Proxus treats Edge Gateways as "cattle, not pets." In plain terms, that means a gateway should be easy to replace instead of something you manually nurse back to health for days. They are designed to be stateless and secure by default, which keeps large deployments simpler to operate.
The Stateless Philosophy
Traditional SCADA Problem
When a site PC fails in a traditional system, the recovery work often looks like this:
- Travel to the remote site
- Reinstall the operating system
- Restore a backup, if one exists
- Reconfigure device drivers
- Test everything again
Downtime: Hours to days.
Proxus Solution
In Proxus, Gateways hold no state. A new unit can take over without rebuilding local configuration:
- Boot: Gateway starts and connects to Central Core using its authentication token
- Fetch: Downloads the latest Deployment Manifest (devices, rules, scripts)
- Execute: Begins polling devices and processing logic immediately
If hardware fails, swap it with a new unit, apply the token, and it is operational in minutes.
Deployment Manifest
When a Gateway connects, it receives a Deployment Manifest. Think of it as the current job list for that gateway:
| Component | Description |
|---|---|
| Device Profiles | Which PLCs and Protocols to read, and how often |
| Rules | The condition-action logic that runs locally |
| Functions | Custom C# code used for transformations or calculations |
| Integrations | Where processed data should be sent next |
The manifest is version-controlled, so you can roll back to a previous version quickly.
Security Model
The Edge-to-Core connection is designed for Zero Trust environments. Gateways connect outward and do not require inbound firewall openings.
Outbound Only
Gateways initiate the connection. No inbound firewall ports required.
Token Authentication
Every Gateway is authenticated via a unique, revocable token.
Sandboxed Functions
User C# scripts run in isolated contexts, so one bad script does not take down the gateway.
Encrypted Transport
All data in motion uses TLS 1.3 encryption.
Store & Forward
Industrial networks are not always stable. Proxus Gateways keep collecting data locally and send it later if the connection drops.
How It Works
Fieldbus Actor
PLC Communication
Dispatcher
Data Routing
Event Log
Local Buffer
Network Link
May Fail
Central Core
Cloud/On-Prem
States
| State | Behavior |
|---|---|
| Online | Data goes straight to the Core |
| Offline | Data is written to a local disk buffer with a configurable size |
| Reconnected | The buffer is drained in FIFO order, preserving timestamps and sequence |
Even during extended outages, no data is lost. You can size the buffer based on disk capacity and the outage window you want to cover.
Gateway Lifecycle
- Provisioning: Create the token in the Management Console
- Deployment: Install the Gateway agent and configure the token
- Activation: The Gateway connects and downloads its manifest
- Operation: It continuously collects and processes data
- Update: A new manifest is applied without a restart
- Replacement: Swap the hardware and apply the same token again
Next Steps
Hub & Spoke Topology
Understand how the Central Core and Gateways work together.
Connect a Gateway
A step-by-step guide to deploying your first Gateway.
Configure the Edge Gateway mode, name, and health metrics interval via the Edge Gateway section in Proxus-config.toml.