Skip to content
qibdo qibdo
Theme
Book a demo

Overview

A distributed object and block storage service for unstructured data, orchestrated across clouds. Storage consolidates your on-premise and public-cloud storage into a single console, so you can provision, access, and repatriate data anywhere with one consistent model, all in one.

Storage is organised around workspace-scoped buckets that hold objects. The management calls that need authorization and an audit trail, creating buckets, defining storage classes, and issuing presigned URLs, run through qibdo’s control-plane API and are gated by IAM; object bytes flow straight to the storage engine over its data plane or through time-limited presigned URLs.

  • Buckets: a workspace-scoped container for objects. A bucket is pinned to a region, can have versioning enabled, and names a default storage class for the objects it holds. It is the parent scope for objects, multipart uploads, and presigned URLs.
  • Objects: identified by their bucket and an object key. Object metadata (size, content type, etag, version id, and so on) is read straight from the storage engine rather than persisted in the control plane.
  • Multipart uploads: a workflow for large objects, with steps to initiate the upload, fetch a presigned URL per part, then complete or abort it.
  • Presigned URLs: time-limited PUT and GET URLs for direct data-plane transfer, valid for one hour by default and up to twenty-four hours.
  • Storage classes: the storage tiers available in a region, defined by an administrator. Each maps to a backing data pool with its own compression and redundancy characteristics, and one per region is marked the default.
  • Operations: every mutating call returns an operation record that captures the request and comes back already completed.