Multi-Instance Operations¶
L2Proxy does not require every protected path to share one service process and one policy set. Service Management runs multiple independent instances, allowing protection to follow industrial ownership, criticality, and access purpose.
What is independent per service¶
| Service property | Operational advantage |
|---|---|
| Service Profile | Different operating purpose and placement |
| Policy Profile | Different equipment and operation permissions |
| Running state | Start, stop, or restart one service independently |
| Service supervision | A fault or restart is localized to the affected instance |
| Local log | Focused troubleshooting for one protected path |
| Event handling | Appropriate recording and archive behavior per service |
| Inspection mode | Inline, passive, or offline operation according to risk and task |
Isolation patterns¶
One VPN user per enforcement service¶
Figure — Multi-instance operations isolate standalone services while hybrid designs combine network and session paths.
Each user receives an isolated access path, a user-specific Policy Profile, and a dedicated instance. A transformer vendor and a protection engineer can therefore have different permissions even when working at the same site.
One plant area per service¶
Production cells, feeders, substations, or package units can be separated into services with different owners and change windows.
One critical function per service¶
High-impact functions such as DER intertie operation, load shedding, or transformer tap control can be isolated from general monitoring policies.
One commissioning activity per service¶
A passive commissioning profile can operate independently while established production services continue inline enforcement.
Service inventory for operations¶
The service inventory presents each profile with its current operating state and summary. Operators can identify a stopped or running service, inspect its activity duration, open its log, and apply the required lifecycle action without using raw process-management commands.
Failure and maintenance containment¶
Independent instances reduce shared operational impact:
- restarting a vendor service does not require restarting a feeder service;
- changing a commissioning profile does not alter a production Policy Profile;
- troubleshooting evidence remains associated with the affected path;
- maintenance and rollback can be planned per service owner;
- a narrow policy can be deployed without expanding a site-wide shared rule set.
Independence does not remove the need for redundancy, capacity planning, or customer change procedures. It provides a cleaner operational boundary for applying them.
Central evidence, local accountability¶
Although instances are independent, their inspection and policy-decision events can be stored in the central event archive. This supports fleet-wide investigation and reporting while retaining the service, policy, user, equipment, and operation context needed to understand each event.
Recommended operating record¶
For every independent service, retain:
- industrial purpose and owner;
- assigned user, plant area, or activity;
- selected Policy Profile and approved revision;
- operating posture and protected path;
- activation and change references;
- validation evidence;
- event-archive destination;
- rollback and recovery procedure.