Eternet Portal Documentation

Cloud Workspace and Compute

Compute route

The current Compute page is a product-selection workspace with IoT bundles, a price calculator, resource rates, quantity selection, and a payment-summary drawer. The displayed plans, rates, and total are currently defined in frontend component data and are not a live quote or confirmed provisioning transaction.

Cloud overview APIs

The shared cloud data layer can load VPS instances, block storage, object storage, virtual networks, and Kubernetes clusters from the Cloud API using the authenticated session. The data loader returns empty lists when the service is unavailable or the request fails; empty results therefore need to be interpreted with connection status in mind.

Provisioning boundary

Do not describe the Compute page’s “Deploy Now” control as completed provisioning: the current workspace opens a local payment-summary drawer and does not execute a verified deployment API flow.

Cloud workspace: what a customer can explore

The Cloud area groups product selection and account resource views, but different routes have different levels of integration. The Compute page currently presents product bundles, a price calculator, resource-rate comparisons, location selection, quantity selection, and a payment-summary drawer. It helps a user compare options and estimate a displayed total based on frontend plan data. That total is illustrative in the current implementation: plan names, rates, and the quantity calculation are component-defined rather than retrieved as a verified customer quote. The “Deploy Now” control currently opens a summary drawer and should not be described as completed provisioning.

Other cloud workspaces retrieve resource lists through a separate Cloud API boundary. The server-side loader uses the signed-in session token to request VPS instances, block storage, object storage, virtual networks, and Kubernetes clusters. This separation means a customer can encounter a populated resource inventory in one area while the product calculator remains a frontend experience. The Cloud pages should therefore be introduced as a combination of resource visibility and product exploration, not as one fully integrated purchasing workflow.

Resource lifecycle and service boundaries

Cloud resources are owned by the Cloud API and underlying infrastructure, not by the portal’s React components. The portal requests authorized data and presents the returned resource records. Any create, resize, delete, deploy, or payment operation must be confirmed in the actual route and API contract before it is represented as available. A visible “available” rate row or plan card is not proof that capacity exists in the selected location. Similarly, an empty list can result from missing authentication, a missing service base URL, a network failure, or a genuinely empty account. The current shared loader converts several failures into empty arrays, so operators should verify connectivity before drawing conclusions from an empty page.

For an accurate customer walkthrough, start with a resource page whose Cloud API connection has been confirmed. Show the returned identifier and status, then explain which resource category it belongs to. If discussing purchase or provisioning, state that the portal’s current Compute calculator is a selection/estimate surface until an end-to-end provisioning endpoint and response have been verified. Keep billing claims separate from Cloud resource listings; billing currently has its own informational settings route.

Customer value and readiness

When connected, the Cloud resource pages can give customers a familiar place to see their infrastructure alongside IoT devices, analytics, and operational services. The combined navigation can help teams move between cloud resources and the workloads that depend on them. Before a production presentation, verify the Cloud API endpoint, authorization scope, response mapping, pagination behavior, and error visibility for each resource type. Validate that row labels represent actual API properties and that any displayed metric has a source and timestamp. The current implementation provides useful entry points, but production purchasing and deployment claims require additional backend confirmation.