWS what the Lambda is allowed to do, for example:

• Read files from S3
• Write to DynamoDB
• Publish messages to SNS

When you create a Lambda, you attach a role to it to grant the necessary permissions.

Example: S3 Read-Only Access

1. Go to the IAM Console → Roles → Create Role
2. Choose AWS Service → Lambda
3. Attach the policy AmazonS3ReadOnlyAccess
4. Give the role a name, for example lambda_s3_readonly_role
5. Use the role ARN when creating your Lambda:

Role='arn:aws:iam::<your-account-id>:role/lambda_s3_readonly_role'

Role='arn:aws:iam::<your-account-id>:role/lambda_s3_readonly_role'

Now the Lambda function can read from S3 buckets but cannot write to them.

Common IAM Policies for Lambda

Using IAM Role in Code

When creating a Lambda function with Boto3:

response = lambda_client.create_function(
    FunctionName='hello_lambda_boto3',
    Runtime='python3.9',
    Role='arn:aws:iam::<your-account-id>:role/lambda_s3_readonly_role',
    Handler='lambda_function.lambda_handler',
    Code={'ZipFile': zipped_code},
)

response = lambda_client.create_function(
    FunctionName='hello_lambda_boto3',
    Runtime='python3.9',
    Role='arn:aws:iam::<your-account-id>:role/lambda_s3_readonly_role',
    Handler='lambda_function.lambda_handler',
    Code={'ZipFile': zipped_code},
)

• Role specifies what permissions the Lambda has.
• Without the correct role or permissions, any attempt to access AWS services will fail.

Key Point

Always attach minimal permissions required by the Lambda function.
This reduces security risks and keeps your AWS account safe.

AWS Lambda Triggers Explained

A trigger is something that tells your Lambda when to run automatically.
Without a trigger, Lambda will only run if you manually invoke it.

S3 Trigger – Run Lambda When a File is Uploaded

Scenario: You want your Lambda to say “Hello” every time someone uploads a file to S3.

Steps in Plain Words:

1. Go to the S3 bucket you want to watch.
2. Click Properties → Event Notifications → Create event notification.
3. Give it a name.
4. Select Event types → All object create events (so it runs whenever a file is uploaded).
5. Choose Destination → Lambda Function.
6. Select your Lambda function (for example, hello_lambda_boto3).
7. Save.

Now: Every time someone uploads a file to that bucket, your Lambda runs automatically.

Optional: Code Example for Lambda to Read the Uploaded File

import json

def lambda_handler(event, context):
    # event contains info about the file uploaded
    file_name = event['Records'][0]['s3']['object']['key']
    bucket_name = event['Records'][0]['s3']['bucket']['name']
    
    print(f"File {file_name} uploaded to bucket {bucket_name}")
    
    return {
        'statusCode': 200,
        'body': json.dumps(f"Processed {file_name} from {bucket_name}")
    }

import json

def lambda_handler(event, context):
    # event contains info about the file uploaded
    file_name = event['Records'][0]['s3']['object']['key']
    bucket_name = event['Records'][0]['s3']['bucket']['name']
    
    print(f"File {file_name} uploaded to bucket {bucket_name}")
    
    return {
        'statusCode': 200,
        'body': json.dumps(f"Processed {file_name} from {bucket_name}")
    }

• event is automatically sent by S3 to Lambda.
• It contains file name, bucket name, and other metadata.

CloudWatch Trigger – Run Lambda on a Schedule

Scenario: You want your Lambda to run every hour, like a scheduled task.

Steps:

1. Open CloudWatch → Rules → Create rule.
2. Choose Event Source → Schedule.
3. Enter a schedule expression like rate(1 hour) (runs every hour).
4. Choose Target → Lambda Function.
5. Save.

Now: Lambda runs automatically every hour.

Optional Lambda Code Example:

def lambda_handler(event, context):
    print("Lambda ran on schedule!")
    return {"statusCode": 200, "body": "Scheduled Lambda executed"}

def lambda_handler(event, context):
    print("Lambda ran on schedule!")
    return {"statusCode": 200, "body": "Scheduled Lambda executed"}

API Gateway Trigger – Run Lambda via HTTP Request

Scenario: You want to call your Lambda function by visiting a web URL.

