파일 저장 및 이벤트를 위한 S3

Welcome! In this lesson, we'll dive into how Amazon S3, a powerful storage service, works hand-in-hand with AWS Lambda.

You'll learn to store and retrieve files, and critically, how S3 can trigger your Lambda functions automatically when something happens to your data.

Amazon S3 (Simple Storage Service) is a highly scalable, durable, and available object storage service.

• It stores data as objects within buckets.
• Objects are files (images, videos, documents) and their metadata.
• Buckets are like top-level folders, globally unique names.
• S3 is perfect for storing large amounts of unstructured data.

You can easily put objects into an S3 bucket. Each object has a key, which is its unique identifier within the bucket.

• Console: Upload files directly via the AWS Management Console.
• CLI: Use the AWS Command Line Interface for scripted uploads.
• SDKs: Programmatically upload from your application using AWS SDKs (e.g., Python Boto3, Node.js SDK).

Getting your data back from S3 is just as simple. You reference the bucket name and the object's key.

S3 also offers features like versioning to keep multiple versions of an object, and pre-signed URLs for temporary, secure access to private objects.

One of S3's most powerful features for serverless is Event Notifications. S3 can notify you when specific events occur in your buckets.

These events include:

• Object creation (PUT, POST, COPY)
• Object deletion
• Object restoration from Glacier

These notifications can be sent to various AWS services, including Lambda functions!

Setting up an S3 trigger for Lambda is done via the S3 console or infrastructure-as-code tools.

You specify:

• The event type (e.g., ObjectCreated:Put).
• The destination (your Lambda function).
• Optional prefix or suffix to filter events (e.g., only trigger for files in the images/ folder or .jpg files).

When an S3 event occurs, Lambda invokes your function and passes the event details as a JSON payload to your handler. Let's see an example in Python.

The event object passed to your Lambda function's handler contains crucial information about the S3 event.

• Records: A list, as multiple events can be batched.
• s3: Contains details about the bucket and object involved.
• bucket.name: The name of the S3 bucket.
• object.key: The full path/key of the object that triggered the event.
• eventName: Describes the specific S3 action (e.g., ObjectCreated:Put).

A classic serverless pattern is image processing. Imagine a mobile app where users upload photos:

1. User uploads a high-res image to an S3 bucket.
2. S3 triggers a Lambda function (e.g., ObjectCreated:Put event).
3. Lambda downloads the image, resizes it to various thumbnails.
4. Lambda uploads the thumbnails to a different S3 bucket.

This is highly scalable and cost-effective!

Which of the following statements are TRUE about S3 Event Notifications for Lambda?

Great job! You've learned how S3 acts as both a storage solution and a powerful event source for Lambda.

• S3 stores objects in buckets, offering high durability and scalability.
• S3 Event Notifications allow specific actions (like object creation/deletion) to trigger Lambda functions.
• Lambda functions receive a JSON payload with event metadata, not the full object.
• This pattern enables powerful serverless workflows like image processing or data transformations.