Connection pool used by any-db

any-db-pool - database agnostic connection pool

Build Status


var anyDB = require('any-db')

var pool = anyDB.createPool('postgres://user:pass@localhost/dbname', {
  min: 5, max: 15,
  reset: function (conn, done) {
    conn.query('ROLLBACK', done)

// Proxies to mysql's connection.query
var q = pool.query('SELECT 1', function (err, res) { })

Note: As shown above, ConnectionPool instances are usually created with anyDB.createPool. The any-db package will be installed alongside any adapters (e.g. any-db-postgres), so most users should depend on their adapter and not on any-db or any-db-pool.


This package contains a database connection pool that can be used with any driver, but it requires an any-db compliant adapter. If you are writing a library that needs to support multiple database backends (e.g. SQLite3 or Postgres or MySQL) then it's strongly recommended that you add any-db toyour peerDependencies and rely on createPool instead of depending on this package directly.


module.exports := (Adapter, adapterConfig: Object, PoolConfig) => ConnectionPool

ConnectionPool := EventEmitter & {
  adapter: String,
  query:   (String, Array?, Continuation<ResultSet>?) => Query,
  acquire: (Continuation<Connection>) => void,
  release: (Connection) => void,
  close:   (Continuation<void>?) => void,

PoolConfig := {
  min: Number?,
  max: Number?,
  idleTimeout: Number?,
  reapInterval: Number?,
  refreshIdle: Boolean?,
  onConnect: (Connection, ready: Continuation<Connection>) => void
  reset: (Connection, done: Continuation<void>) => void
  shouldDestroyConnection: (error: Error) => Boolean


A PoolConfig is generally a plain object with any of the following properties (they are all optional):

  • min (default 0) The minimum number of connections to keep open in the pool.
  • max (default 10) The maximum number of connections to keep open in the pool. When this limit is reached further requests for connections will queue waiting for an existing connection to be released back into the pool.
  • refreshIdle (default true) When this is true, the pool will reap connections that have been idle for more than idleTimeout milliseconds.
  • idleTimeout (default 30000) The maximum amount of time a connection can sit idle in the pool before being reaped.
  • reapInterval (default 1000) How frequently the pool should check for connections that are old enough to be reaped.
  • onConnect Called immediately after a connection is first established. Use this to do one-time setup of new connections. The supplied Connection will not be added to the pool until you pass it to the done continuation.
  • reset Called each time a connection is returned to the pool. Use this to restore a connection to it's original state (e.g. rollback transactions, set the database session vars). If reset fails to call the done continuation the connection will be lost in limbo.
  • shouldDestroyConnection (default function (err) { return true }) - Called when an error is encountered by pool.query or emitted by an idle connection. If shouldDestroyConnection(error) is truthy the connection will be destroyed, otherwise it will be reset.


(String, Array?, Continuation<ResultSet>?) => Query

Implements Queryable.query by automatically acquiring a connection and releasing it when the query completes.


(Continuation<Connection>) => void

Remove a connection from the pool. If you use this method you must return the connection back to the pool using ConnectionPool.release


(Connection) => void

Return a connection to the pool. This should only be called with connections you've manually acquired. You must not continue to use the connection after releasing it.


(Continuation<void>?) => void

Stop giving out new connections, and close all existing database connections as they are returned to the pool.


The string name of the adapter used for this connection pool, e.g. 'sqlite3'.

ConnectionPool events

Acquire event

An 'acquire' event is emitted by a ConnectionPool whenever the pool's acquire() method is invoked.

No arguments are passed to event listeners.

Release event

A 'release' event is emitted by a ConnectionPool whenever the pool's release() method is invoked.

No arguments are passed to event listeners.

Query event

A 'query' event is emitted by a ConnectionPool immediately after the pool's query() method is invoked.

One argument is passed to event listeners:

  • query - a Query object.

Close event

A 'close' event is emitted by a ConnectionPool when the pool has closed all of it's connections. Invoking a pool's close() method would cause a close event to be emitted.

No arguments are passed to event listeners.

Why wouldn't I just use generic-pool?

generic-pool is awesome, but it's very generic. This is a Good Thing for a library with "generic" in the name, but not so good for the very common but slightly more specialized case of pooling stateful SQL database connections. This library uses generic-pool and simply augments it with some added niceties:

  • Hooks for initializing and/or resetting connection state when connections are added or returned to the pool.
  • A query method that allows queries to be performed without the user needing a reference to a connection object (and potentially leaking that reference).

Stop telling me not to use this directly

Ok, if you really want to use this package without using the any-db frontend you should provide a compliant Adapter implementation:

var ConnectionPool = require('any-db-pool')
var adapter = require('my-custom-adapter')
var connectionParams = { user: 'scott', password: 'tiger' }
var poolParams = {
  min: 5, max: 15,
  reset: function (conn, done) {
    conn.query('ROLLBACK', done)
var pool = new ConnectionPool(adapter, connectionParams, poolParams)

However, it would be awesome if you just published your adapter as a package named any-db-$name so that everybody could use it 👍