Steps:

1. Open API Gateway → Create API → HTTP API.
2. Add a route, e.g., /hello.
3. Integrate the route with your Lambda function.
4. Deploy the API.

Now: Visiting the URL triggers your Lambda.

Optional Lambda Code Example:

def lambda_handler(event, context):
    name = event.get("queryStringParameters", {}).get("name", "World")
    return {"statusCode": 200, "body": f"Hello {name}!"}

def lambda_handler(event, context):
    name = event.get("queryStringParameters", {}).get("name", "World")
    return {"statusCode": 200, "body": f"Hello {name}!"}

• Example URL: https://<api-id>.execute-api.<region>.amazonaws.com/hello?name=Emmimal
• Response: "Hello Emmimal!"

Key Takeaways

Packaging Dependencies with Lambda Layers

What Are Lambda Layers?

• Lambda functions have a limited environment with some default Python libraries.
• If your code needs extra libraries (like requests, pandas, numpy), you can’t just pip install inside Lambda.
• Lambda Layers let you package these dependencies separately and attach them to your function.
• Multiple Lambdas can share the same layer, saving space and effort.

Create a Lambda Layer Locally

Step 1: Create a folder for dependencies

mkdir python

mkdir python

Step 2: Install libraries into that folder

pip install requests -t python/

pip install requests -t python/

• -t python/ installs the library into the python folder.
• You can install multiple libraries the same way.

Step 3: Zip the folder

zip -r layer.zip python

zip -r layer.zip python

• layer.zip will contain your dependencies.

Upload the Layer to AWS

1. Go to AWS Console → Lambda → Layers → Create Layer.
2. Give it a name, e.g., requests-layer.
3. Upload layer.zip.
4. Choose Compatible runtimes, e.g., Python 3.9.
5. Click Create.

Attach Layer to Your Lambda Function

• Open your Lambda function in the console.
• Go to Layers → Add a layer → Custom layers → Select your layer.
• Click Add.

Now your Lambda can use the libraries in the layer.

Use the Library in Lambda Code

Example Lambda function using requests:

import requests

def lambda_handler(event, context):
    response = requests.get("https://api.github.com")
    return {
        'statusCode': 200,
        'body': response.text[:200]  # return first 200 characters
    }

import requests

def lambda_handler(event, context):
    response = requests.get("https://api.github.com")
    return {
        'statusCode': 200,
        'body': response.text[:200]  # return first 200 characters
    }

• The requests library is not installed in Lambda by default.
• The layer makes it available.

Key Points

• Layers are reusable across multiple Lambda functions.
• You can update a layer independently from the Lambda code.
• Max 5 layers per Lambda function.

AWS Lambda Logging and Monitoring – Step by Step

Lambda functions don’t run on your computer; they run on AWS servers. To see what’s happening inside your code, you need logs. To track performance, you need metrics. AWS gives us CloudWatch for both.

Logging in Lambda

You can use Python’s built-in logging module to print messages to CloudWatch Logs.

Example:

import logging

# Create a logger object
logger = logging.getLogger()
logger.setLevel(logging.INFO)  # We want info-level messages

def lambda_handler(event, context):
    logger.info("Lambda function started")  # Log message at start

    # Example processing
    name = event.get("name", "World")
    logger.info(f"Processing request for {name}")  # Log the input

    response = {"message": f"Hello {name}!"}
    logger.info(f"Response generated: {response}")  # Log the output

    return {"statusCode": 200, "body": response}

import logging

# Create a logger object
logger = logging.getLogger()
logger.setLevel(logging.INFO)  # We want info-level messages

def lambda_handler(event, context):
    logger.info("Lambda function started")  # Log message at start

    # Example processing
    name = event.get("name", "World")
    logger.info(f"Processing request for {name}")  # Log the input

    response = {"message": f"Hello {name}!"}
    logger.info(f"Response generated: {response}")  # Log the output

    return {"statusCode": 200, "body": response}

Explanation:

• logger.info() sends messages to CloudWatch Logs.
• You can also use logger.error() for errors or logger.warning() for warnings.
• Every time Lambda runs, these messages are automatically stored in CloudWatch.

Viewing Logs

1. Go to AWS Console → CloudWatch → Logs.
2. Find the log group for your Lambda (e.g., /aws/lambda/hello_lambda_boto3).
3. Open the latest log stream.

