File storage
Object storage for files that outlive a deployment, scoped to one environment and served over the CDN.
File storage is object storage for files that have to survive the next deployment: user uploads, generated invoices, exports. Objects live in a bucket, and every bucket belongs to exactly one environment of one Gigadrive Network application.
Environment scoping
Because the environment owns the bucket, production and a preview environment never share one. A preview deployment writes to its own copy, so a test upload cannot land next to real customer files. Deployment credentials carry their environment with them, which is why code running on Network can upload without naming an environment at all.
Buckets are private by default. A private object is reachable only through a signed URL that expires; a public bucket serves every object from a stable CDN hostname. Deleting an object is reversible for 30 days.
Create a bucket, name it, and choose its visibility.
Uploading objectsUpload sessions, the tus endpoint, resuming an interrupted transfer, and object keys.
Client-side uploadsSend a file from the browser without the bytes crossing your server.
Access URLsPublic CDN URLs and signed URLs with an expiry.
S3-compatible APIPoint the AWS CLI, an AWS SDK, boto3 or rclone at your buckets.
S3 credentialsMint, scope and revoke the access keys the S3 endpoint accepts.
BackupsSnapshot a bucket manually or on a schedule, and restore from one.
Trash and restoreRecover a deleted object, or remove it for good.
