การเข้าถึงฐานข้อมูลใน VPC

Welcome! In this lesson, you'll learn how to securely connect your Lambda functions to databases that reside within your private Amazon Virtual Private Cloud (VPC).

This is crucial for serverless applications that need to interact with sensitive or internal data stores like Amazon RDS or other databases that aren't publicly accessible.

You've already configured your Lambda function to operate within a VPC (from the previous lesson). Now, we'll focus on how that Lambda can reach a database that's also inside the same VPC.

This setup ensures secure, internal communication, keeping your database isolated from the public internet.

Security Groups are fundamental here. Think of them as virtual firewalls for your instances and resources within a VPC.

• They control both inbound (incoming) and outbound (outgoing) traffic.
• Both your Lambda function and your database will have associated security groups.
• These groups must be configured to explicitly allow communication between them.

Your database's security group needs an inbound rule to allow connections from your Lambda function.

• Type: Custom TCP
• Port Range: The specific port your database listens on (e.g., 3306 for MySQL, 5432 for PostgreSQL).
• Source: Crucially, this should be the Security Group ID of your Lambda function.

This ensures only your Lambda can initiate connections to the database.

Conversely, your Lambda function's security group needs an outbound rule to allow it to send traffic to your database.

• Type: Custom TCP
• Port Range: Again, the database's specific port (e.g., 3306).
• Destination: This should be the Security Group ID of your database.

This allows your Lambda to successfully reach out and connect to the database.

Beyond security groups, ensure your Lambda function is deployed into subnets that have proper network routing access to the subnets where your database instances reside.

• Typically, both Lambda and the database will be placed in private subnets within the same VPC.
• AWS handles the internal routing within the VPC, but proper subnet association is key for reachability.

Inside your Lambda function's code, you'll use standard database drivers and connection strings, just like any other application.

You'll need these key details:

• Database endpoint: (e.g., your-db.xxxx.rds.amazonaws.com)
• Port: (e.g., 3306)
• Database name
• Username
• Password

Here's a conceptual Java snippet demonstrating how a Lambda might attempt to connect to a database. In a real application, you would handle credentials securely.

This example shows the basic structure; it won't run successfully without a live database and environment variables set.

Never hardcode database credentials directly in your Lambda code! This is a major security risk.

Best practices for managing sensitive information:

• AWS Secrets Manager: The recommended way to store, retrieve, and rotate database credentials securely.
• Environment Variables: Suitable for non-sensitive configuration, but not for passwords or API keys.

Always prioritize security for sensitive data.

Imagine your Lambda function (associated with Security Group SG-A) needs to connect to an RDS database (associated with Security Group SG-B) on port 3306.

Which inbound rule must be configured on SG-B (the database's security group) to allow this connection?

Great job! You've learned the essentials of connecting your Lambda functions to private databases within your VPC.

• Both your Lambda and database must be in the same VPC.
• Security groups are vital for controlling traffic flow between them.
• Configure inbound rules on the database's SG and outbound rules on Lambda's SG.
• Always manage database credentials securely, ideally using AWS Secrets Manager.

This knowledge is key to building secure and robust serverless applications!