Downloads - Tapedrive

Standard S3 read operations, served from decoded and verified content. Reads can be anonymous, subject to the gateway’s metering (rate limits), or signed.

GetObject and HeadObject

Fetch an object or its metadata. Responses carry Content-Type, Content-Length, a quoted ETag (the on-chain commitment), Last-Modified, and Cache-Control. Single-track GETs also advertise Accept-Ranges.

aws CLI
rclone
Presigned URL

aws s3 cp s3://my-bucket/reports/q2.pdf ./q2.pdf --profile tapedrive

rclone copy tapedrive:my-bucket/reports/q2.pdf ./

url=$(aws s3 presign s3://my-bucket/reports/q2.pdf --profile tapedrive --expires-in 3600)
curl -o q2.pdf "$url"

Range requests

A Range header returns 206 Partial Content with the requested bytes. On streamed (multi-segment) objects the gateway resolves the manifest and decodes only the segments your range covers (how streams are laid out). An unsatisfiable range returns 416 (InvalidRange).

curl -H "Range: bytes=0-1048575" -o first-mb.bin "$url"

Listing

ListObjectsV2 supports prefix and delimiter for directory-style browsing, pagination with continuation tokens, and up to 1000 keys per page. The V1 fallback works for older tooling. Listings come from the gateway’s replay-built catalog: a listing page is one range scan, so browsing stays fast at any bucket size (objects).

aws CLI
rclone
boto3

aws s3 ls s3://my-bucket/reports/ --profile tapedrive

rclone lsl tapedrive:my-bucket/reports/

pages = s3.get_paginator("list_objects_v2").paginate(
    Bucket="my-bucket", Prefix="reports/"
)