Summary
IoT gateway deployments require automatic failover mechanisms protecting against data loss during system reboots, updates, or malfunction. NEXCOM's bypass solution implements a software watchdog triggering hardware relay switching to isolate or reroute traffic when the gateway becomes unresponsive. Unique circuit design protects power state during sudden loss, supporting four operational modes (direct, bypass, block, disconnect) across Linux kernel versions 2.6 through 5.x. Integration with NEXCOM's Atlas OS enables RESTful API control for automated failover orchestration.
Problem / Requirements
IoT gateways aggregate sensor data from distributed devices and route information to cloud infrastructure or on-premises analytics platforms. Planned maintenance windows (firmware updates, kernel patches) and unplanned failures (software crashes, power anomalies) can disrupt data collection indefinitely. Organizations require:
- Automatic failover mechanisms activating without manual intervention
- Data preservation during gateway outages and recovery sequences
- Protection against partial failure scenarios (software lockup but power still applied)
- Multiple operational modes enabling flexible failover behavior per deployment
- Compatibility across diverse embedded Linux distributions
Technical Approach
The bypass solution implements a two-layer control architecture. A software watchdog continuously monitors gateway application health through periodic heartbeat signals. If heartbeats cease within a configurable timeout window, the watchdog signals a hardware relay to transition operational mode.
Hardware relays provide four switching configurations:
- Direct: Normal gateway operation with data flowing through appliance
- Bypass: Traffic reroutes around the gateway, connecting upstream and downstream networks directly
- Block: All traffic halts, protecting against malformed data transmission during failures
- Disconnect: Complete electrical isolation, equivalent to physical unplug
Unique circuit design protects relay state during sudden power loss. This overcomes a critical failure mode where unpredictable power transitions could leave relays in an unsafe state. The circuit maintains capacitor charge sufficient to complete relay state transitions even as system power collapses.
Software integration uses RESTful APIs exposed through Atlas OS, enabling centralized control systems to query relay status, manually trigger transitions, and configure watchdog timeout parameters.
Implementation Notes
Linux kernel compatibility spanning 2.6 through 5.x reflects evolution across a decade of embedded Linux development. The bypass mechanism operates independently of specific kernel subsystems, using standard character device drivers to control relay hardware. This compatibility ensures long product lifecycle without forced OS upgrades.
Deployment scenarios include industrial IoT gateways at remote sites where power cycling is the only recovery option, and mission-critical sensor aggregation points where brief data loss could invalidate entire datasets. The bypass mechanism permits graceful degradation: downstream applications receive either filtered data (via bypass mode) or complete connectivity loss (via block mode), avoiding corrupted or partial data that could trigger false analytics.
Atlas OS integration provides centralized gateway management. A control plane queries bypass module status across distributed gateways, identifying which systems require attention. Automated recovery procedures can trigger bypass mode when watchdog timeouts occur, restoring data connectivity while problematic gateways enter recovery sequences.
Challenge-Solution Mapping
/table
Challenge | Solution
Unplanned gateway failures halt data collection indefinitely | Software watchdog detects failure and triggers automatic bypass
Manual failover requires on-site technician intervention | Hardware relay switching operates without human action
Sudden power loss leaves relays in undefined state | Capacitor-protected circuit ensures safe state transitions
Single failure response mode unsuitable for diverse deployments | Four operational modes (direct, bypass, block, disconnect)
Kernel-specific drivers limit OS flexibility | Cross-kernel compatibility spanning 2.6 through 5.x
Remote gateways lack centralized management visibility | Atlas OS RESTful API enables distributed status monitoring
Corrupted data during partial failure damages analytics | Block mode halts transmission rather than permitting bad data
/endtable
Specifications Snapshot
/table
Specification | Detail
Watchdog Mechanism | Software heartbeat monitoring
Failover Trigger | Configurable timeout on missed heartbeats
Relay Switching | Hardware-based, sub-second response
Operational Modes | Direct, bypass, block, disconnect
Power Loss Protection | Capacitor-protected circuit design
Linux Compatibility | Kernel 2.6 through 5.x
Control Interface | RESTful API via Atlas OS
Application Scope | IoT gateways, edge appliances
/endtable
Key Takeaways
1. Software watchdog + hardware relay architecture eliminates single-point-of-failure – Combined software monitoring and hardware failover ensure graceful degradation during gateway malfunctions without data loss.
2. Capacitor-protected circuit prevents undefined relay state during power loss – Specialized circuit design overcomes critical failure mode where system power collapse could leave relays in uncertain state.
3. Four operational modes enable deployment-specific failover behavior – Direct/bypass/block/disconnect modes accommodate diverse recovery requirements from graceful rerouting to complete isolation.
4. Cross-kernel Linux compatibility permits extended product lifecycles – Support spanning kernel 2.6 through 5.x ensures deployment stability across OS version transitions without forced upgrades.
5. Atlas OS integration enables centralized monitoring of distributed gateways – RESTful API queries enable control planes to identify failing systems and trigger automated recovery workflows.
Contact NEXCOM
For specifications, availability, and technical inquiries, contact NEXCOM via the official website.
Source: https://www.nexcom.com/news/Detail/bypass-all-of-your-critical-iot-gateway-issues
