exframe-cg

A simple client for deploying and interaction with choreographer models.

Usage

const cgClient = require('exframe-cg').cgClient({
  serviceUrl: 'http://choreographer.cg:9000',
  headerPrefix: 'harmony-',
  timeToRetry: 1000,
  maxRetryAttempts: 5
});

Configuration Settings

• serviceUrl - The Url and port on which choreographer is listening
• headerPrefix - Prefix to use for http headers
• timeToRetry - default: 1000 number of milliseconds to wait before retrying a model deployment upon failure
• maxRetryAttempts - default: 190 number of times to retry deploying a model before giving up

Client

deploy

Deploys a model to the choreographer.

The deploy method requires a model file path and returns a promise. If the deployment initially fails, the deploy method will retry maxRetryAttempts times before resolving or rejecting the promise.

cgClient.deploy(context, modelPath)
  .then(result => {
    console.log(`${modelPath} has been deployed`);
  })
  .catch(error => {
    console.log(error);
  });

Arguments:

• modelPath string The file path to the model

deployStub

Deploys a model stub to the choreographer.

The deployStub method requires a model file path and a model name. It returns a promise. If the deployment initially fails, the deployStub method will retry maxRetryAttempts times before resolving or rejecting the promise.

cgClient.deployStub(context, modelName, stubPath)
  .then(result => {
    console.log(`${modelName} stub at ${stubPath} has been deployed`);
  })
  .catch(error => {
    console.log(error);
  })

Arguments

• modelName string The name of the model being stubbed
• stubPath string The file path to the model stub

start

Starts a model.

The start method requires a modelName and an object that the model instance uses as input data. It returns a promise that resolves with information about the newly created model instance.

cgClient.start(context, modelName, args)
  .then(result => {
    console.log(`Instance ${result.modelInstanceId} started.`);
  })
  .catch(error => {
    console.log(error);
  })

Arguments

• modelName string The name of the model being started
• args object An object containing input data for the model

instance

Gets status information about a model instance.

The instance method requires a model instance id. It returns a promise that resolves with an object containing instance information.

cgClient.instance(context, modelInstanceId)
  .then(result => {
    console.log(`Model instance ${modelInstanceId} has ${result.completedTasks.length} completed tasks`);
  })
  .catch(error => {
    console.log(error);
  })

Arguments

• modelInstanceId string The instance id of the model of interest.

completeTask

Completes a user task.

The completeTask method requires a model instance id and a task name. It takes an object containing user input that the model instance may use to complete the task. It returns a promise that resolves with an object containing model instance information.

cgClient.complete(context, modelInstanceId, taskName, args)
  .then(result => {
    console.log(`Model instance ${modelInstanceId} has ${result.completedTasks.length} completed tasks`);
  })
  .catch(error => {
    console.log(error);
  })

Arguments

• modelInstanceId string The instance id of the model of interest
• taskName string The name of the task to completed
• args object An object containing input data to be consumed by the model instance.

moveToTask

Moves the execution of the model instance back to a specific user task.

The moveToTask method requires a model instance id and a task name. It returns a promises that resolves with an object containing model instance information.

cgClient.moveToTask(context, modelInstanceId, taskName)
.then(result => {
  console.log(`Model instance ${modelInstanceId} has ${result.completedTasks.length} completed tasks`);
})
.catch(error => {
  console.log(error);
})

Arguments

• modelInstanceId string The instance id of the model of interest
• taskName string The name of the task to move model execution to