Error Handling

This diagram shows how the device handles errors and recovers from failures.

flowchart TD
    Operation[Normal Operation] --> Error{Error
Detected?}
    Error -->|No| Operation
    Error -->|Yes| Classify{Error Type?}

    Classify -->|Sensor Error| SensorFail[Log Sensor Failure]
    Classify -->|Network Error| NetworkFail[Log Network Failure]
    Classify -->|Config Error| ConfigFail[Log Config Failure]
    Classify -->|Critical Error| CriticalFail[Critical Failure]

    SensorFail --> SkipReading[Skip This Reading]
    SkipReading --> Schedule[Schedule Next Attempt]
    Schedule --> Operation

    NetworkFail --> RetryJoin{Join Retry
Count?}
    RetryJoin -->|< Max| WaitRetry[Wait & Retry Join]
    RetryJoin -->|≥ Max| ContinueLocal[Continue Without Network]
    WaitRetry --> Operation
    ContinueLocal --> Operation

    ConfigFail --> UseDefault[Load Default Config]
    UseDefault --> Operation

    CriticalFail --> EnableWatchdog[Enable Watchdog]
    EnableWatchdog --> InfiniteLoop[Enter Infinite Loop]
    InfiniteLoop --> WatchdogReset([Watchdog Timeout
Device Resets])
    WatchdogReset --> Boot([Boot Sequence])
    Boot --> Operation

    style Error fill:#fff4e1
    style CriticalFail fill:#ffe1e1
    style WatchdogReset fill:#ffe1e1
    style Operation fill:#e1ffe1

Error Types

mindmap
  root((Errors))
    I2C Errors
      Slave NACK
      Arbitration Lost
      No Alert Response
      Bus Error
    Network Errors
      Join Failure
      TX Failure
      Link Dead
    Config Errors
      Invalid Magic
      Checksum Error
      Out of Range
    Critical Errors
      Hardware Fault
      Memory Corruption

Error Recovery Strategies

Error Type	Detection	Recovery	Impact
Sensor NACK	I2C transaction fails	Skip measurement, retry next cycle	One data point lost
Sensor Timeout	No response after trigger	Skip measurement, retry next cycle	One data point lost
Join Failure	OTAA join times out	Retry with backoff, max 10 attempts	Delayed operation
TX Failure	Transmission fails	Retry on next cycle	One transmission lost
Link Dead	No response from network	Continue measuring, attempt rejoin	Data buffering needed
Config Invalid	Magic bytes wrong	Use default settings	Settings lost
Memory Corrupt	Read/write errors	Watchdog reset	Device restarts
Critical Fault	Unrecoverable error	Watchdog reset	Device restarts

Watchdog Timer

sequenceDiagram
    participant App as Application
    participant WDT as Watchdog Timer
    participant MCU as Microcontroller

    Note over App,MCU: Normal Operation
    App->>WDT: Pet watchdog (reset timer)
    WDT-->>App: OK

    Note over App: Continue working...

    App->>WDT: Pet watchdog (reset timer)
    WDT-->>App: OK

    Note over App,MCU: Critical Error Occurs

    App->>App: Detect critical error
    App->>WDT: Enable watchdog (15ms)
    App->>App: Enter infinite loop

    Note over WDT: 15ms passes...

    WDT->>MCU: Trigger reset
    MCU->>MCU: System reset

    Note over App,MCU: Device reboots

    MCU->>App: Start application
    App->>App: Initialize & resume

Error Logging

The device maintains an error counter for debugging:

Error Code | Count | Last Occurrence
-----------|-------|----------------
ERR_NONE   |   0   | -
SLAVE_NACK |   3   | 2024-10-30 14:23:15
ARB_LOST   |   0   | -
NO_ALERT   |   1   | 2024-10-30 12:45:30
SMBUS_ERR  |   0   | -

Errors are logged to serial output when DEBUG mode is enabled:

[12345] ERROR: Sensor NACK on address 0x36
[12348] Retrying measurement in next cycle

Recovery Time Estimates

• Sensor Error: Immediate (skip to next cycle)
• Network Join Failure: 10-60 seconds per retry
• Config Error: <1 second (load defaults)
• Watchdog Reset: 2-5 seconds (full reboot)

Best Practices

1. Always validate sensor data before transmission
2. Retry with backoff for transient errors
3. Use watchdog as last resort for critical failures
4. Log errors for debugging and maintenance
5. Fail gracefully - continue operation when possible