Remote Execution

Xenon's remote execution is powered by a high-performance gRPC/NATS event bus, transforming isolated servers into a unified, reactive mesh.

The Hub acts as the strategic intelligence layer, orchestrating device allocation and routing commands across the global grid.

Unified Registry: Real-time visibility into all remote nodes.
Smart Load Balancing: Allocates devices based on node health and latency.
Failover Recovery: Automatic session re-routing if a node becomes unresponsive.

A remote machine hosting physical devices or simulators, running an Appium server with the Xenon plugin active.

Auto-Registration: Nodes use zero-config discovery to join the grid.
High-Fidelity Streaming: Direct NATS streams for ultra-low latency video and logs.

For global-scale deployments, Xenon supports a Cellular Architecture where multiple Hubs share a common state via PostgreSQL.

Cell Isolation: Group infrastructure into regional "cells" (e.g., US-West, EU-Central) to minimize device-to-hub latency.
State Persistence: All session history and device lockers are persisted in PostgreSQL, allowing Hubs to be completely stateless and easily scalable.
Disaster Recovery: If a regional cell fails, the global registry allows workers to be immediately re-provisioned to another cell.

To enable this, configure the databaseProvider and databaseUrl in your xenon.config.json.

Navigate to the host and port of Hub server from the above example it will be http://localhost:31137/xenon
Dashboard should have device list based on the hub configuration.