.NET nanoFramework Seeedstudio HM3301 Basic connectivity

This is a “throw away” .NET nanoFramework application for investigating how Seeedstudio Grove HM3301 Inter Integrated Circuit bus(I²C) connectivity works.

Seeedstudio Grove HM3301 Sensor

My test setup is a simple .NET nanoFramework console application running on an STM32F7691 Discovery board.

Seeedstudio Grove HM3301 + STM32F769 Discovery test rig

The HM3301I2C application has lots of magic numbers from the HM3301 datasheet and is just a tool for exploring how the sensor works.

public static void Main()
{
    I2cConnectionSettings i2cConnectionSettings = new(1, 0x40);

    // i2cDevice.Dispose
    I2cDevice i2cDevice = I2cDevice.Create(i2cConnectionSettings);

    while (true)
    {
        byte[] writeBuffer = new byte[1];
        byte[] readBuffer = new byte[29];

        writeBuffer[0] = 0x88;

        i2cDevice.WriteRead(writeBuffer, readBuffer);

        //i2cDevice.WriteByte(0x88);
        //i2cDevice.Read(readBuffer);

        ushort standardParticulatePm1 = (ushort)(readBuffer[4] << 8);
        standardParticulatePm1 |= readBuffer[5];

        ushort standardParticulatePm25 = (ushort)(readBuffer[6] << 8);
        standardParticulatePm25 |= readBuffer[7];

        ushort standardParticulatePm10 = (ushort)(readBuffer[8] << 8);
                standardParticulatePm10 |= readBuffer[9];

        Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Standard particulate    PM 1.0: {standardParticulatePm1}  PM 2.5: {standardParticulatePm25}  PM 10.0: {standardParticulatePm10} ug/m3");

        ushort atmosphericPm1 = (ushort)(readBuffer[10] << 8);
        atmosphericPm1 |= readBuffer[11];

        ushort atmosphericPm25 = (ushort)(readBuffer[12] << 8);
        atmosphericPm25 |= readBuffer[13];

        ushort atmosphericPm10 = (ushort)(readBuffer[14] << 8);
        atmosphericPm10 |= readBuffer[15];

        Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Atmospheric particulate PM 1.0: {atmosphericPm1:3}  PM 2.5: {atmosphericPm25}  PM 10.0: {atmosphericPm10} ug/m3");


        ushort particulateCountPm03 = (ushort)(readBuffer[16] << 8);
        particulateCountPm03 |= readBuffer[17];

        ushort particulateCountPm05 = (ushort)(readBuffer[18] << 8);
        particulateCountPm05 |= readBuffer[19];

        ushort particulateCountPm1 = (ushort)(readBuffer[20] << 8);
        particulateCountPm1 |= readBuffer[21];

        Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Particulate count       PM 0.3: {particulateCountPm03:3}  PM 0.5: {particulateCountPm05}  PM 1.0: {particulateCountPm1} ug/m3");


        ushort particleCountPm25 = (ushort)(readBuffer[22] << 8);
        particleCountPm25 |= readBuffer[23];

        ushort particleCountPm5 = (ushort)(readBuffer[24] << 8);
        particleCountPm5 |= readBuffer[25];

        ushort particleCountPm10 = (ushort)(readBuffer[26] << 8);
        particleCountPm10 |= readBuffer[27];

        Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Particle count/0.1L     PM 2.5: {particleCountPm25}  PM 5.0: {particleCountPm5}  PM 10.0: {particleCountPm10} particles/0.1L");


        byte checksum = 0;
        for (int i = 0; i < readBuffer.Length - 1; i++)
        {
            checksum += readBuffer[i];
        }
        Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Checksum payload:{readBuffer[28]} calculated:{checksum}");
        Console.WriteLine("");

        Thread.Sleep(5000);
    }
}

The unpacking of the value standard particulate, particulate count and particle count values is fairly repetitive, but I will fix it in the next version.

Visual Studio 2022 Debug output

The checksum calculation isn’t great even a simple cyclic redundancy check(CRC) would be an improvement on summing the 28 bytes of the payload.

.NET nanoFramework Qorvo DW1000 – RAK13801 Device SPI

When developing libraries it’s good to have a selection of different platforms for testing as this can significantly improve the quality and robustness of the implementation. A few months ago I noticed that RAK Wireless have a UWB Module Decawave DW1000 Wisblock so I added one to an order.

My second Qorvo DW1000 setup is a RAK120000 Wisblock Core module, on a RAK19007 WisBlock Base with a RAK13801 WisBlock Wireless module

RAK12000 + RAK19007 + RAK13801 test platform

The Qorvo DW1000 module has a Serial Peripheral Interface (SPI) so the Master In Slave Out(MISO), Master Out Slave In(MOSI), Serial Clock(SCLK) and Chip Slave Select(CSS) pins of the RAK11200 WisBiock Core Module have to be setup using the Configuration.SetPinFunction method of the nanoFramework.Hardware.Esp32 library.

RAK11200 Schematic with SPI pins highlighted.
RAK13801 Schematic with SPI pins highlighted.

I have added a couple of C# processor directives (MAKERFABS_ESP32UWB & RAK11200_RAK1907_RAK13801) so the platform that the Qorvo DW1000 module is running on can be configured.

public class Program
{
#if MAKERFABS_ESP32UWB
    private const int SpiBusId = 1;
    private const int chipSelectLine = Gpio.IO04;
#endif
#if RAK11200_RAK1907_RAK13801
    private const int SpiBusId = 1;
    private const int chipSelectLine = Gpio.IO32;
#endif

    public static void Main()
    {
        Thread.Sleep(5000);

        Debug.WriteLine("devMobile.IoT.Dw1000.ShieldSPI starting");

        try
        {
#if MAKERFABS_ESP32UWB
            Configuration.SetPinFunction(Gpio.IO19, DeviceFunction.SPI1_MISO);
            Configuration.SetPinFunction(Gpio.IO23, DeviceFunction.SPI1_MOSI);
            Configuration.SetPinFunction(Gpio.IO18, DeviceFunction.SPI1_CLOCK);
#endif
#if RAK11200_RAK1907_RAK13801
            Configuration.SetPinFunction(Gpio.IO35, DeviceFunction.SPI1_MISO);
            Configuration.SetPinFunction(Gpio.IO25, DeviceFunction.SPI1_MOSI);
            Configuration.SetPinFunction(Gpio.IO33, DeviceFunction.SPI1_CLOCK);
#endif
            var settings = new SpiConnectionSettings(SpiBusId, chipSelectLine)
            {
                ClockFrequency = 2000000,
                Mode = SpiMode.Mode0,
            };

            using (SpiDevice device = SpiDevice.Create(settings))
            {
                while (true)
                {
                    byte[] writeBuffer = new byte[] { 0x0, 0x0, 0x0, 0x0, 0x0 }; // 0x0 = DEV_ID
                    byte[] readBuffer = new byte[writeBuffer.Length];

                    device.TransferFullDuplex(writeBuffer, readBuffer); // 15, 48, 1, 202, 222

                    uint ridTag = (uint)(readBuffer[4]<< 8 | readBuffer[3]);
                    byte model = readBuffer[2];
                    byte ver = (byte)(readBuffer[1] >> 4);
                    byte rev = (byte)(readBuffer[1] & 0x0f);

                    Debug.WriteLine(String.Format($"RIDTAG 0x{ridTag:X2} MODEL 0x{model:X2} VER 0X{ver:X2} REV 0x{rev:X2}"));

                   Thread.Sleep(10000);
                }
            }
        }
        catch (Exception ex)
        {
            Debug.WriteLine(ex.Message);
        }
    }
}

The alignment of the RAK11200 WisBiock Core Module pins and labels on the circuit diagram tripped me up. My initial configuration caused the device to reboot every time the application started.

