> ## Documentation Index
> Fetch the complete documentation index at: https://docs.getgalaxy.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Amazon S3

> Connecting Amazon S3 buckets to Galaxy

Connect Galaxy to an Amazon S3 bucket to observe objects, prefixes (folders), and organizational hierarchy.

<Frame>
  <img src="https://mintcdn.com/galaxy-d7fe996b/6-Rz1rpIBcXfVZgy/images/s3.jpg?fit=max&auto=format&n=6-Rz1rpIBcXfVZgy&q=85&s=39965044aefc09b4d6e7a4a2f22b99d4" alt="Amazon S3 source configuration" width="1200" height="630" data-path="images/s3.jpg" />
</Frame>

## How to Connect

1. **Select Amazon S3** as the Source type
2. **Provide AWS credentials**:
   * **Access Key ID**: AWS access key ID
   * **Secret Access Key**: AWS secret access key
   * **Region**: AWS region where your buckets are located
   * **Optional**: Role ARN for IAM role authentication
   * **Optional**: Custom endpoint URL
3. **Configure bucket**: Provide the bucket name to observe
4. **Configure streams**: Set up file-based stream configurations for different object types
5. **Optional start date**: Set a start date for syncing historical data

Once connected, Galaxy validates the connection and begins observing the bucket structure.

## Configuration Options

### AWS Credentials

* **Access Key ID**: AWS access key ID
* **Secret Access Key**: AWS secret access key
* **Region**: AWS region where your buckets are located
* **Role ARN**: Optional IAM role ARN for role-based authentication
* **Endpoint**: Optional custom endpoint URL (for S3-compatible services)

Ensure the credentials have appropriate permissions to list and read bucket contents.

### Bucket Configuration

* **Bucket**: S3 bucket name to observe

### Stream Configuration

Configure file-based streams to process different object types:

* **Structured formats**: CSV, JSONL, Parquet, Avro, Excel
* **Unstructured documents**: PDFs, DOCX, Markdown, and other text formats
* **Processing options**: Configure parsing strategies, validation policies, schema discovery, and glob patterns

### Delivery Method

Choose how to deliver data:

* **Replicate Records**: Deliver as structured records
* **Copy Raw Files**: Copy files as-is, optionally preserving directory structure

<Info>
  Don’t see a source type you're looking for? We connect to hundreds of systems - reach out to [support@getgalaxy.io](mailto:support@getgalaxy.io) to request access.
</Info>

## What's Next

<CardGroup cols={2}>
  <Card icon="database" horizontal href="/sources" title="Sources">
    Learn about Sources in general
  </Card>

  <Card icon="diagram-project" horizontal href="/projects" title="Projects">
    Understand how to use Sources in Projects
  </Card>
</CardGroup>