Boto3

 

🧠 Understanding Boto3: Overview

Boto3 is the official AWS SDK for Python, used to interact with AWS services like S3, EC2, Lambda, DynamoDB, etc.

---

⚙️ 1. boto3.client() vs boto3.resource() vs boto3.session()

✅ boto3.client(service_name, ...)

• Low-level client.

• Maps 1:1 to AWS service APIs.

• Returns response dicts (JSON-like).

• Example: client('s3')

✅ boto3.resource(service_name, ...)

• High-level abstraction.

• Uses Python objects.

• Easier for common operations (like bucket.upload_file(...))

⚠️ Not available for all AWS services.

✅ boto3.session.Session(...)

• Used to manage configuration: profiles, credentials, and regions.

• You can have multiple sessions, for example for multi-account or multi-region setups.

---

🔁 When to Use What

Feature	Use client()	Use resource()	Use session()
Needs raw API access	✅ Yes	❌ No	❌ Use session.client()
Object-based actions	❌ Too verbose	✅ Ideal	✅ For multi-profile access
Working across profiles	❌ Only default or env vars	❌ Same	✅ Fully supported
Need flexibility	✅ Advanced control	❌ Less control	✅ Multi-region and credential flexibility

---

✅ Example 1: Using boto3.client to List S3 Buckets

import boto3

s3_client = boto3.client('s3', region_name='us-east-1')

def list_buckets():
    response = s3_client.list_buckets()
    for bucket in response['Buckets']:
        print(f"- {bucket['Name']}")

Why use client?
We want direct access to AWS API to fetch raw data like bucket names.

---

✅ Example 2: Using boto3.resource to Upload File to S3

import boto3

s3 = boto3.resource('s3')
bucket = s3.Bucket('my-bucket-name')

def upload_file():
    bucket.upload_file(Filename='file.txt', Key='uploaded_file.txt')

Why use resource?
This is a high-level operation (upload_file) which is easier with resource than calling put_object() manually with client.

---

✅ Example 3: Using boto3.session for Multiple Profiles

Let's say you have 2 AWS profiles: dev and prod.

import boto3

def get_instance_count(profile_name):
    session = boto3.Session(profile_name=profile_name)
    ec2 = session.client('ec2', region_name='us-east-1')
    
    instances = ec2.describe_instances()
    total = sum(len(reservation['Instances']) for reservation in instances['Reservations'])
    print(f"{profile_name} has {total} EC2 instance(s).")

get_instance_count('dev')
get_instance_count('prod')

Why use session?
Each session uses its own credentials and region. Useful for multi-account management.

---

✅ Example 4: Using boto3.session to Assume Role into Another Account

import boto3

def assume_role_and_list_s3(role_arn):
    base_session = boto3.Session()
    sts_client = base_session.client('sts')

    assumed_role = sts_client.assume_role(
        RoleArn=role_arn,
        RoleSessionName='CrossAccountSession'
    )

    credentials = assumed_role['Credentials']
    temp_session = boto3.Session(
        aws_access_key_id=credentials['AccessKeyId'],
        aws_secret_access_key=credentials['SecretAccessKey'],
        aws_session_token=credentials['SessionToken']
    )

    s3 = temp_session.client('s3')
    buckets = s3.list_buckets()
    for b in buckets['Buckets']:
        print(b['Name'])

assume_role_and_list_s3("arn:aws:iam::123456789012:role/SomeRole")

Why use session?
You can create temporary sessions with assumed roles — essential in enterprise, multi-account setups.

---

🧪 Quick Summary Table

Use Case	Method Used	Why?
List S3 buckets	boto3.client()	Raw API for precise data
Upload files to S3	boto3.resource()	High-level object methods
Switch between dev and prod accounts	boto3.session()	Supports multiple profiles
Cross-account access with STS assume role	boto3.session()	Use temporary credentials via STS

---

🧰 Pro Tip

Use session.client() or session.resource() like this:

session = boto3.Session(profile_name='dev')
s3_client = session.client('s3')

It gives you flexibility + cleaner multi-env support.

✅ 1. S3 File Operations (Upload, List, Download)

python
import boto3

# High-level resource
s3 = boto3.resource('s3')
bucket_name = 'my-demo-bucket'

# Upload a file
s3.Bucket(bucket_name).upload_file('local.txt', 'uploaded.txt')

# List objects
for obj in s3.Bucket(bucket_name).objects.all():
    print(f'File in bucket: {obj.key}')

# Download a file
s3.Bucket(bucket_name).download_file('uploaded.txt', 'downloaded.txt')

✅ Use resource for S3 when you want file operations, cleaner syntax, and auto-pagination.

---

✅ 2. EC2: Launch Instance, List Instances

python
import boto3

ec2 = boto3.resource('ec2')

# Launch new EC2 instance
instances = ec2.create_instances(
    ImageId='ami-0c55b159cbfafe1f0',
    InstanceType='t2.micro',
    MinCount=1,
    MaxCount=1
)

print("Launched instance ID:", instances[0].id)

# List running instances
for instance in ec2.instances.filter(Filters=[{'Name': 'instance-state-name', 'Values': ['running']}]):
    print(instance.id, instance.instance_type, instance.state['Name'])

✅ boto3.resource('ec2') is great for managing instances in an object-oriented way.

---

✅ 3. EKS: List Clusters and Get Cluster Info

python
import boto3

eks = boto3.client('eks')

# List all EKS clusters
response = eks.list_clusters()
print("Clusters:", response['clusters'])

# Get details for a specific cluster
cluster_info = eks.describe_cluster(name='my-eks-cluster')
print("Cluster status:", cluster_info['cluster']['status'])

❗ EKS supports only client, not resource.

---

✅ 4. Lambda: List and Invoke a Function

python
import boto3
import json

lambda_client = boto3.client('lambda')

# List Lambda functions
functions = lambda_client.list_functions()
for func in functions['Functions']:
    print(func['FunctionName'])

# Invoke a function
response = lambda_client.invoke(
    FunctionName='my-function-name',
    InvocationType='RequestResponse',
    Payload=json.dumps({'key1': 'value1'}),
)

print("Function output:", response['Payload'].read().decode())

❗ client is required for AWS Lambda.

---

✅ 5. DynamoDB: Add Item, Query Table

python
import boto3

dynamodb = boto3.resource('dynamodb')
table = dynamodb.Table('MyTable')

# Put item
table.put_item(Item={'id': '123', 'name': 'John Doe'})

# Get item
response = table.get_item(Key={'id': '123'})
print(response['Item'])

✅ resource is perfect for table access in DynamoDB.