Skip to content
Switch branches/tags
Go to file

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time


Build Status

Feathers service for Google Datastore

Note: this service is still in its early stages and may be buggy and change quickly. Feedback is welcome!


npm install feathers-datastore --save


This service follows the standard interface followed by Feathers database services, barring certain caveats as outlined below. Please refer to the feathers-datastore documentation for Feathers service interface.


Proper pagination, sorting and certain query filters are not yet implemented. You cannot use: $in, $nin, $ne, $or. Some of these will be implemented in future, but may not be due to the limitations of Google Datastore querying.

It's also worth noting that unless you index your datastore properly, you may run into issues running many combined filters. See querying Google Datastore.

There is also a performance / cost concern with $select - as Google Datastore cannot project unindexed properties, this library fetches the entire entity from the database and then filters it out before sending the response.


If you're wishing to manually authenticate using credentials, pass a credentials object to the service factory function e.g.

let credentials = {
  client_email: process.env.GCLOUD_CLIENT_EMAIL,
  private_key: process.env.GCLOUD_PRIVATE_KEY

app.use('/user', datastore({ credentials }))

Note: as with Google Cloud, this isn't needed if you're running in an environment which authenticates for you e.g. App Engine or a local machine with the gcloud SDK


Each service instance should be constructed with a kind for Datastore to use e.g.

app.use('/user', datastore({ kind: 'User' }));

This sets this.kind to User.


To namespace any operation on the datastore, add it to the query params e.g.

    query: {
      namespace: 'Marketing'

Datastore Keys

The datastore service has a makeKey method which is called to convert a resource id into a Datastore Entity key. By default it makes a Key with path [ this.kind, id ] and namespace as per the namespace query property.

This method can be overridden in order to serialize / deserialize keys from the ID.

Finding descendants

To specify an ancestor in a query, add it to the query params e.g.

    query: {
      ancestor: 'John'
  .then((users) => {


Datastore will error if entity properties are > 1500 bytes and are set to index, so it's worth defining indexing if you can.

Properties can be excluded from indexing by setting the query param dontIndex e.g.

    name: 'Stacey'
    age: 24,
    description: '...'
  }, {
    query: {
      dontIndex: [ 'description' ]

feathers-datastore can also automatically decide on whether or not to index based on the size of the entity. To enable, set autoIndex to true e.g.

    name: 'Stacey'
    age: 24,
    description: 'Some massive description...'
  }, {
    query: {
      autoIndex: true

in which case if for example description is larger than 1500 bytes, it won't be indexed.

Complete Example

Here's an example of a Feathers server that uses feathers-datastore.

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 datastore = require('feathers-datastore');

const kind = 'User';
const projectId = process.env.GCLOUD_PROJECT;

// Initialize the application
const app = feathers()
  // Needed for parsing bodies (login)
  .use(bodyParser.urlencoded({ extended: true }))
  // Initialize your feathers datastore
  .use('/user', datastore({ kind, projectId }))


console.log('Feathers app started on');


Copyright (c) 2016

Licensed under the MIT license.


No description, website, or topics provided.




No packages published