Skip to content

lconsuegra/puppet-prometheus

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

puppet-prometheus

Build Status Puppet Forge Puppet Forge - downloads Puppet Forge - endorsement Puppet Forge - scores Apache-2 License

Table of Contents


Compatibility

Prometheus Version Recommended Puppet Module Version
>= 0.16.2 latest

node_exporter >= 0.15.0 consul_exporter >= 0.3.0

This module supports below Prometheus architectures:

  • x86_64/amd64
  • i386
  • armv71 (Tested on raspberry pi 3)

The prometheus::ipmi_exporter class has a dependency on saz/sudo Puppet module.

Background

This module automates the install and configuration of Prometheus monitoring tool: Prometheus web site

What This Module Affects

  • Installs the prometheus daemon, alertmanager or exporters(via url or package)
    • The package method was implemented, but currently there isn't any package for prometheus
  • Optionally installs a user to run it under (per exporter)
  • Installs a configuration file for prometheus daemon (/etc/prometheus/prometheus.yaml) or for alertmanager (/etc/prometheus/alert.rules)
  • Manages the services via upstart, sysv, or systemd
  • Optionally creates alert rules
  • The following exporters are currently implemented: node_exporter, statsd_exporter, process_exporter, haproxy_exporter, mysqld_exporter, blackbox_exporter, consul_exporter, redis_exporter, varnish_exporter, graphite_exporter, postgres_exporter, collectd_exporter, grok_exporter, ipsec_exporter, openldap_exporter, openvpn_exporter, ssh_exporter, ssl_exporter

Usage

Notice about breaking changes

Version 5.0.0 and older of this module allowed you to deploy the prometheus server by doing a simple include prometheus. We introduced a new class layout in version 6. By default, including the prometheus class won't deploy the server now. You need to include the prometheus::server class for this (which has the same parameters that prometheus had). An alternative approach is to set the manage_prometheus_server parameter to true in the prometheus class. Background information about this change is described in the related pull request and the issue.

For more information regarding class parameters please take a look at the class docstrings.

Prometheus Server (versions < 1.0.0)

class { 'prometheus::server':
  global_config  => {
    'scrape_interval'     => '15s',
    'evaluation_interval' => '15s',
    'external_labels'     => {'monitor' => 'master'},
  },
  rule_files     => ['/etc/prometheus/alert.rules'],
  scrape_configs => [
    {
      'job_name'        => 'prometheus',
      'scrape_interval' => '10s',
      'scrape_timeout'  => '10s',
      'target_groups'   => [
        {
          'targets' => ['localhost:9090'],
          'labels'  => {'alias' => 'Prometheus'}
        },
      ],
    },
  ],
}

Prometheus Server (versions >= 1.0.0 < 2.0.0)

class { 'prometheus::server':
  version        => '1.0.0',
  scrape_configs => [
    {
      'job_name'        => 'prometheus',
      'scrape_interval' => '30s',
      'scrape_timeout'  => '30s',
      'static_configs'  => [
        {
          'targets' => ['localhost:9090'],
          'labels'  => {
            'alias' => 'Prometheus',
          },
        },
      ],
    },
  ],
  alerts         => [
    {
      'name'         => 'InstanceDown',
      'condition'    => 'up == 0',
      'timeduration' => '5m',
      'labels'       => [
        {
          'name'    => 'severity',
          'content' => 'page',
        },
      ],
      'annotations'  => [
        {
          'name'    => 'summary',
          'content' => 'Instance {{ $labels.instance }} down',
        },
        {
          'name'    => 'description',
          'content' => '{{ $labels.instance }} of job {{ $labels.job }} has been down for more than 5 minutes.',
        },
      ],
    },
  ],
  extra_options  => '-alertmanager.url http://localhost:9093 -web.console.templates=/opt/prometheus-1.0.0.linux-amd64/consoles -web.console.libraries=/opt/prometheus-1.0.0.linux-amd64/console_libraries',
  localstorage   => '/prometheus/prometheus',
}

Prometheus Server (versions >= 2.0.0)

class { 'prometheus::server':
  version        => '2.4.3',
  alerts         => {
    'groups' => [
      {
        'name'  => 'alert.rules',
        'rules' => [
          {
            'alert'       => 'InstanceDown',
            'expr'        => 'up == 0',
            'for'         => '5m',
            'labels'      => {
              'severity' => 'page',
            },
            'annotations' => {
              'summary'     => 'Instance {{ $labels.instance }} down',
              'description' => '{{ $labels.instance }} of job {{ $labels.job }} has been down for more than 5 minutes.'
            }
          },
        ],
      },
    ],
  },
  scrape_configs => [
    {
      'job_name'        => 'prometheus',
      'scrape_interval' => '10s',
      'scrape_timeout'  => '10s',
      'static_configs'  => [
        {
          'targets' => [ 'localhost:9090' ],
          'labels'  => {
            'alias' => 'Prometheus',
          }
        }
      ],
    },
  ],
}

When using prometheus >= 2.0, use the new yaml format for rules and alerts.

