Object storage

Loopback provisions S3-compatible object storage scoped to a project. Each bucket is a first-class object store in the product: it carries provider, region, usage summaries, and credential references. Automation creates the remote bucket and returns endpoint details you use with any AWS S3 SDK-compatible client.

Scope and permissions

Organization owns billing and tenancy; project owns the bucket list.
RBAC is enforced at organization boundary with project context for bucket operations.
Older API v2 list behavior may hide superseded bucket rows for backward compatibility; API v3 lists the full set for the project. Prefer v3 for new integrations unless your operator standardizes otherwise.

Providers

Supported provider families depend on how your operator wired service providers and regions. Commonly referenced families in deployments include:

IONOS DCD - IONOS S3-compatible buckets (regions such as Frankfurt appear in product pickers).
Hetzner Cloud - Hetzner Object Storage.
Partner / dedicated stacks - when the operator runs a non-public object backend.

Exact availability is operator configuration: service providers and compute providers may be pre-linked so bucket creation can resolve credentials without exposing them in API responses.

Lifecycle (conceptual)

Validate name and configuration before create.
Create records the bucket intent and enqueues provider automation (S3-compatible SDKs under the hood).
Credentials are stored by reference; API responses mask raw secrets.
Usage figures (object count, bytes, last sync time) are refreshed by background sync where implemented.

Regions and catalog

v3 exposes a regions helper that aggregates provider-specific region lists for pickers. Exact contents are operator configuration.

Relationship to Kubernetes