You will see something like this:

START RequestId: 12345 Version: $LATEST
Lambda function started
Processing request for Emmimal
Response generated: {'message': 'Hello Emmimal!'}
END RequestId: 12345
REPORT RequestId: 12345  Duration: 45.00 ms  Memory: 128 MB

START RequestId: 12345 Version: $LATEST
Lambda function started
Processing request for Emmimal
Response generated: {'message': 'Hello Emmimal!'}
END RequestId: 12345
REPORT RequestId: 12345  Duration: 45.00 ms  Memory: 128 MB

• START and END are added by Lambda automatically.
• Your logger.info() messages are between them.

Monitoring Metrics

AWS automatically tracks Lambda performance metrics in CloudWatch:

• Open CloudWatch → Metrics → Lambda to see graphs.
• You can set Alarms: for example, get notified if errors > 5 in an hour.

Why This Is Useful

• Logs help you debug your code.
• Metrics help you monitor performance and detect problems early.
• Together, they make your Lambda reliable and maintainable.

Advanced Topics

Asynchronous Invocations

By default, when you call a Lambda, your code waits for the result. This is called synchronous invocation.

Sometimes, you want Lambda to run in the background without waiting. This is asynchronous invocation.

Example:

import boto3

lambda_client = boto3.client('lambda')

# Asynchronous call
response = lambda_client.invoke(
    FunctionName='hello_lambda_boto3',
    InvocationType='Event',  # Event = asynchronous
    Payload=b'{"name":"Emmimal"}'
)

print("Lambda triggered asynchronously")

import boto3

lambda_client = boto3.client('lambda')

# Asynchronous call
response = lambda_client.invoke(
    FunctionName='hello_lambda_boto3',
    InvocationType='Event',  # Event = asynchronous
    Payload=b'{"name":"Emmimal"}'
)

print("Lambda triggered asynchronously")

• InvocationType='Event' tells AWS: don’t wait for a response.
• Lambda runs in the background.
• Useful for tasks that take time or don’t need an immediate response.

Concurrency

Lambda can scale automatically. Concurrency controls how many Lambda instances run at the same time.

• Default: unlimited (AWS scales automatically).
• Reserved concurrency: limits the max number of simultaneous executions.

Example:

lambda_client.put_function_concurrency(
    FunctionName='hello_lambda_boto3',
    ReservedConcurrentExecutions=5  # max 5 instances at once
)

lambda_client.put_function_concurrency(
    FunctionName='hello_lambda_boto3',
    ReservedConcurrentExecutions=5  # max 5 instances at once
)

• If 10 events come in at the same time, only 5 will run, and the rest wait.
• Helps prevent overloading downstream resources like databases.

Dead Letter Queues (DLQ)

Sometimes Lambda fails to process an event.

• A Dead Letter Queue stores failed events so you can debug them later.
• DLQ can be an SQS queue or SNS topic.

How it works:

1. Attach a DLQ to Lambda.
2. When Lambda fails, the event is sent to SQS/SNS.
3. You can inspect it and take corrective action.

Example DLQ setup is usually done in the console or via Boto3 like this:

lambda_client.update_function_configuration(
    FunctionName='hello_lambda_boto3',
    DeadLetterConfig={
        'TargetArn': 'arn:aws:sqs:<region>:<account-id>:my-dlq'
    }
)

lambda_client.update_function_configuration(
    FunctionName='hello_lambda_boto3',
    DeadLetterConfig={
        'TargetArn': 'arn:aws:sqs:<region>:<account-id>:my-dlq'
    }
)

Using Step Functions with Lambda

Step Functions let you combine multiple Lambda functions into a workflow.

• Example workflow:
  1. Lambda A processes data
  2. Lambda B stores results in a database
  3. Lambda C sends a notification
• Step Functions manage:
  • Execution order
  • Error handling
  • Retries

This allows complex applications without writing complicated code in a single Lambda.

