Storage quick start
This path stores one file in a project bucket and reads it back. You need a Hubfly space account with permission to create a project, plus an S3-compatible client or SDK.
Before you begin
Create a new project and select Storage as its platform. Choose the Storage region where the data should live. Open the project’s Storage workspace and note the displayed S3 endpoint and signing region. They are used by your application; the Dashboard API address is different.
Create and upload
- Create a bucket. Choose a name such as
project-assets. Buckets contain objects; an object key such asreports/hello.txtcan include slashes without creating directories. - Create an access key. Give it a name and
readwritepermission. Choose a bucket scope when only one bucket is needed. Copy the access key ID and secret from the creation result; the secret is shown once. - Configure your client. Set the Hubfly space S3 endpoint, the displayed signing region, and the new key pair. Use HTTPS and AWS Signature Version 4. Enable path-style addressing for SDKs that otherwise construct AWS bucket hostnames.
- Upload a file. PUT
reports/hello.txtinto the bucket through the S3 client. Set a usefulContent-Type, such astext/plain.
aws --endpoint-url "$HUBFLY_STORAGE_ENDPOINT" s3api put-object \
--bucket project-assets \
--key reports/hello.txt \
--body ./hello.txt \
--content-type text/plainSet AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, and AWS_DEFAULT_REGION to the values from the Storage workspace before running the command. Use the exact endpoint shown for your project.
Uploading from a browser?
upload_url. The browser PUTs the file there before expires_at (one hour), sending no headers other than Content-Type; that upload URL accepts cross-origin requests. Do not send browser uploads straight to the S3 endpoint, even with a presigned URL: it rejects the browser's CORS preflight. If the upload URL does not fit your app, upload through your own backend.Verify the object
List the bucket or open the object browser in Dashboard. Download the key and compare its bytes with your original file. The S3 API also supports HEAD when you only need metadata.
aws --endpoint-url "$HUBFLY_STORAGE_ENDPOINT" s3api get-object \
--bucket project-assets \
--key reports/hello.txt \
./downloaded-hello.txtDo it with the CLI
The hubfly CLI covers the same path. Uploads and downloads use temporary URLs issued for your account, so you only need an S3 key for your application:
hubfly storage create media # choose the region, confirm, wait until ready
hubfly storage buckets create project-assets --project media
hubfly storage put ./hello.txt project-assets/reports/ --project media
hubfly storage get project-assets/reports/hello.txt ./downloaded-hello.txt --project media
hubfly storage keys create my-app --bucket project-assets --format dotenv --project media # secret printed onceSee the hubfly storage reference for every command.
What next
- Read buckets and objects for listing, versions, multipart uploads, and deletion.
- Register the object as a managed asset if you need delivery controls or image variants.
- See the S3 reference for exact supported operations and an SDK configuration example.