AWS Boto3 for Cloud Automation

AWS Boto3 is the official Python SDK for interacting with Amazon Web Services (AWS). It allows developers to automate cloud tasks such as:
Creating EC2 instances
Managing S3 buckets
Deploying Lambda functions
Handling IAM roles and permissions
Automating CloudWatch monitoring

This guide covers installing Boto3, authenticating AWS access, and automating AWS resources using Python.

1. Installing Boto3 and Configuring AWS Credentials

1.1 Installing Boto3

First, install the Boto3 library:

pip install boto3

1.2 Setting Up AWS Credentials

Boto3 requires AWS credentials for authentication. You can configure them in multiple ways:

Option 1: Using AWS CLI (Recommended)

Run the following command:

aws configure

You’ll be prompted to enter:

AWS Access Key ID: <your_access_key>
AWS Secret Access Key: <your_secret_key>
Default region name: us-east-1
Default output format: json

This saves credentials in ~/.aws/credentials.

Option 2: Using Environment Variables

export AWS_ACCESS_KEY_ID='your_access_key'
export AWS_SECRET_ACCESS_KEY='your_secret_key'
export AWS_DEFAULT_REGION='us-east-1'

Option 3: Using a Credentials File

Create ~/.aws/credentials and add:

[default]
aws_access_key_id=your_access_key
aws_secret_access_key=your_secret_key
region=us-east-1

2. Automating AWS with Boto3

Boto3 has two interfaces:
Client (low-level) – Direct access to AWS APIs
Resource (high-level) – Object-oriented approach

3. Managing AWS Resources with Boto3

3.1 Creating and Managing EC2 Instances

Launching an EC2 instance

import boto3

ec2 = boto3.resource('ec2')

instance = ec2.create_instances(
    ImageId='ami-12345678',  # Replace with a valid AMI ID
    InstanceType='t2.micro',
    MinCount=1,
    MaxCount=1,
    KeyName='your-key-pair',
    SecurityGroupIds=['sg-xxxxxxxx'],
    SubnetId='subnet-xxxxxxxx'
)[0]

print(f'EC2 instance {instance.id} is launching...')

Stopping and Terminating an EC2 instance

ec2_client = boto3.client('ec2')

# Stop an instance
ec2_client.stop_instances(InstanceIds=['i-1234567890abcdef'])

# Terminate an instance
ec2_client.terminate_instances(InstanceIds=['i-1234567890abcdef'])

Fetching all running instances

instances = ec2.instances.filter(Filters=[{'Name': 'instance-state-name', 'Values': ['running']}])

for instance in instances:
    print(instance.id, instance.state)

3.2 Managing S3 Buckets

Creating an S3 bucket

s3 = boto3.client('s3')

s3.create_bucket(
    Bucket='my-boto3-bucket',
    CreateBucketConfiguration={'LocationConstraint': 'us-east-1'}
)

print("S3 bucket created successfully!")

Uploading a file to S3

s3.upload_file('localfile.txt', 'my-boto3-bucket', 'uploadedfile.txt')

Listing all S3 buckets

buckets = s3.list_buckets()

for bucket in buckets['Buckets']:
    print(bucket['Name'])

Deleting an S3 bucket

s3.delete_bucket(Bucket='my-boto3-bucket')

3.3 Automating AWS Lambda Deployment

Deploying a Lambda function

lambda_client = boto3.client('lambda')

with open('lambda_function.zip', 'rb') as f:
    lambda_code = f.read()

response = lambda_client.create_function(
    FunctionName='myLambdaFunction',
    Runtime='python3.8',
    Role='arn:aws:iam::123456789012:role/lambda-role',
    Handler='lambda_function.lambda_handler',
    Code={'ZipFile': lambda_code}
)

print("Lambda function deployed:", response['FunctionArn'])

Invoking a Lambda function

response = lambda_client.invoke(
    FunctionName='myLambdaFunction',
    InvocationType='RequestResponse'
)

print(response['Payload'].read().decode())

Deleting a Lambda function

lambda_client.delete_function(FunctionName='myLambdaFunction')

3.4 Managing IAM Users and Roles

Creating an IAM user

iam = boto3.client('iam')

iam.create_user(UserName='newUser')

print("IAM user created!")

Attaching a policy to an IAM user

iam.attach_user_policy(
    UserName='newUser',
    PolicyArn='arn:aws:iam::aws:policy/AmazonS3ReadOnlyAccess'
)

Deleting an IAM user

iam.delete_user(UserName='newUser')

3.5 Monitoring AWS Services with CloudWatch

Fetching EC2 CPU utilization

cloudwatch = boto3.client('cloudwatch')

response = cloudwatch.get_metric_statistics(
    Namespace='AWS/EC2',
    MetricName='CPUUtilization',
    Dimensions=[{'Name': 'InstanceId', 'Value': 'i-1234567890abcdef'}],
    StartTime=datetime.utcnow() - timedelta(minutes=10),
    EndTime=datetime.utcnow(),
    Period=300,
    Statistics=['Average']
)

print(response['Datapoints'])

Setting up an alarm for high CPU usage

cloudwatch.put_metric_alarm(
    AlarmName='HighCPUUsage',
    MetricName='CPUUtilization',
    Namespace='AWS/EC2',
    Statistic='Average',
    Period=300,
    EvaluationPeriods=2,
    Threshold=80.0,
    ComparisonOperator='GreaterThanThreshold',
    AlarmActions=['arn:aws:sns:us-east-1:123456789012:MyTopic']
)