#node-postgres

Non-blocking PostgreSQL client for node.js. Pure JavaScript and optional native libpq bindings.

Install

$ npm install pg

Intro & Examples

Simple example

var pg = require('pg');

// instantiate a new client
// the client will read connection information from
// the same environment varaibles used by postgres cli tools
var client = new pg.Client();

// connect to our PostgreSQL server instance
client.connect(function (err) {
  if (err) throw err;

  // execute a query on our database
  client.query('SELECT $1::text as name', ['brianc'], function (err, result) {
    if (err) throw err;

    // just print the result to the console
    console.log(result.rows[0]); // outputs: { name: 'brianc' }

    // disconnect the client
    client.end(function (err) {
      if (err) throw err;
    });
  });
});

Client pooling

If you're working on something like a web application which makes frequent queries you'll want to access the PostgreSQL server through a pool of clients. Why? There is ~20-30 millisecond delay (YMMV) when connecting a new client to the PostgreSQL server because of the startup handshake. Also, PostgreSQL can only support a limited number of clients...it depends on the amount of ram on your database server, but generally more than 100 clients at a time is a bad thing. ™️ Finally PostgreSQL can only execute 1 query at a time per connected client.

With that in mind we can imagine a situtation where you have a web server which connects and disconnects a new client for every web request or every query (don't do this!). If you get only 1 request at a time everything will seem to work fine, though it will be a touch slower due to the connection overhead. Once you get >500 simultaneous requests your web server will attempt to open 500 connections to the PostgreSQL backend and 💥 you'll run out of memory on the PostgreSQL server, it will become unresponsive, your app will seem to hang, and everything will break. Boooo!

Good news: node-postgres ships with built in client pooling.

var pg = require('pg');

// create a config to configure both pooling behavior
// and client options
// note: all config is optional and the environment variables
// will be read if the config is not present
var config = {
  user: 'foo', //env var: PGUSER
  database: 'my_db', //env var: PGDATABASE
  password: 'secret', //env var: PGPASSWORD
  port: 5432, //env var: PGPORT
  max: 10, // max number of clients in the pool
  idleTimeoutMillis: 30000, // how long a client is allowed to remain idle before being closed
};

var pool = new pg.Pool(config);

//this initializes a connection pool
//it will keep idle connections open for a 30 seconds
//and set a limit of maximum 10 idle clients
pool.connect(function(err, client, done) {
  if(err) {
    return console.error('error fetching client from pool', err);
  }
  client.query('SELECT $1::int AS number', ['1'], function(err, result) {
    //call `done()` to release the client back to the pool
    done();

    if(err) {
      return console.error('error running query', err);
    }
    console.log(result.rows[0].number);
    //output: 1
  });
});

node-postgres uses pg-pool to manage pooling and includes it and exports it for convienience. If you want, you can require('pg-pool') and use it directly - its the same as the constructor exported at pg.Pool.

It's highly recommend you read the documentation for pg-pool.

Here is an up & running quickly example

More Documentation

Native Bindings

To install the native bindings:

$ npm install pg pg-native

node-postgres contains a pure JavaScript protocol implementation which is quite fast, but you can optionally use native bindings for a 20-30% increase in parsing speed (YMMV). Both versions are adequate for production workloads. I personally use the pure JavaScript implementation because I like knowing whats going on all the way down to the binary on the socket, and it allows for some fancier use cases which are difficult to do with libpq. 😄

To use the native bindings, first install pg-native. Once pg-native is installed, simply replace var pg = require('pg') with var pg = require('pg').native. Make sure any exported constructors from pg are from the native instance. Example:

var pg = require('pg').native
var Pool = require('pg').Pool // bad! this is not bound to the native client
var Client = require('pg').Client // bad! this is the pure JavaScript client

var pg = require('pg').native
var Pool = pg.Pool // good! a pool bound to the native client
var Client = pg.Client // good! this client uses libpq bindings

node-postgres abstracts over the pg-native module to provide exactly the same interface as the pure JavaScript version. Care has been taken to keep the number of api differences between the two modules to a minimum; however, it is recommend you use either the pure JavaScript or native bindings in both development and production and don't mix & match them in the same process - it can get confusing!

Features

• pure JavaScript client and native libpq bindings share the same api
• connection pooling
• extensible js<->postgresql data-type coercion
• supported PostgreSQL features
  • parameterized queries
  • named statements with query plan caching
  • async notifications with LISTEN/NOTIFY
  • bulk import & export with COPY TO/COPY FROM

Extras

node-postgres is by design pretty light on abstractions. These are some handy modules we've been using over the years to complete the picture. Entire list can be found on wiki

Contributing

❤️ contributions!

If you need help getting the tests running locally or have any questions about the code when working on a patch please feel free to email me or gchat me.

I will happily accept your pull request if it:

• has tests
• looks reasonable
• does not break backwards compatibility

Information about the testing processes is in the wiki.

Open source belongs to all of us, and we're all invited to participate!

Support

If at all possible when you open an issue please provide

• version of node
• version of postgres
• smallest possible snippet of code to reproduce the problem

Usually I'll pop the code into the repo as a test. Hopefully the test fails. Then I make the test pass. Then everyone's happy!

If you need help or run into any issues getting node-postgres to work on your system please report a bug or contact me directly. I am usually available via google-talk at my github account public email address. Remember this is a labor of love, and though I try to get back to everything sometimes life takes priority, and I might take a while. It helps if you use nice code formatting in your issue, search for existing answers before posting, and come back and close out the issue if you figure out a solution. The easier you can make it for me, the quicker I'll try and respond to you!

If you need deeper support, have application specific questions, would like to sponsor development, or want consulting around node & postgres please send me an email, I'm always happy to discuss!

I usually tweet about any important status updates or changes to node-postgres on twitter. Follow me @briancarlson to keep up to date.

License

Copyright (c) 2010-2016 Brian Carlson (brian.m.carlson@gmail.com)

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.