How to Parse Syslog Messages

Search…

Syslog is a popular protocol that virtually runs on every server. It is used to collect all kinds of logs. The problem with syslog is that services have a wide range of log formats, and no single parser can parse all syslog messages effectively.
In this tutorial, we will show how to use Fluentd to filter and parse different syslog messages robustly.
Prerequisites
A basic understanding of Fluentd
A running instance of rsyslogd
In this guide, we assume you are running td-agent on Ubuntu.
Setting Up rsyslogd
Open /etc/rsyslogd.conf and append the following line:
1
*.* @127.0.0.1:5140
Copied!
Then restart the rsyslogd service:
1
$ sudo systemctl restart syslog
Copied!
This tells rsyslogd to forward logs to port 5140 to which Fluentd will listen.
Setting up Fluentd
In this section, we will evolve our Fluentd configuration step-by-step.
Step 1: Listening to syslog Messages
First, let's configure Fluentd to listen to syslog messages.
Open /etc/td-agent/td-agent.conf and put the following configuration:
1
<source>
2
  @type syslog
3
  port 5140
4
  tag system
5
</source>
6
​
7
<match system.**>
8
  @type stdout
9
</match>
Copied!
This is the most basic setup: it listens to all syslog messages and outputs them to the standard output.
Now please restart td-agent:
1
$ sudo systemctl restart td-agent
Copied!
Let's confirm data is coming in:
1
$ less /var/log/td-agent/td-agent.log
Copied!
Step 2: Extract syslog Messages from sudo
Now, let's look at a sudo message like this one:
1
2018-09-27 16:00:01.000000000 +0900 system.authpriv.info: {"host":"localhost",
2
"ident":"sudo","message":"pam_unix(sudo:session): session opened for user root
3
by admin(uid=0)"}
Copied!
For security reasons, it is worth knowing which user performed what using sudo. In order to do so, we need to parse the message field. In other words, we need to extract syslog messages from sudo and handle them differently.
For this purpose, we can use the grep filter plugin. It examines the fields of events, and filter them based on regular expression patterns. In the following example, Fluentd filters out events that come from sudo and contain command data:
1
<source>
2
  @type syslog
3
  port 42185
4
  tag system
5
</source>
6
​
7
<filter system.**>
8
  @type grep
9
  <regexp>
10
    key ident
11
    pattern /^sudo$/
12
  </regexp>
13
  <regexp>
14
    key message
15
    pattern /COMMAND/
16
  </regexp>
17
</filter>
18
​
19
<match system.**>
20
  @type stdout
21
</match>
Copied!
Step 3: Extract Information from Messages
Now let's extract some information from syslog messages. For this purpose, we use another plugin called filter-parser. With this plugin, you can parse the content of a field using a regular expression.
Here is the final configuration:
1
<source>
2
  @type syslog
3
  port 5140
4
  tag system
5
</source>
6
​
7
<filter system.**>
8
  @type grep
9
  <regexp>
10
    key ident
11
    pattern /^sudo$/
12
  </regexp>
13
  <regexp>
14
    key message
15
    pattern /COMMAND/
16
  </regexp>
17
</filter>
18
​
19
<filter system.**>
20
  @type parser
21
  key_name message
22
  <parse>
23
    @type regexp
24
    expression /USER=(?<sudoer>[^ ]+) ; COMMAND=(?<command>.*)$/
25
  </parse>
26
</filter>
27
​
28
<match system.**>
29
  @type stdout
30
</match>
Copied!
Then restart td-agent:
1
$ sudo systemctl restart td-agent
Copied!
Let's execute some comment with sudo:
1
$ sudo cat /var/log/auth.log
Copied!
Now, you should have a line like this in /var/log/td-agent/td-agent.log:
1
2018-09-27 16:00:01.000000000 +0900 system.authpriv.notice: {"sudoer":"root","command":"/bin/cat"}
Copied!
There it is, as you can see in the line!
Conclusion
Fluentd makes it easy to ingest syslog events. You can immediately send data to the output systems like MongoDB and Elasticsearch, but also you can do filtering and further parsing inside Fluentd before passing the processed data onto the output destinations.
If this article is incorrect or outdated, or omits critical information, please let us know. Fluentd is an open-source project under Cloud Native Computing Foundation (CNCF). All components are available under the Apache 2 License.

Email Alerting like Splunk

Cloud Data Logging with Raspberry Pi

Last modified 4mo ago

Copy link

Contents

Prerequisites

Setting Up rsyslogd

Setting up Fluentd

Step 1: Listening to syslog Messages

Step 2: Extract syslog Messages from sudo

Step 3: Extract Information from Messages

Conclusion

Prerequisites

Setting Up `rsyslogd`

Setting up Fluentd

Step 1: Listening to `syslog` Messages

Step 2: Extract `syslog` Messages from `sudo`

Step 3: Extract Information from Messages

Conclusion