Skip to Content
ArchitectureVirtual Devices

Virtual Devices

A virtual device is a cloud-hosted compute node that participates in your IronFlock project exactly like a physical edge device. It runs the same IronFlock agent, connects to the same message broker, and can run any containerized application.

Virtual devices bridge the gap between edge and cloud. They give you persistent, centrally hosted compute power within the same project context as your physical devices — without setting up separate infrastructure.

How Virtual Devices Work

When you provision a virtual device, IronFlock creates a virtual machine in the same cloud infrastructure that hosts the central services. This VM:

  1. Runs the IronFlock agent, just like a physical device
  2. Joins your project’s messaging realm on the WAMP broker
  3. Appears in your device list alongside physical devices
  4. Can install and run any app from the App Store or your own repository
  5. Receives OTA updates through the same pipeline as physical devices

From the perspective of your apps and your AI agents, a virtual device is indistinguishable from a physical one. An app running on a physical edge device can communicate with an app running on a virtual device through the message broker — as if they were on the same local network.

Common Use Cases

Fleet-Wide Services

Run services that need a persistent, always-on host with a fleet-wide perspective:

  • Grafana — Centralized dashboarding connected to FleetDB
  • Node-RED — Visual flow programming that orchestrates data from multiple devices
  • Jupyter — Interactive data science notebooks with direct access to fleet telemetry
  • Netdata — Infrastructure monitoring across your project

Data Processing Pipelines

Host applications that aggregate, transform, or analyze data from multiple physical devices:

  • Collect telemetry from 100 sensors and run a central quality model
  • Aggregate production metrics from multiple lines into a unified report
  • Run batch processing jobs on historical data stored in FleetDB

AI & Compute-Intensive Workloads

Run workloads that require more compute power than edge hardware provides:

  • Large AI inference models that need GPU resources
  • Training pipelines that process data from the entire fleet
  • Computer vision models that analyze camera feeds from multiple devices

Protocol Bridges & Integrations

Connect external enterprise systems to your IronFlock project:

  • SAP or ERP integration services
  • Cloud API connectors (AWS, Azure, Google Cloud)
  • Custom REST endpoints that expose fleet data to third-party systems
  • MQTT brokers that bridge external data sources into your WAMP messaging realm

Virtual Devices vs. Central Services

Virtual devices are not central services — they are project-scoped compute nodes that you configure and manage:

AspectCentral ServicesVirtual Devices
Managed byIronFlock (or your ops team on-prem)You, through the IronFlock UI
PurposeFleet-wide data, UI, AI, coordinationCustom compute for your specific use case
AppsFixed system servicesAny app from the store or your own
ScopeSystem-widePer-project
ScalingHandled by IronFlockProvision as many as you need

Think of central services as the operating system of IronFlock, and virtual devices as additional servers you spin up for your own applications.

Provisioning

Virtual devices are provisioned from the IronFlock web UI:

  1. Navigate to your project’s device list
  2. Click Add Virtual Device
  3. Choose a compute tier (CPU, memory, storage)
  4. The virtual device boots, connects, and appears in your device list

From that point, you install apps, configure dashboards, and manage it the same way as any physical device.

Last updated on