MongoDB Native: The Official Node.js MongoDB driver

The Node.js MongoDB driver was started in 2009 and has grown to become one of the most popular database drivers for the Node.js Platform.

MongoDB is a document database that allows the storage of documents in json format. It’s main benefits come from it’s ability to store unstructured data while maintaining traditional databases ability to query and index the document efficiently. MongoDb also provides a whole scaling up path.

A whole ecosystem has sprung out on top of the driver including ODM mappers such as Mongoose and whole frameworks such as Meteor (a fantastic real time framework). The driver is constantly evolving and there are some future spin-off projects planned around the driver.

what where

Bugs / Feature Requests

Think you’ve found a bug? Want to see a new feature in node-mongodb-native? Please open a case in our issue management tool, JIRA:

Bug reports in JIRA for all driver projects (i.e. NODE, PYTHON, CSHARP, JAVA) and the Core Server (i.e. SERVER) project are public.

Questions and Bug Reports

Change Log


To install the most recent release from npm, run:

npm install mongodb

That may give you a warning telling you that bugs[‘web’] should be bugs[‘url’], it would be safe to ignore it (this has been fixed in the development version)

To install the latest from the repository, run::

npm install path/to/node-mongodb-native

Live Examples


This is a node.js driver for MongoDB. It’s a port (or close to a port) of the library for ruby at

A simple example of inserting a document.

  var MongoClient = require('mongodb').MongoClient
    , format = require('util').format;

  MongoClient.connect('mongodb://', function(err, db) {
    if(err) throw err;

    var collection = db.collection('test_insert');
    collection.insert({a:2}, function(err, docs) {
      collection.count(function(err, count) {
        console.log(format("count = %s", count));

      // Locate all the entries using find
      collection.find().toArray(function(err, results) {
        // Let's close the db

Data types

To store and retrieve the non-JSON MongoDb primitives (ObjectID, Long, Binary, Timestamp, DBRef, Code).

In particular, every document has a unique _id which can be almost any type, and by default a 12-byte ObjectID is created. ObjectIDs can be represented as 24-digit hexadecimal strings, but you must convert the string back into an ObjectID before you can use it in the database. For example:

  // Get the objectID type
  var ObjectID = require('mongodb').ObjectID;

  var idString = '4e4e1638c85e808431000003';
  collection.findOne({_id: new ObjectID(idString)}, console.log)  // ok
  collection.findOne({_id: idString}, console.log)  // wrong! callback gets undefined

Here are the constructors the non-Javascript BSON primitive types:

  // Fetch the library
  var mongo = require('mongodb');
  // Create new instances of BSON types
  new mongo.Long(numberString)
  new mongo.ObjectID(hexString)
  new mongo.Timestamp()  // the actual unique number is generated on insert.
  new mongo.DBRef(collectionName, id, dbName)
  new mongo.Binary(buffer)  // takes a string or Buffer
  new mongo.Code(code, [context])
  new mongo.Symbol(string)
  new mongo.MinKey()
  new mongo.MaxKey()
  new mongo.Double(number)  // Force double storage

The C/C++ bson parser/serializer

If you are running a version of this library has the C/C++ parser compiled, to enable the driver to use the C/C++ bson parser pass it the option native_parser:true like below

  // using native_parser:
    , {db: {native_parser: true}}, function(err, db) {})

The C++ parser uses the js objects both for serialization and deserialization.

GitHub information

The source code is available at You can either clone the repository or download a tarball of the latest release.

Once you have the source you can test the driver by running

$ make test

in the main directory. You will need to have a mongo instance running on localhost for the integration tests to pass.


For examples look in the examples/ directory. You can execute the examples using node.

$ cd examples
$ node queries.js


The GridStore class allows for storage of binary files in mongoDB using the mongoDB defined files and chunks collection definition.

For more information have a look at Gridstore


For more information about how to connect to a replicaset have a look at the extensive documentation Documentation

Primary Key Factories

Defining your own primary key factory allows you to generate your own series of id’s (this could f.ex be to use something like ISBN numbers). The generated the id needs to be a 12 byte long “string”.

Simple example below

  var MongoClient = require('mongodb').MongoClient
    , format = require('util').format;    

  // Custom factory (need to provide a 12 byte array);
  CustomPKFactory = function() {}
  CustomPKFactory.prototype = new Object();
  CustomPKFactory.createPk = function() {
    return new ObjectID("aaaaaaaaaaaa");

  MongoClient.connect('mongodb://', {'pkFactory':CustomPKFactory}, function(err, db) {
    if(err) throw err;

    db.dropDatabase(function(err, done) {
      db.createCollection('test_custom_key', function(err, collection) {
        collection.insert({'a':1}, function(err, docs) {
          collection.find({'_id':new ObjectID("aaaaaaaaaaaa")}).toArray(function(err, items) {
            // Let's close the db


If this document doesn’t answer your questions, see the source of Collection or Cursor, or the documentation at MongoDB for query and update formats.


The find method is actually a factory method to create Cursor objects. A Cursor lazily uses the connection the first time you call nextObject, each, or toArray.

The basic operation on a cursor is the nextObject method that fetches the next matching document from the database. The convenience methods each and toArray call nextObject until the cursor is exhausted.


  var cursor = collection.find(query, [fields], options);

  cursor.nextObject(function(err, doc) {});
  cursor.each(function(err, doc) {});
  cursor.toArray(function(err, docs) {});

  cursor.rewind()  // reset the cursor to its initial state.

