Skip to main content
A concentrator is a “thing of things”: one XMPP identity representing many physical or virtual nodes organized into data-source trees.

​
Discovery operations

Clients can:
  • query capabilities;
  • list all, root, or child data sources;
  • test whether a node exists;
  • get one or many nodes;
  • list root/child/ancestor nodes and inheritance;
  • localize node parameters and messages.
Use paged or hierarchical traversal. getAllNodes can be inappropriate for a large industrial deployment.

​
Node management

When supported, the interface exposes editable parameters, common parameters across nodes, addable node types, new-node creation, deletion, and move up/down operations. Before editing, request the form/schema for the exact node type. Optimistic updates should include the version or timestamp returned by the concentrator where available.

​
Commands and queries

Nodes can expose commands. A command can be immediate, parameterized, or a long-running query with progress and abort operations. Treat query IDs as scoped to the concentrator and authenticated account.

​
Events

Subscribe to structural events:
Reconcile from the server after reconnect because an event stream can have gaps.

​
Troubleshooting sniffers

The concentrator interface can register a remote sniffer for a node where permitted. Captures can contain device data and credentials; require an operator privilege, limit duration, and record who enabled it.

​
Stable client cache

Key cached nodes by full extended address. Store display name, node type, capabilities, status, and last update separately. A move changes tree position, not necessarily node identity.