Skip to content

Capability Profiles

Capability profiles describe what each enrolled target contributes to VeeaONE infrastructure.

A VeeaHub, Linux / Ubuntu system, NVIDIA Jetson device, and specialized hardware node may all run containerized applications, but they add different value to the managed edge environment. Capability profiles make that value clear so developers can deploy each workload to the right place.

Capability Profile

Use a capability profile to describe the target in operational terms:

Capability What It Means
Identity Runtime ID, serial number, displayed model, site, mesh, and ownership.
Compute CPU, memory, architecture, and expected workload class.
Storage Persistent app data, image storage, logs, backups, and available capacity.
Networking IP addressing, DNS, routing, firewall posture, local network access, and cloud reachability.
App lifecycle Install, start, stop, restart, redeploy, upgrade, rollback, and uninstall behavior.
Operations VeeaONE visibility, logs, health, events, alerts, and support paths.
Device access Serial, USB, camera, sensor, radio, industrial bus, or local hardware access.
Acceleration GPU, NPU, TPU, FPGA, or other accelerator access from containers.
Security Credentials, certificates, registry access, permissions, and privilege boundaries.

Match Workloads to Targets

Workload Need Capability to Look For
Customer web app or API App lifecycle, networking, persistent storage, operations visibility.
Local database or file workflow Storage capacity, backup behavior, disk performance, lifecycle behavior.
AI inference or video analytics Accelerator access, camera or stream access, compute, thermal profile.
Industrial integration Device I/O, local network reachability, permissions, ruggedized placement.
Site-adjacent aggregation Network placement, storage, compute, cloud connectivity, operational visibility.

Operational Readiness

Before placing a workload on a target, confirm the capability profile covers:

  1. runtime identity and enrollment
  2. VeeaONE operations visibility
  3. container runtime behavior
  4. storage persistence
  5. network reachability
  6. application lifecycle
  7. workload-specific hardware access
  8. support and ownership notes

Language Guide

Use precise, positive support language:

  • Use native VeeaHub target for Veea hardware with the full managed runtime profile.
  • Use qualified edge node when a non-hub target has a defined runtime profile.
  • Use runtime target when describing any node that can run VeeaONE Runtime.
  • Use capability profile when describing what a target contributes to the edge environment.

Avoid broad claims such as "runs anywhere" or "any hardware." The stronger story is that VeeaONE infrastructure can grow across the right targets for the right workloads.