Useful chainable methods of cursor. These can optionally be options of find instead of method calls:

  • .limit(n).skip(m) to control paging.
  • .sort(fields) Order by the given fields. There are several equivalent syntaxes:
  • .sort({field1: -1, field2: 1}) descending by field1, then ascending by field2.
  • .sort([['field1', 'desc'], ['field2', 'asc']]) same as above
  • .sort([['field1', 'desc'], 'field2']) same as above
  • .sort('field1') ascending by field1

Other options of find:

  • fields the fields to fetch (to avoid transferring the entire document)
  • tailable if true, makes the cursor tailable.
  • batchSize The number of the subset of results to request the database to return for every request. This should initially be greater than 1 otherwise the database will automatically close the cursor. The batch size can be set to 1 with batchSize(n, function(err){}) after performing the initial query to the database.
  • hint See Optimization: hint.
  • explain turns this into an explain query. You can also call explain() on any cursor to fetch the explanation.
  • snapshot prevents documents that are updated while the query is active from being returned multiple times. See more details about query snapshots.
  • timeout if false, asks MongoDb not to time out this cursor after an inactivity period.

For information on how to create queries, see the MongoDB section on querying.

  var MongoClient = require('mongodb').MongoClient
    , format = require('util').format;    

  MongoClient.connect('mongodb://', function(err, db) {
    if(err) throw err;

    var collection = db
      .toArray(function(err, docs) {



  collection.insert(docs, options, [callback]);

where docs can be a single document or an array of documents.

Useful options:

  • safe:true Should always set if you have a callback.

See also: MongoDB docs for insert.

  var MongoClient = require('mongodb').MongoClient
    , format = require('util').format;    

  MongoClient.connect('mongodb://', function(err, db) {
    if(err) throw err;
    db.collection('test').insert({hello: 'world'}, {w:1}, function(err, objects) {
      if (err) console.warn(err.message);
      if (err && err.message.indexOf('E11000 ') !== -1) {
        // this _id was already inserted in the database

Note that there’s no reason to pass a callback to the insert or update commands unless you use the safe:true option. If you don’t specify safe:true, then your callback will be called immediately.

Update: update and insert (upsert)

The update operation will update the first document that matches your query (or all documents that match if you use multi:true). If safe:true, upsert is not set, and no documents match, your callback will return 0 documents updated.

See the MongoDB docs for the modifier ($inc, $set, $push, etc.) formats.


  collection.update(criteria, objNew, options, [callback]);

Useful options:

  • safe:true Should always set if you have a callback.
  • multi:true If set, all matching documents are updated, not just the first.
  • upsert:true Atomically inserts the document if no documents matched.

Example for update:

  var MongoClient = require('mongodb').MongoClient
    , format = require('util').format;    

  MongoClient.connect('mongodb://', function(err, db) {
    if(err) throw err;

    db.collection('test').update({hi: 'here'}, {$set: {hi: 'there'}}, {w:1}, function(err) {
      if (err) console.warn(err.message);
      else console.log('successfully updated');

Find and modify

findAndModify is like update, but it also gives the updated document to your callback. But there are a few key differences between findAndModify and update:

  1. The signatures differ.
  2. You can only findAndModify a single item, not multiple items.


    collection.findAndModify(query, sort, update, options, callback)

The sort parameter is used to specify which object to operate on, if more than one document matches. It takes the same format as the cursor sort (see Connection.find above).

See the MongoDB docs for findAndModify for more details.

Useful options:

  • remove:true set to a true to remove the object before returning
  • new:true set to true if you want to return the modified object rather than the original. Ignored for remove.
  • upsert:true Atomically inserts the document if no documents matched.

Example for findAndModify:

  var MongoClient = require('mongodb').MongoClient
    , format = require('util').format;    

  MongoClient.connect('mongodb://', function(err, db) {
    if(err) throw err;
    db.collection('test').findAndModify({hello: 'world'}, [['_id','asc']], {$set: {hi: 'there'}}, {}, function(err, object) {
      if (err) console.warn(err.message);
      else console.dir(object);  // undefined if no matching object exists.


The save method is a shorthand for upsert if the document contains an _id, or an insert if there is no _id.

Release Notes



  1. 10gen
  2. Google Closure Library
  3. Jonas Raoni Soares Silva


Aaron Heckmann, Christoph Pojer, Pau Ramon Revilla, Nathan White, Emmerman, Seth LaForge, Boris Filipov, Stefan Schärmeli, Tedde Lundgren, renctan, Sergey Ukustov, Ciaran Jessup, kuno, srimonti, Erik Abele, Pratik Daga, Slobodan Utvic, Kristina Chodorow, Yonathan Randolph, Brian Noguchi, Sam Epstein, James Harrison Fisher, Vladimir Dronnikov, Ben Hockey, Henrik Johansson, Simon Weare, Alex Gorbatchev, Shimon Doodkin, Kyle Mueller, Eran Hammer-Lahav, Marcin Ciszak, François de Metz, Vinay Pulim, nstielau, Adam Wiggins, entrinzikyl, Jeremy Selier, Ian Millington, Public Keating, andrewjstone, Christopher Stott, Corey Jewett, brettkiefer, Rob Holland, Senmiao Liu, heroic, gitfy


Copyright 2009 - 2013 MongoDb Inc.

Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.