Amazon MemoryDB for Redis Hands-On

Demo

MemoryDB = Primary Database (durable, persistent storage) ElastiCache = Cache Layer (fast, temporary storage)

Fundamental Difference

In this demo, we will:

  1. Set up networking infrastructure (VPC and security groups)
  2. Create a MemoryDB subnet group
  3. Deploy a MemoryDB cluster with Multi-AZ configuration
  4. Configure parameter groups for optimal performance
  5. Connect to the cluster using EC2 instance
  6. Interact with MemoryDB using CLI and Python Program
  7. Clean up resources

Agenda

Amazon MemoryDB Dashboard

memorydb-demo

Create VPC

Number of Availability Zones

Create VPC

View VPC

Security Groups

Create Subnet Group

memorydb-demo-subnet-group
Subnet group for MemoryDB demo cluster

Create subnet group

Selected subnets

The subnet group was created successfully

Create MemoryDB Parameter Group

memorydb-demo-params

Create parameter group

Custom parameters for demo cluster

Edit parameters

maxmemory-policy

Edit parameters

Create cluster

Cluster settings

memorydb-demo-cluster

Cluster info

Demo cluster for session storage

Connectivity

Vector search

View default settings

Create

Create SSM Role for EC2

Select trusted entity

Add permissions

Name, review, and create

SSMInstanceRole

Step 1: Select trusted entities

Step 2: Add permissions

Create EC2 Machine

Launch EC2 Instance for Testing

memorydb-client

Amazon Machine Image (AMI)

Instance type

Key pair (login)

Network settings

Configure storage

IAM instance profile

Session Manager

Connect and Test MemoryDB Cluster

# Update package cache
sudo dnf update -y 

# Install Redis directly
sudo dnf install -y redis6 

# Verify installation
redis6-cli --version

Install Redis CLI

ENDPOINT=
redis6-cli -h ${ENDPOINT} -p 6379 --tls
# Set a key
SET session:user123 "{'id':'123','name':'John','loginTime':'2024-01-15T10:30:00Z'}"
# Get the key
GET session:user123
# Set key with expiration (30 minutes)
SETEX session:user456 1800 "{'id':'456','name':'Jane'}"
# Check time to live
TTL session:user456

Connect to MemoryDB cluster:

Test basic operations:

Creates a key with automatic expiration after 1800 seconds (30 minutes) for session timeout

# Test data structures
HSET user:1001 name "Alice" email "alice@example.com" lastLogin "2024-01-15"
HGETALL user:1001
# List operations
LPUSH recent:logins "user123" "user456" "user789"
LRANGE recent:logins 0 -1
# Test persistence
SET persistent:data "This will survive a failover"

Uses Redis hash data structure to store multiple user attributes efficiently

Pushes user IDs to a list to track recent login activity in chronological order

Sets a test key to verify data persistence through cluster failover events

sudo dnf install -y python3 python3-pip 
pip3 install redis

Install Python and Redis client

Test Application Integration 

cat > test_memorydb.py << EOF
import redis
import json
import sys
from datetime import datetime

# Check for command line argument
if len(sys.argv) != 2:
    print("Usage: python3 test_memorydb.py <your-cluster-endpoint>")
    print("Example: python3 test_memorydb.py memorydb-demo-cluster.abc123.memorydb.us-east-1.amazonaws.com")
    sys.exit(1)

cluster_endpoint = sys.argv[1]

# Configure connection
try:
    r = redis.Redis(
        host=cluster_endpoint,
        port=6379,
        ssl=True,
        decode_responses=True
    )
    
    # Test connection
    r.ping()
    print(f"✓ Successfully connected to MemoryDB cluster: {cluster_endpoint}")
    
except redis.ConnectionError as e:
    print(f"✗ Failed to connect to {cluster_endpoint}")
    print(f"Error: {e}")
    print("\nPlease check:")
    print("  - Cluster endpoint is correct")
    print("  - Security group allows port 6379 from this instance")
    print("  - Instance and cluster are in the same VPC")
    sys.exit(1)

# Session management example
def create_session(user_id, username):
    session_data = {
        'user_id': user_id,
        'username': username,
        'login_time': datetime.now().isoformat()
    }
    # Store session with 30-minute expiration
    r.setex(f'session:{user_id}', 1800, json.dumps(session_data))
    print(f"✓ Session created for {username}")

def get_session(user_id):
    session = r.get(f'session:{user_id}')
    if session:
        return json.loads(session)
    return None

# Test the functions
print("\n--- Testing Session Management ---")
create_session('user001', 'alice')
session = get_session('user001')
print(f"✓ Retrieved session: {session}")

# Cache example with automatic expiration
print("\n--- Testing Cache with TTL ---")
r.setex('cache:api_response', 300, json.dumps({'data': 'cached response'}))
print(f"✓ Cache created with TTL: {r.ttl('cache:api_response')} seconds")

# Additional tests for data persistence
print("\n--- Testing Data Persistence ---")
r.set('persistent:data', 'This will survive a failover')
print(f"✓ Persistent data stored")

# Test various data structures
print("\n--- Testing Redis Data Structures ---")

# Hash
r.hset('user:1001', mapping={
    'name': 'Alice',
    'email': 'alice@example.com',
    'lastLogin': datetime.now().isoformat()
})
print(f"✓ Hash created: {r.hgetall('user:1001')}")

# List
r.lpush('recent:logins', 'user001', 'user002', 'user003')
print(f"✓ List created: {r.lrange('recent:logins', 0, -1)}")

# Set
r.sadd('active:users', 'alice', 'bob', 'charlie')
print(f"✓ Set created with {r.scard('active:users')} members")

print("\n✓ All tests completed successfully!")
EOF
# Replace with your actual MemoryDB cluster endpoint
python3 test_memorydb.py ${ENDPOINT}

Run the Python Script

Clean Up

Delete MemoryDB Cluster

Delete EC2 Instances

Delete Subnet Group

Delete Parameter Group

Delete VPC

Delete NAT Gateway

🙏

Thanks

for

Watching

Amazon MemoryDB - Hands-On Demo

By Deepak Dubey

Amazon MemoryDB - Hands-On Demo

Amazon MemoryDB - Hands-On Demo

  • 29

More from Deepak Dubey