Visual Studio 2022 Debug window displaying the decoded value from Register 0x0

At the top of test applications, I usually have a brief delay i.e Thread.Sleep(5000) so I can attach the debugger or erase the flash before the application crashes.

.NET nanoFramework Qorvo DW1000 – Makerfabs Device SPI

The Makerfabs ESP32 UWB(Ultra Wideband) module has a Qorvo DW1000 and Espressif ESP32 module. The Espressif ESP32 module can run the .NET nanoFramework but does not have a Qorvo DW1000 library. (March2023)

Makerfabs ESP32 UWB(Ultra Wide Band) module

Before any coding I used nanoff to “flash” the Espressif ESP32 module with the latest version of .NET nanoFramework

Flashing Makerfabs ESP32 UWB module with nanoff

The Qorvo DW1000 module has a Serial Peripheral Interface (SPI) so the Master In Slave Out(MISO), Master Out Slave In(MOSI), Serial Clock(SCLK) and Chip Slave Select(CSS) pins have to be configured using the Configuration.SetPinFunction method of the nanoFramework.Hadware.Esp32 library

Makerfabs ESP32 UWB module schematic

Even though SPI is an industry standard there are often subtle differences which need to be taken into account when reading from/writing to registers. The DW1000 has a static “Device Identifier” which I used to debug my “proof of concept” code.

DW1000 Datasheet Register Map documentation for Register 0x00

The DeviceSPI program reads register 0x00 and then displays the decoded payload.

public class Program
{
#if MAKERFABS_ESP32UWB
    private const int SpiBusId = 1;
    private const int chipSelectLine = Gpio.IO04;
#endif

    public static void Main()
    {
        Thread.Sleep(5000);

        Debug.WriteLine("devMobile.IoT.Dw1000.ShieldSPI starting");

        try
        {
#if MAKERFABS_ESP32UWB
            Configuration.SetPinFunction(Gpio.IO19, DeviceFunction.SPI1_MISO);
            Configuration.SetPinFunction(Gpio.IO23, DeviceFunction.SPI1_MOSI);
            Configuration.SetPinFunction(Gpio.IO18, DeviceFunction.SPI1_CLOCK);
#endif
            var settings = new SpiConnectionSettings(SpiBusId, chipSelectLine)
            {
                ClockFrequency = 2000000,
                Mode = SpiMode.Mode0,
            };

            using (SpiDevice device = SpiDevice.Create(settings))
            {
                Thread.Sleep(500);

                while (true)
                {
                    /*
                    byte[] writeBuffer = new byte[] { 0x0, 0x0, 0x0, 0x0, 0x0 }; // 0x0 = DEV_ID
                    byte[] readBuffer = new byte[writeBuffer.Length];

                    device.TransferFullDuplex(writeBuffer, readBuffer); // 15, 48, 1, 202, 222
                    */

                    byte[] writeBuffer = new byte[] { 0x0 }; // 0x0 = DEV_ID
                    byte[] readBuffer = new byte[5];

                    device.TransferFullDuplex(writeBuffer, readBuffer); // 15, 48, 1, 202, 222
                       
                    uint ridTag = (uint)(readBuffer[4]<< 8 | readBuffer[3]);
                    byte model = readBuffer[2];
                    byte ver = (byte)(readBuffer[1] >> 4);
                    byte rev = (byte)(readBuffer[1] & 0x0f);

                    Debug.WriteLine(String.Format($"RIDTAG 0x{ridTag:X2} MODEL 0x{model:X2} VER 0X{ver:X2} REV 0x{rev:X2}"));

                    Thread.Sleep(10000);
                }
            }
        }
        catch (Exception ex)
        {
            Debug.WriteLine(ex.Message);
        }
    }
}
Visual Studio 2022 Debug window displaying the decoded value from Register 0x0

The DW1000 User Manual is > 240 pages, with roughly 140 pages of detailed documentation about the DW1000 register set so progress will be slow.

.NET nanoFramework SHT20 Basic connectivity

A couple of years ago I wrote a .NET Core library for the Sensirion SHT20 temperature and humidity(Waterproof) sensor from DFRobot. This .NET nanoFramework version was “inspired” by the .NET Core library version, though I have added some message validation functionality.

DF Robot SHT20 Waterproof sensor

My test setup is a simple .NET nanoFramework console application running on an STM32F7691 Discovery board.

Discovery STM32F769 + SHT20 Testrig

The SH20DeviceI2C application has lots of magic numbers from the SHT20 datasheet and was just a tool for exploring how the sensor works.

 public static void Main()
{
    I2cConnectionSettings i2cConnectionSettings = new(1, 0x40);

    // i2cDevice.Dispose in final program
    I2cDevice i2cDevice = I2cDevice.Create(i2cConnectionSettings);

    while (true)
    {
        byte[] readBuffer = new byte[3] { 0, 0, 0 };

        // First temperature
        i2cDevice.WriteByte(0xF3);

        //Thread.Sleep(50); // no go -46.8
        //Thread.Sleep(60);
        Thread.Sleep(70);
        //Thread.Sleep(90);
        //Thread.Sleep(110);

        i2cDevice.Read(readBuffer);

        ushort temperatureRaw = (ushort)(readBuffer[0] << 8);
        temperatureRaw += readBuffer[1];

        //Debug.WriteLine($"Raw {temperatureRaw}");

        double temperature = temperatureRaw * (175.72 / 65536.0) - 46.85;

        // Then read the Humidity
        i2cDevice.WriteByte(0xF5);

        //Thread.Sleep(50);  
        //Thread.Sleep(60);  
        Thread.Sleep(70);  
        //Thread.Sleep(90);  
        //Thread.Sleep(110);   
                
        i2cDevice.Read(readBuffer);

        ushort humidityRaw = (ushort)(readBuffer[0] << 8);
        humidityRaw += readBuffer[1];

        //Debug.WriteLine($"Raw {humidityRaw}");

        double humidity = humidityRaw * (125.0 / 65536.0) - 6.0;

        //Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Temperature:{temperature:F1}°C");
        //Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Humidity:{humidity:F0}%");
        Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Temperature:{temperature:F1}°C Humidity:{humidity:F0}%");

        Thread.Sleep(1000);
    }
}

While tinkering with the sensor I found that having a short delay between initiating the temperature reading (TemperatureNoHold = 0xF3 was used so as not to hang up the I2C bus) and reading the value was important.

