mircomasa.filebeat

Overview Versions Insights

Filebeat Role

Ansible

Summary

This role:

installs filebeat on Ubuntu, CentOS, Windows
copies prepared configuration file (log path, connect to elasticsearch etc.)

Role tasks

Prepare server (add elastic repo)
[Optional] Create folder(s) for custom paths
Install filebeat
Copy configuration file

Requirements

Minimal Version of the ansible for installation: 2.5
Supported OS:
- CentOS
  - 6, 7
- Ubuntu
  - 16.04, 18.04
- Debian
  - 8, 9
- Windows

Role Variables

You can override any variable below by setting "variable: value" in playbook.

filebeat_version Is used to select main Filebeat branch to be installed (5.x or 6.x current stable versions). Default value is 6.
filebeat_last_version Is used to select specific Filebeat version to be installed. Default value is 6.6.0
elastic_gpg_key GPG-key from elasticsearch repository. Default value is https://artifacts.elastic.co/GPG-KEY-elasticsearch
filebeat_node_name Name of the filebeat node. Default value is {{ inventory_hostname }}. If this options is not defined, the hostname is used.
filebeat_ssl_enabled Turns on/off SSL connection between filebeat and logstash/elasticsearch. SSL options should be set by corresponding dict fields like shown below:

  ssl:
    key: "/etc/pki/tls/private/server.key"
    certificate: "/etc/pki/tls/certs/server.crt"
    certificate_authorities: "/etc/pki/CA/ca-root.pem"

in case of Windows setup:

 ssl:
   key: 'c:\tls\private\server.key'
   certificate: 'c:\tls\certs\server.pem'
   certificate_authorities: 'c:\CA\ca-root.pem'

The path section of the configuration options defines where Filebeat looks for its files. For example, Filebeat looks for the Elasticsearch template file in the configuration path and writes log files in the logs path. Filebeat looks for its registry files in the data path. Default values for Linux host are set up this way:

path:
  home: /usr/share/filebeat
  config: /etc/filebeat
  data: /var/lib/filebeat
  logs: /var/log/filebeat

in case of Windows setup default paths look like:

path:
  home: 'c:\program files\filebeat'
  config: 'c:\program files\filebeat'
  data: 'c:\programdata\filebeat'
  logs: 'c:\programdata\filebeat\logs'

win_download_path Temp directory for Windows to download and upzip Filebeat package. Default value is '{{ ansible_env.TEMP }}/filebeat' (ansible_env.TEMP value solves idempotence issue)
input_logpath Path to log files.

Default value for *NIX OS family is "/var/log/*.log"

Default value for WIN OS family is 'c:\windows\*.log'

Variable filebeat_inputs defines type of logs that will be processed by pipeline, their log paths and Elasticsearch index that should store this type of logs. You can specify several inputs with various paths, logtypes and index names using yaml format like in example below:

    filebeat_inputs:
      - name: hybris
        paths:
          - '/var/log/console*.log'
        fields:
          logtype: hybris
          index_name: hybris-console
      - name: access
        paths:
          - '/var/log/access*.log'
          - '/var/log/nginx_access*.log'
        fields:
          logtype: access
          index_name: nginx-access

Output customization:

filebeat_output Is used to configure what output to use when sending data (elasticsearch or logstash). Default value is elasticsearch
filebeat_elasticsearch_output.hosts Array of hosts to connect to. Default value is localhost
filebeat_elasticsearch_output.port Value for setting custom port. Default value is 9200
filebeat_logstash_outputs Array of hosts and ports to connect to. Default value is localhost:5044

Advanced config parameters:

The filebeat(systemd)\initd section of the configuration options defines which init script will be used to manage filebeat service depending on the *nix OS. Custom paths will be taken into account (if configured).

filebeat_service_name Name of nssm\init script, which manages filebeat service
filebeat_bulk_max_size Maximum number of events to bulk in a single Logstash request. Default value is 500
filebeat_worker Number of workers per Elasticsearch host. Default value is 1
filebeat_logging_to_syslog Send all logging output to syslog. Default value is false
filebeat_logging_to_files Send all logging output to rotating files. Default value is true
filebeat_rotateeverybytes Defines log file size limit. Defalt value is 104857600 = 100MB
filebeat_keepfiles Number of log files to keep. Default value is 30
filebeat_ignore_older Value (any time strings like 2h, 5m can be used) above which files will be ignored. Default value is 0 (disabled)
filebeat_scan_frequency Defines how often filebeat checks file updates. Default value is 15s
filebeat_harvester_buffer_size Defines the buffer size. Default value is 65535
filebeat_logname Name of the logging files. Default value is "filebeat.log"

Dependencies

ca-cert (only for installation with SSL)

Example Playbook

Installing Filebeat 6.x version:

- name: Install filebeat
  hosts: all
  roles:
    - role: ansible-role-filebeat

Installing Filebeat 6.x version with custom path to log files and elasticsearch output:

- name: Install filebeat
  hosts: all
  roles:
    - role: ansible-role-filebeat
  vars:
    input_logpath: "/var/log/messages"
    filebeat_elasticsearch_output:
      hosts:
        - elasticsearch.example.com
      port: 9200