node.dbslayer.js

node.dbslayer.js is a very basic and easy-to-use library to connect to a DBSlayer server, which effectively provides non-blocking and scalable MySQL support for Node.JS.

DBSlayer benefits include:

• It's Node.JS/V8/JavaScript friendly, since the the messages are delivered in JSON format over HTTP.

• Developed by the New York Times, it's designed with scalability in mind, doing connection pooling for you. This is what makes DBSlayer arguably better than implementing an async MySQL client directly into Node (through mysac for example).

Forked version enhancements/modifications

• returning results as JSON objects with column names as attributes
• supports query parsing with ? placeholder
• uses addListener instead of addCallback

Requirements

• Node.js (tested with v0.5.0-pre)
• DBSlayer (tested with beta-12)

How to Use

From your node.js script, require the dbslayer package

var db = require('dbslayer');

Initialize a connection

var connection = db.Server('localhost', 9090);

and then perform a query:

connection.query("SELECT * FROM table");

To be truly non-blocking, Server::fetch has to return a promise and not the result immediately. This means that in order to be able to perform queries in a designated order or access the result, you'll have to use callbacks:

var q = connection.query("SELECT * FROM TABLE");
q.addListener("success", function(results){
	for (var i = 0, l = results.length; i < l; i++){
		var row = results[i];
		// do something with the data
	}
});

This forked version supports query parsing. ? characters can be used as placeholders for variable length arguments that will be safely escaped before sending the final query:

connection.query("INSERT INTO TABLE SET col1 = ?, col2 = ?, col3 = ?", integer_value, "string value", json_object)

This forked version's results are returned as JSON objects. Each rows' values are stored as attributes:

connection.query("SELECT * FROM TABLE WHERE col1 = ?", integer_value).addListener("success", function(results) {
	for(i in results) {
		console.log(results[i].col1, results[i].col2, results[i].col3);
		// or results[i]["col1"]
	}
});

If you want to capture MySQL errors, subscribe to the 'error' event

var q = connection.query("SELECT * FROM NONEXISTENT_TABLE");
q.addListener("error", function(error, errno, request_obj) {
	console.log("error:", error);
	console.log("MySQL error number:", errno);
	console.log("SQL query that caused error:", request_obj.SQL);
});

Aside from query, the commands stat, client_info, host_info, server_version and client_version are available, which provide the respective information about the server.

Installing DBSlayer

Compile it according to the instructions here.

Then create a /etc/dbslayer.conf file defining a database. Here I'm defining the cool server which connects to my mysql database

[cool]
database=mysql
host=localhost
user=root
pass=1234

Then run DBSlayer for that connection:

dbslayer -c /etc/dbslayer.conf -s cool

Test it by running test.js like this:

node test.js "SELECT * FROM help_category"

If you get a bunch of entries like in this screenshot then dbslayer (and node.dbslayer.js) work!

Author

Guillermo Rauch <http://devthought.com>