Ansible Role: Elasticsearch Curator

An Ansible Role that installs Elasticsearch Curator on RedHat/CentOS or Debian/Ubuntu.

Requirements

None, but it's a lot more helpful if you have Elasticsearch running somewhere!

On RedHat/CentOS, make sure you have the EPEL repository configured, so the python-pip package can be installed. You can install the EPEL repo by simply adding geerlingguy.repo-epel to your playbook's roles.

Role Variables

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

elasticsearch_curator_version: ''

The version of elasticsearch-curator to install. Available versions are listed on the Python Package Index. By default, no version is specified, so the latest version will be installed.

elasticsearch_curator_cron_jobs:
  - name: "Run elasticsearch curator actions."
    job: "/usr/local/bin/curator ~/.curator/action.yml"
    minute: 0
    hour: 1

A list of cron jobs. Typically you would run one cron job using the actions defined in action.yml, but you can split up cron jobs or use the curator_cli to run actions individually instead of via an action file. You can add any of minute, hour, day, weekday, and month to the cron jobs—values that are not explicitly set will default to *. You can also use state to define whether the job should be present or absent.

elasticsearch_curator_config_directory: ~/.curator

The directory inside which Curator's configuration (and action file) will be stored.

elasticsearch_curator_hosts:
  - 'localhost:9200'
elasticsearch_curator_http_auth: ''

These variables control parameters used in the default elasticsearch_curator_yaml. If you define your own custom elasticsearch_curator_yaml, you may not need to define or override these variables. _hosts is a list of hosts with ports, and _http_auth is a basic http_auth user:pass string, if your Elasticsearch instance requires basic HTTP authorization.

elasticsearch_curator_yaml: |
  ---
  client:
    hosts: {{ elasticsearch_curator_hosts | to_yaml }}
    url_prefix:
    use_ssl: False
    certificate:
    client_cert:
    client_key:
    ssl_no_validate: False
    http_auth: {{ elasticsearch_curator_http_auth }}
    timeout: 30
    master_only: False
  logging:
    loglevel: INFO
    logfile:
    logformat: default
    blacklist: ['elasticsearch', 'urllib3']

This YAML goes into the file ~/.curator/curator.yml and stores the connection details Elasticsearch Curator uses to when connecting to Elasticsearch, as well as Curator logging configuration.

elasticsearch_curator_action_yaml: |
  ---
  actions:
    1:
      action: delete_indices
      options:
        ignore_empty_list: True
        disable_action: False
      filters:
      - filtertype: pattern
        kind: prefix
        value: logstash-
        exclude:
      - filtertype: age
        source: name
        direction: older
        timestring: '%Y.%m.%d'
        unit: days
        unit_count: 45
        exclude:

This YAML goes into the file ~/.curator/action.yml and defines the actions Curator performs when the default cron job is run. See documentation: Curator actions file.

elasticsearch_curator_pip_package: 'python3-pip'
elasticsearch_curator_pip_executable: "{{ 'pip3' if elasticsearch_curator_pip_package.startswith('python3') else 'pip' }}"

System pip package which needs to be installed, and the pip executable to run. For older OSes or when using Python 2, you may need to override this and change these to python-pip, and pip, respectively.

Dependencies

• geerlingguy.repo-epel (RedHat/CentOS only)

Example Playbook

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

License

MIT / BSD

Author Information

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