docs/self-hosting/aws.md
Host BentoPDF on AWS for maximum control and scalability.
User → CloudFront (CDN) → S3 (Static Files)
# Create bucket
aws s3 mb s3://your-bentopdf-bucket --region us-east-1
# Enable static website hosting
aws s3 website s3://your-bentopdf-bucket \
--index-document index.html \
--error-document index.html
# Build the project (optionally with custom branding)
# VITE_BRAND_NAME="AcmePDF" VITE_BRAND_LOGO="images/acme-logo.svg" npm run build
npm run build
# Sync to S3
aws s3 sync dist/ s3://your-bentopdf-bucket \
--delete \
--cache-control "max-age=31536000"
# Set correct MIME types for WASM
aws s3 cp s3://your-bentopdf-bucket/ s3://your-bentopdf-bucket/ \
--recursive \
--exclude "*" \
--include "*.wasm" \
--content-type "application/wasm" \
--metadata-directive REPLACE
aws cloudfront create-distribution \
--origin-domain-name your-bentopdf-bucket.s3.amazonaws.com \
--default-root-object index.html
Or use the AWS Console:
index.htmlLibreOffice-based conversions (Word, Excel, PowerPoint to PDF) require SharedArrayBuffer, which needs specific response headers. Create a CloudFront Response Headers Policy:
| Header | Value |
|---|---|
Cross-Origin-Embedder-Policy | require-corp |
Cross-Origin-Opener-Policy | same-origin |
Cross-Origin-Resource-Policy | cross-origin |
Or via CLI:
aws cloudfront create-response-headers-policy \
--response-headers-policy-config '{
"Name": "BentoPDF-COEP-COOP",
"CustomHeadersConfig": {
"Quantity": 3,
"Items": [
{"Header": "Cross-Origin-Embedder-Policy", "Value": "require-corp", "Override": true},
{"Header": "Cross-Origin-Opener-Policy", "Value": "same-origin", "Override": true},
{"Header": "Cross-Origin-Resource-Policy", "Value": "cross-origin", "Override": true}
]
}
}'
The LibreOffice WASM files are pre-compressed (.wasm.gz, .data.gz). Set the correct Content-Type and Content-Encoding so browsers decompress them:
# Set correct headers for soffice.wasm.gz
aws s3 cp s3://your-bentopdf-bucket/libreoffice-wasm/soffice.wasm.gz \
s3://your-bentopdf-bucket/libreoffice-wasm/soffice.wasm.gz \
--content-type "application/wasm" \
--content-encoding "gzip" \
--metadata-directive REPLACE
# Set correct headers for soffice.data.gz
aws s3 cp s3://your-bentopdf-bucket/libreoffice-wasm/soffice.data.gz \
s3://your-bentopdf-bucket/libreoffice-wasm/soffice.data.gz \
--content-type "application/octet-stream" \
--content-encoding "gzip" \
--metadata-directive REPLACE
::: warning Important
Without the response headers policy, SharedArrayBuffer is unavailable and LibreOffice WASM conversions will hang at ~55%. Without the correct Content-Encoding on the .gz files, the browser receives raw gzip bytes and WASM compilation fails.
:::
Allow CloudFront to access the bucket:
{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowCloudFrontAccess",
"Effect": "Allow",
"Principal": {
"Service": "cloudfront.amazonaws.com"
},
"Action": "s3:GetObject",
"Resource": "arn:aws:s3:::your-bentopdf-bucket/*",
"Condition": {
"StringEquals": {
"AWS:SourceArn": "arn:aws:cloudfront::ACCOUNT_ID:distribution/DISTRIBUTION_ID"
}
}
}
]
}
Configure 404 to return index.html for SPA routing:
/index.html| Resource | Estimated Cost |
|---|---|
| S3 Storage (~500MB) | ~$0.01/month |
| CloudFront (1TB transfer) | ~$85/month |
| CloudFront (10GB transfer) | ~$0.85/month |
::: tip Use S3 Intelligent Tiering for cost optimization on infrequently accessed files. :::
# main.tf
resource "aws_s3_bucket" "bentopdf" {
bucket = "your-bentopdf-bucket"
}
resource "aws_cloudfront_distribution" "bentopdf" {
origin {
domain_name = aws_s3_bucket.bentopdf.bucket_regional_domain_name
origin_id = "S3Origin"
}
enabled = true
default_root_object = "index.html"
default_cache_behavior {
allowed_methods = ["GET", "HEAD"]
cached_methods = ["GET", "HEAD"]
target_origin_id = "S3Origin"
viewer_protocol_policy = "redirect-to-https"
}
}