Feathers service adapters for managing MongoDB databases, users and collections
With NPM npm install feathers-mongodb-management --save
or Yarn yarn add feathers-mongodb-management
v1.x is expected to work with Feathers v3 (a.k.a Buzzard) and NodeJS v12
v2.x is expected to work with Feathers v5 (a.k.a Dove) and NodeJS v16
If a large number of users require it a more complete feathers-mongodb-management documentation will be created. For now as the API is pretty straigth forward the example and tests should be sufficient to understand how it works.
The initial use case was to simplify data segregation for a SaaS application where each user could belong to different organisations like in GitHub. Indeed, one key aspect of access control is to filter the resources a user can reach. In this specific use case the users always work in the context of an organisation acting as a filter applied on the manipulated resources. This issue is often tackled by applying a client-side filter on a global service like /users?org=orgId
and storing on each resouce the organisation it belongs to. We consider this approach as more harder to maintain in the long-term due to additional filtering and relations in the data model. As a consequence we created this plugin to "physically" separate the data of each organisation in a dedicated DB instead of doing this "logically" using filtering.
You can find more information by reading this article.
Here's an example of a Feathers server that uses feathers-mongodb-management
.
const feathers = require('feathers');
const rest = require('feathers-rest');
const hooks = require('feathers-hooks');
const bodyParser = require('body-parser');
const errorHandler = require('feathers-errors/handler');
const mongodb = require('mongodb');
const plugin = require('feathers-mongodb-management');
// Initialize the application
const app = feathers()
.configure(rest())
.configure(hooks())
// Needed for parsing bodies (login)
.use(bodyParser.json())
.use(bodyParser.urlencoded({ extended: true }))
.use(errorHandler());
// Connect to Mongo instance
const client = await mongodb.MongoClient.connect('mongodb://127.0.0.1:27017')
// Initialize your feathers plugin to manage databases
app.use('/mongo/databases', plugin.database({ adminDb: client.db('feathers-test').admin(), client }));
let dbService = app.service('/mongo/databases');
// Now create a new database
let db = await dbService.create({ name: 'test-db' })
// The objects provided through the plugin services are just metadata and not MongoDB driver instances
// We need to retrieve it to create collection/user services that require the DB instance
db = client.db('test-db');
// Now create services binded to this database to manage collections/users
app.use('/mongo/test-db/collections', plugin.collection({ db }));
let collectionService = app.service('/mongo/test-db/collections');
const collection = await collectionService.create({ name: 'test-collection' })
app.use('/mongo/test-db/users', plugin.user({ db }));
let userService = app.service('/mongo/test-db/users');
const user = await userService.create({ name: 'test-user', password: 'test-password', roles: ['readWrite'] })
// Perform other operations using these services if required
...
// Then start the app
app.listen(3030);
console.log('Feathers app started on 127.0.0.1:3030');
Copyright (c) 2016
Licensed under the MIT license.