Temperature value without Thread.Sleep

When I ran the application without a Thread.Sleep(70) the temperature and/or humidity the values were incorrect and sometimes quite random.

Temperature value with Thread.Sleep(70)
Humidity value without Thread.Sleep
Humidity value with Thread.Sleep(70)
Temperature and Humidity values with Thread.Sleep(70)

The .NET Core library didn’t validate the message payload Cyclic Redundancy Check (CRC) so I have added that in this version

void CheckCrc(byte[] bytes, byte bytesLen, byte checksum)
{
    var crc = 0;

    for (var i = 0; i < bytesLen; i++)
    {
        crc ^= bytes[i];
        for (var bit = 8; bit > 0; --bit)
        {
            crc = ((crc & 0x80) == 0x80) ? ((crc << 1) ^ CrcPolynomial) : (crc << 1);
        }
    }

    if (crc != checksum)
    {
        throw new Exception("CRC Error");
    }
}

The CheckCrc is called in Temperature and Humidity methods.

public double Temperature()
{
    byte[] readBuffer = new byte[3] { 0, 0, 0 };
    if (_i2cDevice == null)
    {
        throw new ArgumentNullException(nameof(_i2cDevice));
    }

    _i2cDevice.WriteByte(TemperatureNoHold);

    Thread.Sleep(ReadingWaitmSec);

    _i2cDevice.Read(readBuffer);

    CheckCrc(readBuffer, 2, readBuffer[2]);

    ushort temperatureRaw = (ushort)(readBuffer[0] << 8);
    temperatureRaw += readBuffer[1];

    double temperature = temperatureRaw * (175.72 / 65536.0) - 46.85;

    return temperature;
}

I’m going to soak test the library for a week to check that is working okay, then refactor the code so it can be added to the nanoFramework IoT.Device Library repository.

Swarm Space – Asset Tracker Payload Formatter

After writing Swarm Space – Payload Formatter Debugging I then tested it creating a new payload formatter for my new Swarm Asset Tracker.

Swarm Asset Tracker device

The Swarm Asset Tracker has a slightly different payload to the Swarm Eval Kit which is detailed in the product manual.

Swarm Asset Tracker JSON payload

The first message sent shortly after I powered up the device had the latitude and longitude of Null Island

The Asset Tracker UserApplicationId is 65002 and the payload is similar to the Swarm Eval Kit. I created some message payloads (location of Christchurch Cathedral) for testing.

The JSON payload sent by my Swarm Asset Tracker

{
  "dt": 1677396395,
  "lt": -43.5333,
  "ln": 172.6333,
  "al": 25,
  "sp": 0,
  "hd": 126,
  "gj": 92,
  "gs": 1,
  "bv": 4103,
  "tp": 20,
  "rs": -110,
  "tr": -107,
  "ts": 3,
  "td": 1677396357,
  "hp": 166,
  "vp": 187,
  "tf": 36526
}

The Base64 representation of the payload sent by my Swarm Asset Tracker

ew0KICAiZHQiOiAxNjc3Mzk2Mzk1LA0KICAibHQiOiAtNDMuNTMzMywNCiAgImxuIjogMTcyLjYzMzMsDQogICJhbCI6IDI1LA0KICAic3AiOiAwLA0KICAiaGQiOiAxMjYsDQogICJnaiI6IDkyLA0KICAiZ3MiOiAxLA0KICAiYnYiOiA0MTAzLA0KICAidHAiOiAyMCwNCiAgInJzIjogLTExMCwNCiAgInRyIjogLTEwNywNCiAgInRzIjogMywNCiAgInRkIjogMTY3NzM5NjM1NywNCiAgImhwIjogMTY2LA0KICAidnAiOiAxODcsDQogICJ0ZiI6IDM2NTI2DQp9

The initial version of my payload formatter

using System;
using System.Collections.Generic;
using System.Globalization;
using Newtonsoft.Json.Linq;

public class FormatterUplink : PayloadFormatter.IFormatterUplink
{
    public JObject Evaluate(IDictionary<string, string> properties, uint organisationId, uint deviceId, byte deviceType, ushort userApplicationId, JObject payloadJson, string payloadText, byte[] payloadBytes)
    {
        JObject telemetryEvent = new JObject();

        if ((payloadText != "") && (payloadJson != null))
        {
            JObject location = new JObject();

            location.Add("lat", payloadJson.GetValue("lt"));
            location.Add("lon", payloadJson.GetValue("ln"));
            location.Add("alt", payloadJson.GetValue("al"));

            telemetryEvent.Add("DeviceLocation", location);
        }

        // Course & speed
        telemetryEvent.Add("Course", payloadJson.GetValue("hd"));
        telemetryEvent.Add("Speed", payloadJson.GetValue("sp"));

        // Battery voltage
        telemetryEvent.Add("BatteryVoltage", payloadJson.GetValue("bv"));

        // RSSI
        telemetryEvent.Add("RSSI", payloadJson.GetValue("rs"));

        properties.Add("iothub-creation-time-utc", DateTimeOffset.FromUnixTimeSeconds((long)payloadJson.GetValue("dt")).ToString("s", CultureInfo.InvariantCulture));

        return telemetryEvent;
    }
}

The PayloadFormatterMaintenanceApplication command line I used for testing my Swarm Asset Tracker payload formatter

The console output of my Swarm Asset Tracker payload formatter

The PayloadFormatterMaintenanceApplication is better than trying to debug a payload formatter in a staging/production environment.

Currently the payload formatters still have to be manually uploaded to the application’s Azure Blob Storage for final testing.

Swarm Space – Payload Formatter Debugging

After Swarm Space – Uplink Payload Formatters revisited I wrote a couple of payload formatters and they were easy to get wrong and the Azure Application Insights error messages were unhelpful.

namespace PayloadFormatter // Additional namespace for shortening interface when usage in formatter code
{
    using System.Collections.Generic;

    using Newtonsoft.Json.Linq;

    public interface IFormatterUplink
    {
        public JObject Evaluate(IDictionary<string, string> properties, uint organisationId, uint deviceId, byte deviceType, ushort userApplicationId, JObject payloadJson, string payloadText, byte[] payloadBytes);
    }

    public interface IFormatterDownlink
    {
        public byte[] Evaluate(IDictionary<string, string> properties, uint organisationId, uint deviceId, byte deviceType, ushort userApplicationId, JObject payloadJson, string payloadText, byte[] payloadBytes);
    }
}

The definitions of the uplink & downlink payload formatter evaluator interfaces have been updated and shifted to a new project.

