Eternet Portal Documentation
Service Operations
Current controls
The page requests service records through a portal API route and offers a start/stop action based on the reported container state. It updates the displayed status after a successful action response.
Metrics caveat
CPU and memory values are read from the service response when present. Disk and traffic values, capacity labels, and the “Last Update” display are currently hard-coded presentation values in the page and must not be reported as measured live metrics.
Operational caution
Starting or stopping a service changes runtime state. Confirm the target service and authorization; verify the resulting container state from the service API or infrastructure rather than relying only on an optimistic UI update.
Service inventory and control workflow
The Services page requests a list of service/container records from a portal API route and presents the name, runtime state, and any CPU or memory values returned in the record. For a service that has an identifier, the page offers a start or stop action based on its reported status. The request is sent through the portal server boundary so infrastructure credentials do not need to be placed in browser code. This gives operators a convenient entry point for basic runtime control without requiring them to navigate directly to a container host.
Container start/stop actions can interrupt dependencies or change the availability of user-facing services. Administrators should confirm the service identity and the operational impact before acting. After sending the request, the UI updates its local state when the response succeeds. That immediate status change should still be checked against an authoritative runtime state, because a control API may accept a request before the process has fully started or stopped. A robust runbook verifies health checks, dependent services, and application availability after the action.
Metrics and presentation caveats
CPU and memory values may come from the service response through the HostConfig fields. They should be shown with units, collection time, and a clear source. In the current page, disk and traffic values, capacity labels, and the Last Update label are hard-coded presentation values rather than measured telemetry. Do not tell a customer that these figures represent their actual VPS utilization. The displayed “Running” or “Stopped” status is based on a state field and can be stale until the list is refreshed. It does not necessarily mean that the application inside a container passes its health check.
Service list failures are surfaced as an error, while an empty result is rendered as no services found. Operators should distinguish a genuinely empty account from a missing service connection or insufficient permission. For production use, the backend should return explicit error and timestamp details, and the page should refresh after state changes to confirm the final state.
Customer value and safe operation
When backed by a properly scoped service API, this page can shorten incident response by making service state and basic controls available within the portal. It complements, rather than replaces, infrastructure observability and orchestration tooling. A safe customer demo should use a non-critical service in a test environment, show a read-only status first, and avoid stopping shared production dependencies. Present current metrics only when their source is verified. The portal provides an authenticated control surface; container lifecycle, resource telemetry, and application health remain owned by the infrastructure and backend services.