Installation

Install the dependency:

yarn install @corvina/device-client

Run the device using the given runner:

import dotenv from "dotenv"
dotenv.config()

import { DeviceRunnerService, DeviceService } from '@corvina/device-client';

const devRunner = new DeviceRunnerService(new DeviceService());

devRunner.run();

The DeviceRunnerService is responsible for translating the environment configuration to the JSON configuration to be used to init the class DeviceService.

Usage in a Nestjs application

In your app module, import the device client module:

import { DeviceClientModule } from '@corvina/device-client/device.module';

@Module({
  imports: [DeviceClientModule],
  controllers: [AppController],
  providers: [AppService],
})
export class AppModule {}

In your main.ts you can start the device using the DeviceRunner service. This service reads the configuration from a .env file.

import { DeviceRunnerService } from '@corvina/device-client';

async function bootstrap() {
  const app = await NestFactory.create(AppModule);
  await app.listen(3000);

  app.get(DeviceRunnerService).run();
}

Sending data

The actual sending occurs according to the configuration received from the cloud. Posting a tag not configured from the cloud will result in an error.

In order to send data the DeviceClient class exposes the post method. The method allow to post a list of data points.

A data point is defined by:

• a tagName, representing the device identifier of the data point
• a timestamp, representing the UTC timestamp the value was originate
• a value

The post method accepts further options:

• qos, the MQTT QoS required
• cb, a function callback to catch sending errors or confirmations
• forceImmediateSend, to bypass the publish policies configured from the cloud and just send the data immediately
• recurseNotifyOnlyWholeObject, when posting a full JSON, just post the full object not every single path

Simulation

The device can be set up using environment variables to simulate data sending.

1. set the available tags to simulate (the list will also be published to the cloud):

  AVAILABLE_TAGS=[{"name":"Tag1","type":"integer"},{"name":"Tag2","type":"integer"},{"name":"Tag3","type":"integer"},{"name":"PositionNow","type":"integer"},{"name":"InputTag","type":"integer"}]

2. enable simulation:

  SIMULATE_TAGS=1

The same can be done for alarms:

1. set up the alarms to simulate:

  AVAILABLE_ALARMS=[{"name":"Alarm10","severity":1,"source":"Tag1","desc":{"en":"Tag above normal : [Tag1]"},"ack_required":true,"reset_required":true,"simulation":{"f":"{ return Math.random() > 0.5 }"}}]

2. enable simulation:

  SIMULATE_ALARMS=1

Receiving data

In order to receive data on a tag the tag must be configured writable in the cloud mapping

When data is written from the cloud a write event is emitted and can be used to handle the write request.

The event will report an object containing:

• modelPath: the written model path
• v: the written value

For example:

    app.get(DeviceService).on("write", (event) => {
        console.log("Write event received", event);
    });

Environment variables

See envs.md for a detailed description of environment variables.

Author

Arrigo Zanette

License

Licensed under the MIT License - see the LICENSE file for details.