Visual Studio 2022 Solution with payloadformatter maintenance application

I built a console application to help with developing and debugging uplink or downlink formatters. The application has a number of command line parameters which specify the formatter to be used, UserApplicationId, OrganizationId, DeviceType etc.

public class CommandLineOptions
{
    [Option('d', "Direction", Required = true, HelpText = "Test Uplink or DownLink formatter")]
	public string Direction { get; set; }

    [Option('p', "filename", HelpText = "Uplink or Downlink Payload file name")]
    public string PayloadFilename { get; set; } = string.Empty;

    [Option('o', "OrganisationId", Required = true, HelpText = "Organisation unique identifier")]
    public uint OrganizationId { get; set; }

    [Option('i', "DeviceId", Required = true, HelpText = "Device unique identitifer")]
    public uint DeviceId { get; set; }

    [Option('t', "DeviceType", Required = true, HelpText = "Device type number")]
    public byte DeviceType { get; set; }

    [Option('u', "UserApplicationId", Required = true, HelpText = "User Application Id")]
    public ushort UserApplicationId { get; set; }

    [Option('h', "SwarmHiveReceivedAtUtc", HelpText = "Swarm Hive received at time UTC")]
    public DateTime? SwarmHiveReceivedAtUtc { get; set; }

    [Option('w', "UplinkWebHookReceivedAtUtc", HelpText = "Webhook received at time UTC")]
    public DateTime? UplinkWebHookReceivedAtUtc { get; set; }

    [Option('s', "Status", HelpText = "Uplink local file system file name")]
    public byte? Status { get; set; }

    [Option('c', "Client", HelpText = "Uplink local file system file name")]
    public string Client { get; set; } 
 }

The downlink formatter (similar approach for uplink) loads the sample file as an array of bytes, then tries to convert it to text, and finally to JSON. Then the formatter code is “compiled” and the executed with the file payload and command line parameters.

private static async Task DownlinkFormatterCore(CommandLineOptions options)
{
    Dictionary<string, string> properties = new Dictionary<string, string>();

    string formatterFolder = Path.Combine(Environment.CurrentDirectory, "downlink");
    Console.WriteLine($"Downlink- uplinkFormatterFolder: {formatterFolder}");

    string formatterFile = Path.Combine(formatterFolder, $"{options.UserApplicationId}.cs");
    Console.WriteLine($"Downlink- UserApplicationId: {options.UserApplicationId}");
    Console.WriteLine($"Downlink- Payload formatter file: {formatterFile}");

    PayloadFormatter.IFormatterDownlink evalulator;
    try
    {
        evalulator = CSScript.Evaluator.LoadFile<PayloadFormatter.IFormatterDownlink>(formatterFile);
     }
    catch (CSScriptLib.CompilerException cex)
    {
        Console.Write($"Loading or compiling file:{formatterFile} failed Exception:{cex}");
        return;
    }

    string payloadFilename = Path.Combine(formatterFolder, options.PayloadFilename);
    Console.WriteLine($"Downlink- payloadFilename:{payloadFilename}");
    byte[] uplinkBytes;

    try
    {
        uplinkBytes = File.ReadAllBytes(payloadFilename);
    }
    catch (DirectoryNotFoundException dex)
    {
        Console.WriteLine($"Uplink payload filename directory {formatterFolder} not found:{dex}");
        return;
    }
    catch (FileNotFoundException fnfex)
    {
        Console.WriteLine($"Uplink payload filename {payloadFilename} not found:{fnfex}");
        return;
    }
    catch (FormatException fex)
    {
        Console.WriteLine($"Uplink payload file invalid format {payloadFilename} not found:{fex}");
        return;
    }

    // See if payload can be converted to a string
    string uplinkText = string.Empty;
    try
    {
        uplinkText = Encoding.UTF8.GetString(uplinkBytes);
    }
    catch (FormatException fex)
    {
        Console.WriteLine("Encoding.UTF8.GetString failed:{0}", fex.Message);
    }

    // See if payload can be converted to JSON
    JObject uplinkJson;
    try
    {
        uplinkJson = JObject.Parse(uplinkText);
    }
    catch (JsonReaderException jrex)
    {
        Console.WriteLine("JObject.Parse failed Exception:{1}", jrex);

        uplinkJson = new JObject();
    }

    Console.WriteLine("Properties");
    foreach (var property in properties)
    {
        Console.WriteLine($"{property.Key}:{property.Value}");
    }

    // Transform the byte and optional text and JSON payload
    Byte[] payload;
    try
    {
        payload = evalulator.Evaluate(properties, options.OrganizationId, options.DeviceId, options.DeviceType, options.UserApplicationId, uplinkJson, uplinkText, uplinkBytes);
    }
    catch (Exception ex)
    {
        Console.WriteLine($"evalulatorUplink.Evaluate failed Exception:{ex}");
        return;
    }

    Console.WriteLine("Payload");
    Console.WriteLine(Convert.ToBase64String(payload));
}

The sample JSON payload is what would be sent by Azure IoT Central to a device to configure the fan speed

Azure IoT Central M138 Breakout device template with the Fan Status command selected
{
  "FanStatus": 2
}

If the downlink payload formatter is compiled and executes successfully the Base64 representation output is displayed

using System;
using System.Collections.Generic;
using Newtonsoft.Json.Linq;

public class FormatterDownlink : PayloadFormatter.IFormatterDownlink
{
    public byte[] Evaluate(IDictionary<string, string> properties, uint organisationId, uint deviceId, byte deviceType, ushort userApplicationId, JObject payloadJson, string payloadText, byte[] payloadBytes)
    {
        byte? status = payloadJson.Value<byte?>("FanStatus");

        if ( status.HasValue ) 
        { 
            return new byte[] { status.Value };
        }

        return new byte[]{};
    }
}

If the downlink payload formatter syntax is incorrect e.g. { status.Value ; }; an error message with the line and column is displayed.

using System;
using System.Collections.Generic;
using Newtonsoft.Json.Linq;

public class FormatterDownlink : PayloadFormatter.IFormatterDownlink
{
    public byte[] Evaluate(IDictionary<string, string> properties, uint organisationId, uint deviceId, byte deviceType, ushort userApplicationId, JObject payloadJson, string payloadText, byte[] payloadBytes)
    {
        byte? status = payloadJson.Value<byte?>("FanStatus");

        if ( status.HasValue ) 
        {
            return new byte[] { status.Value ; };
        }

        return new byte[]{};
    }
}

If the downlink payload formatter syntax is correct but execution fails (in the example code division by zero) an error message is displayed.

using System;
using System.Collections.Generic;
using Newtonsoft.Json.Linq;

public class FormatterDownlink : PayloadFormatter.IFormatterDownlink
{
    public byte[] Evaluate(IDictionary<string, string> properties, uint organisationId, uint deviceId, byte deviceType, ushort userApplicationId, JObject payloadJson, string payloadText, byte[] payloadBytes)
    {
        byte? status = payloadJson.Value<byte?>("FanStatus");

        if ( status.HasValue ) 
        {
            int divideByZero = 10;

            divideByZero = divideByZero / 0;

            return new byte[] { status.Value };
        }

        return new byte[]{};
    }
}

The PayloadFormatterMaintenanceApplication makes it significantly easier to develop formatters. Currently the payload formatters have to be manually uploaded to the application’s Azure Blob Storage for final testing.

