gelf-http

Node module to send Graylog Information via HTTP. Because sometimes that's all you need.

configuration

{
	url: "http://localhost:8082/gelf", // URL to the Graylog Node
	gzip : false // Set to true if you want to save some bandwidth.
}

Functions

init(url[,defaultMessage])

Initializes the module with the given configuration

var gl2 = require("gelf-http");
gl2.init('http://localhost:8082/gelf');

Parameters

• (string or instanceof require('url')) url : URL of the HTTP Input on your Graylog server. gelf-http supports both HTTP and HTTPS

• (object) defaultMessage : default object that will be used as a base message for all the messages sent. Useful if you want to provide some fields with all your messages. Will be merged with the message parameter of send().

send(message, level [, callback])

Sends a message to the Graylog Server.

• message : Either a string or a object. If an object is given, it will be flattened to be correctly analyzed by Graylog.
• level : from 0 (panic) to 7 (debug), as described in the GELF specifications.
• callback : optional function(err, (boolean) success) fired on the result of the HTTP request.

Shorthands functions to send()

So you don't need to remember which level is which number :)

• debug(message [, callback])
• info(message [, callback])
• notice(message [, callback])
• warn(message [, callback])
• error(message [, callback])
• critical(message [, callback])
• alert(message [, callback])
• panic(message [, callback])

Metrics

gelf-http also provides two functions specifically used for sending metrics to Graylog.

• metric(name, value[, callback])
• metrics(values[, callback])

Examples

let gl2 = require('gelf-http');

gl2.init('http://localhost:8082/gelf');

/* Graylog will receive a basic DEBUG message with "This is a debug line" as message */
gl2.debug("This is a debug line");

/* Graylog will receive a INFO message with variables : 
    {
        "_cpu_load" : 0.509439832,
        "_ram_load" : 0.2984398,
        "_db_connections" :298
    }
*/
gl2.info({
    "cpu_load":0.509439832,
    "ram_load":0.2984398,
    "db_connections":298
});

/* Graylog will receive a INFO message with the following payload : 
    {
        "_label" : "cpu_load",
        "_value" : 50
    }
*/
gl2.metric("cpu_load",50);

Contribute

Issues

I use the module in some of my personal/professional projects, so any major bug would be detected by myself. Anyway, submit an issue if you encounter any trouble. I'll be glad to help. :)

Code

Fork the repository, do your modifications, and send a PR :) I don't intend to add much more features to this module, as it already does all I need. But if you have any idea, don't hesitate!

Run unit tests

Basic tests with only a mock server

$ npm install
$ mocha test

You have a real Graylog server with an HTTP input? Test the module with it too!

$ npm install
$ env GELF_HTTP_SERVER="url_of_your_http_input"
$ mocha test

Changelog

• 1.0 (2017-11-15) : Published to npmjs.com