Ansible Role: Logstash

An Ansible Role that installs Logstash on RedHat/CentOS Debian/Ubuntu.

Note that this role installs a syslog grok pattern by default; if you want to add more filters, please add them inside the /etc/logstash/conf.d/ directory. As an example, you could create a file named 13-myapp.conf with the appropriate grok filter and restart logstash to start using it. Test your grok regex using the Grok Debugger.

Requirements

Though other methods are possible, this role is made to work with Elasticsearch as a backend for storing log messages.

Role Variables

Available variables are listed below, along with default values (see defaults/main.yml):

logstash_listen_port_tcp: 5000
logstash_listen_port_udp: 5000

The TCP and UDP ports over which logstash will listen for syslog messages.

logstash_elasticsearch_host: localhost

The host on which Elasticsearch resides.

logstash_ssl_dir: /etc/pki/logstash
logstash_ssl_certificate_file: logstash-forwarder-example.crt
logstash_ssl_key_file: logstash-forwarder-example.key

SSL configuration for Logstash to accept requests from logstash-forwarder running on remote hosts. Security note: On production or public-facing (e.g. any non-test) servers, you should create your own key/certificate pair and use that instead of the included default! You can use OpenSSL to create the key and certificate files, with a command like the following: openssl req -x509 -batch -nodes -newkey rsa:2048 -keyout lumberjack.key -out lumberjack.crt -subj /CN=logstash.example.com.

For the logstash_ssl_certificate_file and logstash_ssl_key_file, you can provide a path relative to the role directory, or an absolute path to the file.

logstash_local_syslog_path: /var/log/syslog
logstash_monitor_local_syslog: true

Whether configuration for local syslog file (defined as logstash_local_syslog_path) should be added to logstash. Set this to false if you are monitoring the local syslog differently, or if you don't care about the local syslog file. Other local logs can be added by your own configuration files placed inside /etc/logstash/conf.d.

logstash_enabled_on_boot: yes

Set this to no if you don't want logstash to run on system startup.

Other Notes

If you are seeing high CPU usage from one of the logstash processes, and you're using Logstash along with another application running on port 80 on a platform like Ubuntu with upstart, the logstash-web process may be stuck in a loop trying to start on port 80, failing, and trying to start again, due to the restart flag being present in /etc/init/logstash-web.conf. To avoid this problem, either change that line to add a limit to the respawn statement, or set the logstash-web service to enabled=no in your playbook, e.g.:

- name: Ensure logstash-web process is stopped and disabled.
  service: name=logstash-web state=stopped enabled=no

Example Playbook

- hosts: search
  roles:
    - { role: geerlingguy.elasticsearch }
    - { role: geerlingguy.logstash }

License

MIT / BSD

Author Information

This role was created in 2014 by Jeff Geerling, author of Ansible for DevOps.