Swarm Space – Replacing the OpenAPI Client

At the start of this project I used NSwag and Open API Swagger definition file (provided by Swarm Space technical support) to generate a Swarm Space Bumble bee hive client and the core of a simulator.

Swarm Space Bumble hive classes in Visual Studio 2022

My SwarmSpaceAzureIoTConnector project only needed to login, get a list of devices and send messages so all the additional functionality was never going to be used. The method to send a message didn’t work, the class used for the payload (UserMessage) appears to be wrong.

OpenAPI Swagger docs for sending a message

The Open API Swagger definition for sending a message to a device

"post": {
        "tags": [ "messages" ],
        "summary": "POST user messages",
        "description": "<p>This endpoint submits a JSON formatted UserMessage object for delivery to a Swarm device. A JSON object is returned with a newly assigned <code>packetId</code> and <code>status</code> of<code>OK</code> on success, or <code>ERROR</code> (with a description of the error) on failure.</p><p>The current user must have access to the <code>userApplicationId</code> and <code>device</code> given inside the UserMessage JSON. The device must also have the ability to receive messages from the Hive (\"two-way communication\") enabled. If these conditions are not met, a response with status code 403 (Forbidden) will be returned.</p><p>Note that the <code>data</code> field is the <b>Base64-encoded</b> version of the data to be sent. This allows the sending of binary, as well as text, data.</p>",
        "operationId": "addApplicationMessage",
        "requestBody": {
          "content": { "application/json": { "schema": { "$ref": "#/components/schemas/UserMessage" } } },
          "required": true
        },
        "responses": {
          "401": {
            "description": "Unauthorized",
            "content": { "*/*": { "schema": { "$ref": "#/components/schemas/ApiError" } } }
          },
          "403": {
            "description": "Forbidden",
            "content": { "*/*": { "schema": { "$ref": "#/components/schemas/ApiError" } } }
          },
          "400": {
            "description": "Bad Request",
            "content": { "*/*": { "schema": { "$ref": "#/components/schemas/ApiError" } } }
          },
          "200": {
            "description": "OK",
            "content": { "application/json": { "schema": { "$ref": "#/components/schemas/PacketPostReturn" } } }
          }
        }
      }
    },

The Open API Swagger definition for a UserMessage

[System.CodeDom.Compiler.GeneratedCode("NJsonSchema", "13.17.0.0 (NJsonSchema v10.8.0.0 (Newtonsoft.Json v13.0.0.0))")]
    public partial class UserMessage
    {
        /// <summary>
        /// Swarm packet ID
        /// </summary>
        [Newtonsoft.Json.JsonProperty("packetId", Required = Newtonsoft.Json.Required.Always)]
        public long PacketId { get; set; }

        /// <summary>
        /// Swarm message ID. There may be multiple messages for a single message ID. A message ID represents an intent to send a message, but there may be multiple Swarm packets that are required to fulfill that intent. For example, if a Hive -&gt; device message fails to reach its destination, automatic retry attempts to send that message will have the same message ID.
        /// </summary>
        [Newtonsoft.Json.JsonProperty("messageId", Required = Newtonsoft.Json.Required.DisallowNull, NullValueHandling = Newtonsoft.Json.NullValueHandling.Ignore)]
        public long MessageId { get; set; }

        /// <summary>
        /// Swarm device type
        /// </summary>
        [Newtonsoft.Json.JsonProperty("deviceType", Required = Newtonsoft.Json.Required.Always)]
        public int DeviceType { get; set; }

        /// <summary>
        /// Swarm device ID
        /// </summary>
        [Newtonsoft.Json.JsonProperty("deviceId", Required = Newtonsoft.Json.Required.Always)]
        public int DeviceId { get; set; }

        /// <summary>
        /// Swarm device name
        /// </summary>
        [Newtonsoft.Json.JsonProperty("deviceName", Required = Newtonsoft.Json.Required.DisallowNull, NullValueHandling = Newtonsoft.Json.NullValueHandling.Ignore)]
        public string DeviceName { get; set; }

        /// <summary>
        /// Direction of message
        /// </summary>
        [Newtonsoft.Json.JsonProperty("direction", Required = Newtonsoft.Json.Required.Always)]
        public int Direction { get; set; }

        /// <summary>
        /// Message data type, always = 6
        /// </summary>
        [Newtonsoft.Json.JsonProperty("dataType", Required = Newtonsoft.Json.Required.Always)]
        public int DataType { get; set; }

        /// <summary>
        /// Application ID
        /// </summary>
        [Newtonsoft.Json.JsonProperty("userApplicationId", Required = Newtonsoft.Json.Required.Always)]
        public int UserApplicationId { get; set; }

        /// <summary>
        /// Organization ID
        /// </summary>
        [Newtonsoft.Json.JsonProperty("organizationId", Required = Newtonsoft.Json.Required.Always)]
        public int OrganizationId { get; set; }

        /// <summary>
        /// Length of data (in bytes) before base64 encoding
        /// </summary>
        [Newtonsoft.Json.JsonProperty("len", Required = Newtonsoft.Json.Required.DisallowNull, NullValueHandling = Newtonsoft.Json.NullValueHandling.Ignore)]
        public int Len { get; set; }

        /// <summary>
        /// Base64 encoded data string
        /// </summary>
        [Newtonsoft.Json.JsonProperty("data", Required = Newtonsoft.Json.Required.Always)]
        [System.ComponentModel.DataAnnotations.Required(AllowEmptyStrings = true)]
        public byte[] Data { get; set; }

        /// <summary>
        /// Swarm packet ID of acknowledging packet from device
        /// </summary>
        [Newtonsoft.Json.JsonProperty("ackPacketId", Required = Newtonsoft.Json.Required.DisallowNull, NullValueHandling = Newtonsoft.Json.NullValueHandling.Ignore)]
        public long AckPacketId { get; set; }

        /// <summary>
        /// Message status. Possible values:
        /// <br/>0 = incoming message (from a device)
        /// <br/>1 = outgoing message (to a device)
        /// <br/>2 = incoming message, acknowledged as seen by customer. OR a outgoing message packet is on groundstation
        /// <br/>3 = outgoing message, packet is on satellite
        /// <br/>-1 = error
        /// <br/>-3 = failed to deliver, retrying
        /// <br/>-4 = failed to deliver, will not re-attempt
        /// <br/>
        /// </summary>
        [Newtonsoft.Json.JsonProperty("status", Required = Newtonsoft.Json.Required.DisallowNull, NullValueHandling = Newtonsoft.Json.NullValueHandling.Ignore)]
        public int Status { get; set; }

