Varkes

Overview

Varkes (Greek for “small boats”) is a framework which mocks applications. This "small boat" allows you to mock even a complex application (a "big ship") and develop your functionality in an easy way. Instead of setting up and deploying an often complex application, you simply pair the mock with Kyma to develop and experiment with it. When you are done, simply switch the "ships" to connect to and work with the target application.

Varkes framework provides the following node modules to make the application mocks work:

API Server serves an API to pair with a Kyma cluster to register Events and APIs and send Events. For details, see API Server.
App Connector is a library used by the API-Server covering all communication to a Kyma cluster. For details, see App Connector.
Cockpit a UI of the API server for a convenient way of management. For details, see Cockpit.
OpenAPI Mock which mocks the application APIs based on the OpenAPI specification. For details, see OpenAPI mock.
OData Mock which mocks the application APIs based on the OData specification. For details, see OData mock.
Configuration which manages loading and validation of Varkes configuration files, see Configuration.

Architecture

The diagram shows how an application mock integrates with a Kyma cluster and which functionality is taken over by which framework modules:

Examples & Usage

Use the modules as express middleware in your express based NodeJs application. Depending on the modules you require, add the dependencies to your package.json like that:

  "dependencies": {
    "@varkes/configuration": "*",
    "@varkes/api-server": "*",
    "@varkes/odata-mock": "*",
    "@varkes/openapi-mock": "*",
    "@varkes/cockpit": "*"
  }

In your application enable the modules in your express app like that:

const config = require("@varkes/configuration")
const odataMock = require("@varkes/odata-mock")
const openapiMock = require("@varkes/openapi-mock")
const server = require("@varkes/api-server")
const cockpit = require("@varkes/cockpit")
const app = require('express')()

let configuration = await config.resolveFile("./varkes_config.json")
app.use(await odataMock.init(configuration))
app.use(await openapiMock.init(configuration))
app.use(await server.init(configuration))
app.use(await cockpit.init(configuration))

app.listen(10000, function () {
       console.info("Started application on port %d", 10000)
})

Furthermore a configuration is required, specified in the ./varkes_config.json file. The simplest configuration is {}. For examples on configurations look at the examples in the examples folder. They include different mocks based on OpenAPI and OData packages.

A reference documentation on the configuration you can find at the Type definition

NOTE: Using examples in the master branch without cloning the repository is currently impossible because of the development version numbers used in the package.json files which are not published to the npm registry. If you want to use the examples without cloning the whole repository, use the tags in the release branch. There you can find the latest version of the dependencies already included in the package.json files.

Development

This project is maintained with Lerna. To start developing perform these steps:

If you don't have lerna, install it by running npm install -g lerna.
Clone this repo and run make resolve in the project root to install dependencies and link local dependencies.
Introduce your changes and commit freely. There is no need to update the package versions in dependent subprojects.

To increase the version number, run lerna version --no-git-tag-version, stating the new version number. When omitting the flag, it also creates a new git tag with the given version number. This command also updates the dependency version in the package.json of subprojects.

To see how CI operates on Lerna, check the makefile in the root folder.

NOTE: To create Docker images of examples, use the release branch, where the npm installed in Docker containers can get the latest version from the registry.