Summary of Advanced Topics

Best Practices

Keep Functions Small and Focused

• Each Lambda should do one thing only.
• Benefits:
  • Easier to maintain
  • Faster execution
  • Simpler debugging
• Example:
  • One Lambda processes uploaded files
  • Another Lambda sends notifications

Use Environment Variables for Configuration

• Store secrets, API keys, or configuration values outside your code.
• Makes your Lambda flexible and secure.
• Example:

• Change STAGE without touching code.

import os

def lambda_handler(event, context):
    stage = os.environ.get('STAGE', 'dev')
    return {"stage": stage}

import os

def lambda_handler(event, context):
    stage = os.environ.get('STAGE', 'dev')
    return {"stage": stage}

Monitor Execution Time and Memory Usage

• Lambda charges are based on compute time and memory.
• Use CloudWatch metrics to check:
  • Duration
  • Memory usage
  • Errors
• Optimize code to reduce cost and improve performance.

Version Your Lambda Functions

• Use versions to track stable releases of your Lambda.
• Allows safe deployment and rollbacks.
• Example:

# Publish a new version via Boto3
lambda_client.publish_version(FunctionName='hello_lambda_boto3')

# Publish a new version via Boto3
lambda_client.publish_version(FunctionName='hello_lambda_boto3')

• You can also use aliases to point to production or testing versions.

Use Layers for Shared Dependencies

• If multiple Lambdas use the same libraries, put them in a Layer.
• Benefits:
  • Saves deployment size
  • Easier to update shared libraries
• Example: Layer for requests or pandas.

mkdir python
pip install requests -t python/
zip -r layer.zip python

mkdir python
pip install requests -t python/
zip -r layer.zip python

• Upload layer.zip to Lambda Layers and attach it to functions.

Summary

• Small, single-purpose functions
• Configurable via environment variables
• Monitor and optimize performance
• Version functions for safety
• Use layers for shared libraries

Following these practices keeps your Lambda functions efficient, maintainable, and secure.

Wrapping Up

We’ve covered everything from basic to advanced AWS Lambda with Boto3 and Python.
You learned how to:

• Create Lambda functions manually and programmatically.
• Invoke, update, and delete them using Boto3.
• Use environment variables, triggers, and permissions.
• Handle advanced topics like concurrency, DLQ, and Step Functions.

With this knowledge, you can confidently build and manage serverless applications on AWS.

Recommended External References

Official Documentation

1. AWS Lambda Developer Guide
   https://docs.aws.amazon.com/lambda/latest/dg/welcome.html
   • Comprehensive guide from AWS covering Lambda setup, triggers, environment variables, and best practices.
2. Boto3 Documentation
   https://boto3.amazonaws.com/v1/documentation/api/latest/index.html
   • The official Python SDK documentation for AWS. Includes examples for Lambda management, S3, DynamoDB, and more.
3. AWS Lambda Function Configuration
   https://docs.aws.amazon.com/lambda/latest/dg/configuration.html
   • Details about memory, timeout, environment variables, concurrency, and execution roles.
4. AWS Lambda Layers
   https://docs.aws.amazon.com/lambda/latest/dg/configuration-layers.html
   • Explains how to package dependencies in layers for reusable code.
5. AWS Step Functions with Lambda
   https://docs.aws.amazon.com/step-functions/latest/dg/welcome.html
   • Guide for orchestrating multiple Lambda functions into workflows.

Tutorials and Guides

6. Serverless Framework – AWS Lambda Tutorial
   https://www.serverless.com/framework/docs/providers/aws/guide/intro/
   • Step-by-step tutorials for building serverless applications with Lambda.
7. AWS Blog – Lambda Best Practices
   https://aws.amazon.com/blogs/compute/category/compute/aws-lambda/
   • Latest updates, best practices, and real-world use cases for AWS Lambda.

Forums and Community

9. GitHub – AWS Labs and Examples
   https://github.com/awslabs

• Open-source projects and Lambda examples maintained by AWS Labs.

FAQs

About The Author

Emmimal Alexander

Emmimal Alexander is an AI & Machine Learning Expert, passionate educator, and the author of “Neural Networks and Deep Learning with Python.” As the founder of EmiTechLogic, she’s on a mission to make complex tech topics accessible, engaging, and empowering for learners at every level.

With deep expertise in Python, HTML, JavaScript, and CSS, Emmimal brings a strong coding foundation to her tutorials and educational resources. Her work focuses on blending theoretical understanding with real-world application—so readers not only learn how things work, but also why they matter.

Through EmiTechLogic, she creates hands-on guides, detailed breakdowns, and project-based learning content that bridges the gap between academic concepts and practical implementation. Whether you’re exploring AI for the first time or fine-tuning your neural networks, you’re in the right place.

See author's posts