        /// <summary>
        /// Time that the message was received by the Hive
        /// </summary>
        [Newtonsoft.Json.JsonProperty("hiveRxTime", Required = Newtonsoft.Json.Required.Always)]
        [System.ComponentModel.DataAnnotations.Required(AllowEmptyStrings = true)]
        public System.DateTimeOffset HiveRxTime { get; set; }

        private System.Collections.Generic.IDictionary<string, object> _additionalProperties;

        [Newtonsoft.Json.JsonExtensionData]
        public System.Collections.Generic.IDictionary<string, object> AdditionalProperties
        {
            get { return _additionalProperties ?? (_additionalProperties = new System.Collections.Generic.Dictionary<string, object>()); }
            set { _additionalProperties = value; }
        }

    }

After several attempts I gave up and have rebuilt the required Bumble bee hive integration with RestSharp

public async Task SendAsync(uint organisationId, uint deviceId, byte deviceType, ushort userApplicationId, byte[] data, CancellationToken cancellationToken)
{
    await TokenRefresh(cancellationToken);

    _logger.LogInformation("SendAsync: OrganizationId:{0} DeviceType:{1} DeviceId:{2} UserApplicationId:{3} Data:{4} Enabled:{5}", organisationId, deviceType, deviceId, userApplicationId, Convert.ToBase64String(data), _bumblebeeHiveSettings.DownlinkEnabled);

    Models.MessageSendRequest message = new Models.MessageSendRequest()
    {
        OrganizationId = (int)organisationId,
        DeviceType = deviceType,
        DeviceId = (int)deviceId,
        UserApplicationId = userApplicationId,
        Data = data,
    };

    RestClientOptions restClientOptions = new RestClientOptions()
    {
        BaseUrl = new Uri(_bumblebeeHiveSettings.BaseUrl),
        ThrowOnAnyError = true,
    };

    using (RestClient client = new RestClient(restClientOptions))
    {
        RestRequest request = new RestRequest("api/v1/messages", Method.Post);

        request.AddBody(message);

        request.AddHeader("Authorization", $"bearer {_token}");

        // To save the limited monthly allocation of mesages downlinks can be disabled
        if (_bumblebeeHiveSettings.DownlinkEnabled)
        {
           var response = await client.PostAsync<Models.MessageSendResponse>(request, cancellationToken);

            _logger.LogInformation("SendAsync-Result:{Status} PacketId:{PacketId}", response.Status, response.PacketId);
        }
    }
}

The new Data Transfer Objects(DTOs) were “inspired” by the NSwag generated ones.

public partial class MessageSendRequest
{
    /// <summary>
    /// Swarm device type
    /// </summary>
    [Newtonsoft.Json.JsonProperty("deviceType", Required = Newtonsoft.Json.Required.Always)]
    public int DeviceType { get; set; }

    /// <summary>
    /// Swarm device ID
    /// </summary>
    [Newtonsoft.Json.JsonProperty("deviceId", Required = Newtonsoft.Json.Required.Always)]
    public int DeviceId { get; set; }

    /// <summary>
    /// Application ID
    /// </summary>
    [Newtonsoft.Json.JsonProperty("userApplicationId", Required = Newtonsoft.Json.Required.Always)]
    public int UserApplicationId { get; set; }

    /// <summary>
    /// Organization ID
    /// </summary>
    [Newtonsoft.Json.JsonProperty("organizationId", Required = Newtonsoft.Json.Required.Always)]
    public int OrganizationId { get; set; }

    /// <summary>
    /// Base64 encoded data string
    /// </summary>
    [Newtonsoft.Json.JsonProperty("data", Required = Newtonsoft.Json.Required.Always)]
    [System.ComponentModel.DataAnnotations.Required(AllowEmptyStrings = true)]
    public byte[] Data { get; set; }
}

public class MessageSendResponse
{
    /// <summary>
    /// Swarm packet ID.
    /// </summary>
    [Newtonsoft.Json.JsonProperty("packetId", Required = Newtonsoft.Json.Required.Always)]
    public long PacketId { get; set; }

    /// <summary>
    /// Submission status, "OK" or "ERROR" with a description of the error.
    /// </summary>
    [Newtonsoft.Json.JsonProperty("status", Required = Newtonsoft.Json.Required.Always)]
    [System.ComponentModel.DataAnnotations.Required(AllowEmptyStrings = true)]
    public string Status { get; set; }
}

The RestSharp based approach is significantly smaller and less complex….

Azure Functions Isolated Worker support for VB.Net 4.8

As part of my “day job” I spend a bit of time working with VB.Net 4.X “legacy” projects doing upgrades, and bug fixes. Currently I am updating a number of Windows Service applications to run as Microsoft Azure Functions. With the release of the Azure functions runtime V4 Isolated Worker Processes with .NET Framework V4.8 support this is the last post in my Azure Functions with VB.Net 4.X and Azure Functions with VB.Net on .NET Core V6 series.

I have published source code for Azure Storage BlobTrigger, Azure Storage QueueTrigger, and TimerTriggers.

Visual Studio Solution explorer Azure Functions projects

All of the examples now have a program.vb which initialises the Trigger.

Namespace VBNet....TriggerIsolated
    Friend Class Program
        Public Shared Sub Main(ByVal args As String())
            Call FunctionsDebugger.Enable()

            Dim host = New HostBuilder().ConfigureFunctionsWorkerDefaults().Build()

            host.Run()
        End Sub
    End Class
End Namespace

All of the Isolated worker process Triggers displayed this message which appeared to be benign.

Csproj not found in C:\Users\..\VBNetHttpTriggerIsolated\bin\Debug\net48 directory tree. Skipping user secrets file configuration.

There were a lot of articles about problems building Docker images but the only relevant ones appeared to talk about getting F# and other .NET Core languages to work in Azure Functions.

Namespace devMobile.Azure.VBNetBlobTriggerIsolated
    Public Class BlobTrigger
        Private ReadOnly _logger As ILogger

        Public Sub New(ByVal loggerFactory As ILoggerFactory)
            _logger = loggerFactory.CreateLogger(Of BlobTrigger)()
        End Sub

        <[Function]("vbnetblobtriggerisolated")>
        Public Sub Run(
        <BlobTrigger("vbnetblobtriggerisolated/{name}", Connection:="blobendpoint")> ByVal myBlob As String, ByVal name As String)

            _logger.LogInformation($"VB.Net NET 4.8 Isolated Blob trigger function Processed blob Name: {name}  Data: {myBlob}")
        End Sub
    End Class
End Namespace

I used Azure Storage Explorer to upload files containing Lorem Ipsum for testing the BlobTrigger.

