NRPE Configuration

An Ansible role to handle the installation and rollout of the Nagios NRPE Daemon.

This role currently supports:

• Red Hat Enterprise Linux 7
• CentOS 7
• Fedora versions

Requirements

• EPEL needs to be installed, and configured, on Red Hat, Scientific Linux, CentOS and any other RHEL-derivative. Fedora systems should function properly without this requirement.

Role Information

This role gives you the ability to deploy plugins on a global and per-server basis. This can be done by:

• putting plugins into files/plugins/global or
• by creating a folder in files/plugins/ that is the servers FQDN.

You can find out your servers FQDN by running the Ansible Setup module.

OS Specific Role variables

These following variable are Operating System, or at least Operating System family specific. They are configurable, but change at your own risk!

• nagios_nrpe_server_pid: /var/run/nrpe/nrpe.pid
• nagios_nrpe_server_user: nrpe
• nagios_nrpe_server_group: nrpe
• nagios_nrpe_server_repo_redhat: epel
• nagios_nrpe_server_service: nrpe
• nagios_nrpe_server_dir: /etc/nagios

Role Variables

• nrpe_data: Filesystem tree (local-only) where extra checks need to reside.
• nrpe: base configuration variable for NRPE. This variable acts as a dictionary (hash) of nrpe configuration options.
• nrpe_defaults: Default options are available in the vars/RedHat.yml file. The defaults are used of no corresponding nrpe[option] variable is set.
• nrpe_<configuration_option>: The override option exists for all nrpe variables. To override the default and the base configuration variable -- for instance, in a specific host_vars folder -- use this configuration option.
• nrpe_command: This variable contains the default, and system-specific, nrpe check options.

Default Role variables

---
#
# Check for virtualization types. If we're virtualized (mainly testing), then
# don't try to manage or reload the service. Do not change unless you know what
# you are doing!
#
nrpe_manage_service: {}
nrpe_allow_restart: {}

#
# Should we skip the defaults provided by the operating system? This should
# typically be set to false. If you are generating a completely independent
# configuration file, then you should probably set this to try and override
# the nrpe variable below.
#
nrpe_skip_defaults: false

#
# By default, we set the os_supported variable to "no". This way, we can
# override it in the vars defaults files for any OS that we can support with
# this role. This short-circuits the execution of the playbook for not EL-
# derivatives.
#
nrpe_os_supported: no

#
# By default, inherit from the Red Hat nrpe_defaults variables.
#
nrpe: {}
nrpe_command: {}

$> cat group_vars/all
nrpe:
  server_address: 127.0.0.1
  bind_address: 127.0.0.1
  allowed_hosts:
    - nagios-server.example.com
    - 127.0.0.1
  port: 5666
  dont_blame_nrpe: 0
  log_facility: daemon

nrpe_command:
  default:
    oracle_tnsping:
      script: check_oracle_health
      option: --mode tnsping
    oracle_connection-time:
      script: check_oracle_health
      option: --mode connection-time
  'hostname.domain':
    check_uptime:
      script: check_uptime
      option: -u days -w 90 -c 120

$> cat host_vars/host.example.com
---
#
# NRPE runs on port 5667 on this host, so this variable overrides nrpe['port']
# in the group_vars/all file.
#
nrpe_port: 5667

Dependencies

none

Example Playbook

- hosts: servers
  roles:
     - ISU-Ansible.nrpe
  vars:
    nrpe:
      allowed_hosts:
        - 192.168.0.1
        - 127.0.0.1
    nrpe_command:
      default:
        oracle_tnsping:
          script: check_oracle_health
          option: --mode tnsping
        oracle_connection-time:
          script: check_oracle_health
          option: --mode connection-time
      'hostname.domain':
        check_uptime:
          script: check_uptime
          option: -u days -w 90 -c 120

License

MIT

Author Information

• Barry Britt

This role was adapted from the nagios-nrpe-server role by jloh. Checkout his blog here.