Powered by GitBook

Data Collection

Data Import from Python Applications

The 'fluent-logger-python', library is used to post records from Python applications to Fluentd.

This article explains how to use the fluent-logger-python library.

Prerequisites

  • Basic knowledge of Python
  • Basic knowledge of Fluentd
  • Python 2.6 or higher

Installing Fluentd

Please refer to the following documents to install fluentd.

Modifying the Config File

Next, please configure Fluentd to use the forward Input plugin as its data source.

<source>
  type forward
  port 24224
</source>
<match fluentd.test.**>
  type stdout
</match>

Please restart your agent once these lines are in place.

# for rpm/deb only
$ sudo /etc/init.d/td-agent restart

Using fluent-logger-python

First, install the fluent-logger library via pip.

$ pip install fluent-logger

Next, initialize and post the records as shown below.

# test.py
from fluent import sender
from fluent import event
sender.setup('fluentd.test', host='localhost', port=24224)
event.Event('follow', {
  'from': 'userA',
  'to':   'userB'
  })

Executing the script will send the logs to Fluentd.

$ python test.py

The logs should be output to /var/log/td-agent/td-agent.log or stdout of the Fluentd process via the stdout Output plugin.

Production Deployments

Output Plugins

Various output plugins are available for writing records to other destinations:

High-Availablability Configurations of Fluentd

For high-traffic websites (more than 5 application nodes), we recommend using a high availability configuration of td-agent. This will improve data transfer reliability and query performance.

Monitoring

Monitoring Fluentd itself is also important. The article below describes general monitoring methods for td-agent.