Azure BlobTrigger function running in the desktop emulator
Azure BlobTrigger Function logging in Application Insights

I used Telerik Fiddler to POST messages to the desktop emulator and Azure endpoints.

Namespace VBNetHttpTriggerIsolated
    Public Class HttpTrigger
        Private Shared executionCount As Int32
        Private ReadOnly _logger As ILogger

        Public Sub New(ByVal loggerFactory As ILoggerFactory)
            _logger = loggerFactory.CreateLogger(Of HttpTrigger)()
        End Sub

        <[Function]("Notifications")>
        Public Function Run(
        <HttpTrigger(AuthorizationLevel.Anonymous, "get", "post")> ByVal req As HttpRequestData) As HttpResponseData
            Interlocked.Increment(executionCount)

            _logger.LogInformation("VB.Net NET 4.8 Isolated HTTP trigger Execution count:{executionCount} Method:{req.Method}", executionCount, req.Method)

            Dim response = req.CreateResponse(HttpStatusCode.OK)
            response.Headers.Add("Content-Type", "text/plain; charset=utf-8")

            Return response
        End Function
    End Class
End Namespace
Azure HttpTrigger Function running in the desktop emulator
Azure HttpTrigger Function logging in Application Insights

I used Azure Storage Explorer to create messages for testing the QueueTrigger

Namespace devMobile.Azure.VBNetQueueTriggerIsolated

    Public Class QueueTrigger
        Private Shared _logger As ILogger
        Private Shared _concurrencyCount As Integer = 0
        Private Shared _executionCount As Integer = 0

        Public Sub New(ByVal loggerFactory As ILoggerFactory)
            _logger = loggerFactory.CreateLogger(Of QueueTrigger)()
        End Sub

        <[Function]("VBNetQueueTriggerIsolated")>
        Public Sub Run(
        <QueueTrigger("vbnetqueuetriggerisolated", Connection:="QueueEndpoint")> ByVal message As String)
            Interlocked.Increment(_concurrencyCount)
            Interlocked.Increment(_executionCount)

            _logger.LogInformation("VB.Net .NET 4.8 Isolated Queue Trigger Concurrency:{_concurrencyCount} ExecutionCount:{_executionCount} Message:{message}", _concurrencyCount, _executionCount, message)

            Interlocked.Decrement(_concurrencyCount)
        End Sub
    End Class
End Namespace
Azure QueueTrigger Function running in the desktop emulator
Azure QueueTrigger Function logging in Application Insights
Namespace devMobile.Azure.VBNetTimerTriggerIsolated
    Public Class TimerTrigger
        Private Shared _logger As ILogger
        Private Shared _executionCount As Integer = 0

        Public Sub New(ByVal loggerFactory As ILoggerFactory)
            _logger = loggerFactory.CreateLogger(Of TimerTrigger)()
        End Sub

        <[Function]("Timer")>
        Public Sub Run(
        <TimerTrigger("0 */1 * * * *")> ByVal myTimer As MyInfo)

            Interlocked.Increment(_executionCount)
            _logger.LogInformation("VB.Net Isolated TimerTrigger next trigger:{0} Execution count:{1}", myTimer.ScheduleStatus.Next, _executionCount)
        End Sub
    End Class
Azure TimerTrigger Function running in the desktop emulator
Azure TimerTrigger Function logging in Application Insights

The development, debugging and deployment of these functions took a lot of time. Initially Azure Application Insights didn’t work when the Azure Isolated Worker triggers were deployed to Azure. After some experimentation I found that Application Insights Connection Strings worked and Application Instrumentation Keys did not.

With the Microsoft: ‘We Do Not Plan to Evolve Visual Basic as a Language this should hopefully be my last post about VB.Net ever.

Swarm Space – Underlying Architecture sorted

After figuring out that calling an Azure Http Trigger function to load the cache wasn’t going to work reliably, I have revisited the architecture one last time and significantly refactored the SwarmSpaceAzuureIoTConnector project.

Visual Studio 2022 solution

The application now has a StartUpService which loads the Azure DeviceClient cache (Lazy Cache) in the background as the application starts up. If an uplink message is received from a SwarmDevice before, it has been loaded by the FunctionsStartup the DeviceClient information is cached and another connection to the Azure IoT Hub is not established.

...
using Microsoft.Azure.Functions.Extensions.DependencyInjection;

[assembly: FunctionsStartup(typeof(devMobile.IoT.SwarmSpaceAzureIoTConnector.Connector.StartUpService))]
namespace devMobile.IoT.SwarmSpaceAzureIoTConnector.Connector
{
...
    public class StartUpService : BackgroundService
    {
        private readonly ILogger<StartUpService> _logger;
        private readonly ISwarmSpaceBumblebeeHive _swarmSpaceBumblebeeHive;
        private readonly Models.ApplicationSettings _applicationSettings;
        private readonly IAzureDeviceClientCache _azureDeviceClientCache;

        public StartUpService(ILogger<StartUpService> logger, IAzureDeviceClientCache azureDeviceClientCache, ISwarmSpaceBumblebeeHive swarmSpaceBumblebeeHive, IOptions<Models.ApplicationSettings> applicationSettings)//, IOptions<Models.AzureIoTSettings> azureIoTSettings)
        {
            _logger = logger;
            _azureDeviceClientCache = azureDeviceClientCache;
            _swarmSpaceBumblebeeHive = swarmSpaceBumblebeeHive;
            _applicationSettings = applicationSettings.Value;
        }

        protected override async Task ExecuteAsync(CancellationToken cancellationToken)
        {
            await Task.Yield();

            _logger.LogInformation("StartUpService.ExecuteAsync start");

            try
            {
                _logger.LogInformation("BumblebeeHiveCacheRefresh start");

                foreach (SwarmSpace.BumblebeeHiveClient.Device device in await _swarmSpaceBumblebeeHive.DeviceListAsync(cancellationToken))
                {
                    _logger.LogInformation("BumblebeeHiveCacheRefresh DeviceId:{DeviceId} DeviceName:{DeviceName}", device.DeviceId, device.DeviceName);

                    Models.AzureIoTDeviceClientContext context = new Models.AzureIoTDeviceClientContext()
                    {
                        OrganisationId = _applicationSettings.OrganisationId,
                        DeviceType = (byte)device.DeviceType,
                        DeviceId = (uint)device.DeviceId,
                    };

                    await _azureDeviceClientCache.GetOrAddAsync(context.DeviceId, context);
                }

                _logger.LogInformation("BumblebeeHiveCacheRefresh finish");
            }
            catch (Exception ex)
            {
                _logger.LogError(ex, "StartUpService.ExecuteAsync error");

                throw;
            }

            _logger.LogInformation("StartUpService.ExecuteAsync finish");
        }
    }
}

The uplink and downlink payload formatters are stored in Azure Blob Storage are compiled (CS-Script) as they are loaded then cached (Lazy Cache)

