A DigitalOcean Function that backs up the complete contents of a DigitalOcean Spaces bucket by archiving all files into a ZIP and uploading it to a destination bucket.

• Downloads all objects from a source Spaces bucket
• Creates a compressed ZIP archive
• Uploads the archive to a destination bucket
• Uses streaming for memory-efficient processing
• Handles pagination for large buckets
• Configurable via environment variables
• Supports cross-region backups

• DigitalOcean Account
• DigitalOcean CLI (doctl)
• Two DigitalOcean Spaces buckets (source and destination)
• Spaces API credentials (Access Key and Secret)

# macOS
brew install doctl

# Authenticate
doctl auth init

doctl serverless connect

Copy the example environment file and fill in your values:

cp .env.example .env

Edit .env with your configuration

npm install

Deploy the function to DigitalOcean:

doctl serverless deploy .

The function will be available at a URL like:

https://faas-nyc1-x.doserverless.co/api/v1/web/fn-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx/backup/backup

You can invoke the function via HTTP request or using the CLI:

# Using doctl
doctl serverless functions invoke backup/backup

# Using curl (replace with your function URL)
curl -X POST https://your-function-url/backup/backup

To run backups automatically, you can:

1. Use DigitalOcean Functions Triggers (when available)
2. Use a cron job with doctl:

   # Add to crontab to run daily at 2 AM
   0 2 * * * /usr/local/bin/doctl serverless functions invoke backup/backup

3. Use a third-party service like cron-job.org to hit the function URL

Successful backup:

{
  "statusCode": 200,
  "body": {
    "message": "Backup completed successfully",
    "sourceBucket": "my-source-bucket",
    "destinationBucket": "my-backup-bucket",
    "archiveName": "backups/backup-my-source-bucket-2025-12-19T10-30-00-000Z.zip",
    "filesBackedUp": 150,
    "archiveSize": 104857600,
    "durationSeconds": 45.32
  }
}

Error response:

{
  "statusCode": 500,
  "body": {
    "error": "Backup failed",
    "message": "Error description",
    "stack": "..."
  }
}

Configured in project.yml:

• Timeout: 15 minutes (900,000 ms)
• Memory: 1GB RAM

Adjust these if needed for larger buckets.

1. List Objects: The function retrieves a complete list of all objects in the source bucket using @aws-sdk/client-s3, handling pagination automatically
2. Create Archive: Uses the archiver library (v7) to create a streaming ZIP archive
3. Stream Download: Each object is streamed from the source bucket using the AWS SDK v3 GetObjectCommand
4. Stream Upload: The archive is simultaneously streamed to the destination bucket using @aws-sdk/lib-storage for efficient multipart uploads
5. Complete: Returns a summary with file count, size, and duration

• Size: Limited by function memory (1GB) and timeout (15 minutes)
• Large Buckets: For very large buckets (100GB+), consider:
  • Increasing function memory and timeout in project.yml
  • Splitting the backup into multiple archives by prefix
  • Using a DigitalOcean Droplet instead
• Bandwidth: Subject to DigitalOcean bandwidth limits

The function uses the following key libraries:

• @aws-sdk/client-s3 (v3) - S3-compatible operations for DigitalOcean Spaces
• @aws-sdk/lib-storage (v3) - Managed multipart uploads with streaming support
• archiver (v7) - Streaming ZIP archive creation

doctl serverless activations list
doctl serverless activations get <activation-id> --logs

MIT

• DigitalOcean Functions Documentation
• DigitalOcean Spaces Documentation
• AWS SDK for JavaScript v3 - Client S3
• AWS SDK for JavaScript v3 - Lib Storage
• Archiver Documentation