winston-lumberjack

A logstash transport for winston based on node-lumberjack-protocol, a lumberjack protocol implementation for Node.js. It allows for sending encrypted logs from your node.js app to your logstash server instance.

Usage

Node

var winston = require('winston');
// Simply requiring winston-lumberjack will assign winston.transports.Lumberjack
require('winston-lumberjack');

winston.add(winston.transports.Lumberjack, {
    "timestamp": true,
    "level": "debug",
    "serverAddress": {
        "host": "localhost",
        "port": 5000
    },
    "sslCrt": "./logstash-forwarder.crt"
});

winston.log('debug', 'Now my debug messages are written to logstash!');

Lumberjack Configuration

• sslCert (optional)

  Filename of the trusted CA certificate which has signed the logstash server certificate. If this is omitted several well known "root" CAs will be used. These are used to authorize connections.

• serverAddress

  An object to specify the host and port of the logstash server. Additionally, all options supported by tls.connect() may be provided. If the ca property is the content of the file referenced by the sslCert property (if provided) will be set as ca.

  "serverAddress": { "host": "localhost", "port": 5000 }

• level (optional)

  Level of messages this transport should log. Defaults to 'info'.

• timestamp (optional)

  Boolean flag indicating if log messages shall include a timestamp. Defaults to false.

• clientOptions (optional)

  An object with additional configuration option passed on to the lumberjack protocol stack:

  • windowSize - the windowSize to send to the receiver. Defaults to 1000. See Caveats section of node-lumberjack-protocol README for details.

  • maxQueueSize - the maximum number of messages to queue while disconnected. If this limit is hit, all messages in the queue will be filtered with allowDrop(data). Only messages which this function returns true for will be removed from the queue. If there are still too many messages in the queue at this point the the oldest messages will be dropped. Defaults to 500.

  • allowDrop(data) - this will be called when deciding which messages to drop. By dropping lower priority messages (info and debug level messages, for example) you can increase the chances of higher priority messages getting through when the Client is having connection issues, or if the receiver goes down for a short period of time. This function is used both to drop messages from the queue while disconnected, and to drop messages if the receiver is taking too long to acknowledge messages.

    Note that this function will be called on all messages in the queue every time the queue grows too large - if this function does not return true for any messages, then it could be called for every message in the queue every time a message is queued.

  • options.reconnect - time, in ms, to wait between reconnect attempts. Defaults to 3 seconds.

License

Copyright (c) 2015-2018, Marcus Wittig All rights reserved.

MIT