Azure Storage explorer displaying list of uplink payload formatter blobs.
Azure Storage explorer displaying list of downlink payload formatter blobs.
private async Task<IFormatterDownlink> DownlinkLoadAsync(int userApplicationId)
{
    BlobClient blobClient = new BlobClient(_payloadFormatterConnectionString, _applicationSettings.PayloadFormattersDownlinkContainer, $"{userApplicationId}.cs");

    if (!await blobClient.ExistsAsync())
    {
        _logger.LogInformation("PayloadFormatterDownlink- UserApplicationId:{0} Container:{1} not found using default:{2}", userApplicationId, _applicationSettings.PayloadFormattersUplinkContainer, _applicationSettings.PayloadFormatterUplinkBlobDefault);

        blobClient = new BlobClient(_payloadFormatterConnectionString, _applicationSettings.PayloadFormatterDownlinkBlobDefault, _applicationSettings.PayloadFormatterDownlinkBlobDefault);
    }

    BlobDownloadResult downloadResult = await blobClient.DownloadContentAsync();

    return CSScript.Evaluator.LoadCode<PayloadFormatter.IFormatterDownlink>(downloadResult.Content.ToString());
}

The uplink and downlink formatters can be edited in Visual Studio 2022 with syntax highlighting (currently they have to be manually uploaded).

The SwarmSpaceBumbleebeehive module no longer has public login or logout methods.

    public interface ISwarmSpaceBumblebeeHive
    {
        public Task<ICollection<Device>> DeviceListAsync(CancellationToken cancellationToken);

        public Task SendAsync(uint organisationId, uint deviceId, byte deviceType, ushort userApplicationId, byte[] payload);
    }

The DeviceListAsync and SendAsync methods now call the BumblebeeHive login method after configurable period of inactivity.

public async Task<ICollection<Device>> DeviceListAsync(CancellationToken cancellationToken)
{
        if ((_TokenActivityAtUtC + _bumblebeeHiveSettings.TokenValidFor) < DateTime.UtcNow)
        {
            await Login();
        }

        using (HttpClient httpClient = _httpClientFactory.CreateClient())
       {
            Client client = new Client(httpClient);

            client.BaseUrl = _bumblebeeHiveSettings.BaseUrl;

            httpClient.DefaultRequestHeaders.Add("Authorization", $"bearer {_token}");

            return await client.GetDevicesAsync(null, null, null, null, null, null, null, null, null, cancellationToken);
        }
}

I’m looking at building a webby user interface where users an interactivity list, create, edit, delete formatters with syntax highlighter support, and the executing the formatter with sample payloads.

Swarm Space Azure IoT Connector Identity Translation Gateway Architecture

This approach uses most of the existing building blocks, and that’s it no more changes.

Swarm Space – DeviceClient Cache warming with HTTPTrigger

For C2D messaging to work a device must have a DeviceClient “connection” established to the Azure IoT Hub which is a problem for irregularly connect devices. Sometimes establishing a connection on the first D2C messages is sufficient, especially for devices which only support D2C messaging. An Identity Translation Gateway establishes a connection for each device (see discussion about AMQP Connection multiplexing) so that C2D messages can be sent immediately.

I initially tried building a cache loader with BackgroundService so that the DeviceClient cache would start loading as the application started but interdependencies became problem.

public partial class Connector
{
    [Function("BumblebeeHiveCacheRefresh")]
    public async Task<IActionResult> BumblebeeHiveCacheRefreshRun([HttpTrigger(AuthorizationLevel.Function, "get")] CancellationToken cancellationToken)
    {
        _logger.LogInformation("BumblebeeHiveCacheRefresh start");

        await _swarmSpaceBumblebeeHive.Login(cancellationToken);

        foreach (SwarmSpace.BumblebeeHiveClient.Device device in await _swarmSpaceBumblebeeHive.DeviceListAsync(cancellationToken))
        {
            _logger.LogInformation("BumblebeeHiveCacheRefresh DeviceId:{DeviceId} DeviceName:{DeviceName}", device.DeviceId, device.DeviceName);

            Models.AzureIoTDeviceClientContext context = new Models.AzureIoTDeviceClientContext()
            {
                // TODO seems a bit odd getting this from application settings
                OrganisationId = _applicationSettings.OrganisationId, 
                //UserApplicationId = device.UserApplicationId, deprecated
                DeviceType = (byte)device.DeviceType,
                DeviceId = (uint)device.DeviceId,
            };

            switch (_azureIoTSettings.ApplicationType)
            {
                case Models.ApplicationType.AzureIotHub:
                    switch (_azureIoTSettings.AzureIotHub.ConnectionType)
                    {
                        case Models.AzureIotHubConnectionType.DeviceConnectionString:
                             await _azureDeviceClientCache.GetOrAddAsync<DeviceClient>(device.DeviceId.ToString(), (ICacheEntry x) => AzureIoTHubDeviceConnectionStringConnectAsync(device.DeviceId.ToString(), context));
                            break;
                        case Models.AzureIotHubConnectionType.DeviceProvisioningService:
                             await _azureDeviceClientCache.GetOrAddAsync<DeviceClient>(device.DeviceId.ToString(), (ICacheEntry x) => AzureIoTHubDeviceProvisioningServiceConnectAsync(device.DeviceId.ToString(), context, _azureIoTSettings.AzureIotHub.DeviceProvisioningService));
                            break;
                        default:

                        _logger.LogError("Azure IoT Hub ConnectionType unknown {0}", _azureIoTSettings.AzureIotHub.ConnectionType);

                            throw new NotImplementedException("AzureIoT Hub unsupported ConnectionType");
                    }
                    break;

                case Models.ApplicationType.AzureIoTCentral:
                    await _azureDeviceClientCache.GetOrAddAsync<DeviceClient>(device.DeviceId.ToString(), (ICacheEntry x) => AzureIoTHubDeviceProvisioningServiceConnectAsync(device.DeviceId.ToString(), context, _azureIoTSettings.AzureIoTCentral.DeviceProvisioningService));
                break;

                default:
                    _logger.LogError("AzureIoT application type unknown {0}", _azureIoTSettings.ApplicationType);

                    throw new NotImplementedException("AzureIoT unsupported ApplicationType");
            }
        }

        _logger.LogInformation("BumblebeeHiveCacheRefresh finish");

        return new OkResult();
    }
}

The HTTP WEBSITE_WARMUP_PATH environment variable is used to call the Azure HTTPTrigger Function and this is secured with an x-functions-key header.

Azure Function App Configuration

In the short-term loading the cache with a call to an Azure HTTPTrigger Function works but may timeout issues. When I ran the connector with my 100’s of devices simulator the function timed out every so often.