Ansible Role: Nagios

Installs Nagios, NRPE (both client and server), and configures Nagios automatically based on Ansible's existing inventory of hosts. The goal is to have a complete Nagios monitoring system in a single Ansible role, and without too much manual configuration required.

Prerequisites

Make sure that DNS or /etc/hosts name resolution is working. For example that you can ping web3 at web3.example.com. Email is a recommended prerequisite, with choices of Exim, Postfix, and others.

The templates assume that a dynamic inventory is being used with Openstack to extract the instance UID as the hostname and instance name as the alias.

Role Variables

All the variables in the defaults/ directory may be overridden or changed.

Contacts/Users:

Add users who should be allowed to login to the Nagios web GUI in the nagios_users variable:
nagios_users:

user: nagiosadmin
pass: Password1change
email: nagiosadmin@example.com

Another and better way to add both nagios contacts and nagios admins is the users variable, compatible with https://github.com/mivok/ansible-users

This playbook will look for users in the sysadmin group and those will become nagios admins.

Create the users var, a good place is group_vars/all. Here is an example.

users:

username: foo
name: Foo Barrington
groups: ['sysadmin']
uid: 1001
ssh_key:
- "ssh-rsa AAAAA.... foo@machine"
- "ssh-rsa AAAAB.... foo2@machine"
  htpasswd: $apr1$SheSL4Et$xry6RljdWWvUVrh42s7OA0
  nagios:
  pager: "nagiosadmin_pager@example.com"
  email: "nagiosadmin@example.com"

Where did that htpasswd value come from? Generate the htpasswds manually, and then paste them into users.
htpasswd -n mario
mario:$apr1$SheSL4Et$xry6RljdWWvUVrh42s7OA0

You must configure at minimum one sysadmin user, as just explained, or nagios contacts won't work.

Nagios Commands:

The nagios commands are in the nagios_commands variable, see defaults/main.yml. You can add more commands by adding to, or overriding, the variable.

Nagios Services:

The nagios services are currently in the "checks" per hostgroup in nagios_host_groups, see defaults/main.yml. You can add more services by adding to, or overriding, the variable. Service checks per host are handled similarly in the nagios_hosts variable in defaults/main.yml.

Nagios Hosts:

Ansible hosts in the 'all' group are converted to nagios monitored hosts. No configuration required.

Nagios Hostgroups:

Ansible groups are converted to nagios hostgroups. No configuration required.

Skipping Hosts and Hostgroups

The following values are set in defaults/main.yml :

nagios_hosts_ignore: ""
nagios_groups_ignore: ""

They can be overridden to include Hosts or Hostgroups that should be skipped entirely. Examples:

nagios_hosts_ignore:
  - host1_skip_this_host
nagios_groups_ignore:
  - hostgroup_dev_to_skip
  - and_this_one

This determines which hosts the nagios server will ignore/skip, because these hosts will be excluded from the configs.

Another consideration is whether the ansible playbook should run on all servers. The example playbook test.yml has "- hosts: all", which installs the nagios client on all known servers. You could adjust your top-level playbook and set a different "hosts: " directive. However it's probably fine to install the monitoring client on all servers.

Example Playbook

Refer to test.yml in the root of this role.

To install the client:
roles:
- { role: 'darrylweaver.nagios', run_nagios_client: true }

To install the server:
roles:
- { role: 'darrylweaver.nagios', run_nagios_server: true }

Add nagios servers to the monitoring-servers group in the Ansible inventory.

License

BSD

Author Information

By Darryl Weaver, 2017. Based on pre-existing roles, see ACKNOWLEDGEMENTS.md file. Feedback and bug reports welcome.

darrylweaver / ansible-nagios