Which in Puppet means the alerts looks like this:

alerts => {
  'groups' => [
    {
      'name' => 'alert.rules',
      'rules' => [
        {
          'alert'  => 'InstanceDown',
          'expr'   => 'up == 0',
          'for'    => '5m',
          'labels' => {
            'severity' => 'page',
          },
          'annotations' => {
            'summary'     => 'Instance {{ $labels.instance }} down',
            'description' => '{{ $labels.instance }} of job {{ $labels.job }} has been down for more than 5 minutes.',
          }
        }
      ],
    },
  ],
},

And that results in this YAML configuration.

---
alerts:
  groups:
    - name: 'alert.rules'
      rules:
        - alert: 'InstanceDown'
          expr: 'up == 0'
          for: '5m'
          labels:
            severity: 'page'
          annotations:
            summary: 'Instance {{ $labels.instance }} down'
            description: '{{ $labels.instance }} of job {{ $labels.job }} has been down for more than 5 minutes.'

Monitored Nodes

include prometheus::node_exporter

or:

class { 'prometheus::node_exporter':
  version            => '0.12.0',
  collectors_disable => ['loadavg', 'mdadm'],
  extra_options      => '--collector.ntp.server ntp1.orange.intra',
}

Example

Real Prometheus >=2.0.0 setup example including alertmanager and slack_configs.

class { 'prometheus':
  manage_prometheus_server => true,
  version                  => '2.0.0',
  alerts                   => {
    'groups' => [
      {
        'name'  => 'alert.rules',
        'rules' => [
          {
            'alert'       => 'InstanceDown',
            'expr'        => 'up == 0',
            'for'         => '5m',
            'labels'      => {'severity' => 'page'},
            'annotations' => {
              'summary'     => 'Instance {{ $labels.instance }} down',
              'description' => '{{ $labels.instance }} of job {{ $labels.job }} has been down for more than 5 minutes.'
            },
          },
        ],
      },
    ],
  },
  scrape_configs           => [
    {
      'job_name'        => 'prometheus',
      'scrape_interval' => '10s',
      'scrape_timeout'  => '10s',
      'static_configs'  => [
        {
          'targets' => ['localhost:9090'],
          'labels'  => {'alias' => 'Prometheus'}
        }
      ],
    },
    {
      'job_name'        => 'node',
      'scrape_interval' => '5s',
      'scrape_timeout'  => '5s',
      'static_configs'  => [
        {
          'targets' => ['nodexporter.domain.com:9100'],
          'labels'  => {'alias' => 'Node'}
        },
      ],
    },
  ],
  alertmanagers_config     => [
    {
      'static_configs' => [{'targets' => ['localhost:9093']}],
    },
  ],
}

class { 'prometheus::alertmanager':
  version   => '0.13.0',
  route     => {
    'group_by'        => ['alertname', 'cluster', 'service'],
    'group_wait'      => '30s',
    'group_interval'  => '5m',
    'repeat_interval' => '3h',
    'receiver'        => 'slack',
  },
  receivers => [
    {
      'name'          => 'slack',
      'slack_configs' => [
        {
          'api_url'       => 'https://hooks.slack.com/services/ABCDEFG123456',
          'channel'       => '#channel',
          'send_resolved' => true,
          'username'      => 'username'
        },
      ],
    },
  ],
}

And if you want to use Hiera to declare the values instead, you can simply include the prometheus class and set your Hiera data as shown below:

Puppet Code

include prometheus

Hiera Data (in yaml)

---
prometheus::manage_prometheus_server: true

prometheus::version: '2.0.0'

prometheus::alerts:
  groups:
    - name: 'alert.rules'
      rules:
        - alert: 'InstanceDown'
          expr: 'up == 0'
          for: '5m'
          labels:
            severity: 'page'
          annotations:
            summary: 'Instance {{ $labels.instance }} down'
            description: '{{ $labels.instance }} of job {{ $labels.job }} has been
              down for more than 5 minutes.'

prometheus::scrape_configs:
  - job_name: 'prometheus'
    scrape_interval: '10s'
    scrape_timeout: '10s'
    static_configs:
      - targets:
          - 'localhost:9090'
        labels:
          alias: 'Prometheus'
  - job_name: 'node'
    scrape_interval: '10s'
    scrape_timeout: '10s'
    static_configs:
      - targets:
          - 'nodexporter.domain.com:9100'
        labels:
          alias: 'Node'

prometheus::alertmanagers_config:
  - static_configs:
      - targets:
          - 'localhost:9093'

prometheus::alertmanager::version: '0.13.0'

prometheus::alertmanager::route:
  group_by:
    - 'alertname'
    - 'cluster'
    - 'service'
  group_wait: '30s'
  group_interval: '5m'
  repeat_interval: '3h'
  receiver: 'slack'

prometheus::alertmanager::receivers:
  - name: 'slack'
    slack_configs:
      - api_url: 'https://hooks.slack.com/services/ABCDEFG123456'
        channel: "#channel"
        send_resolved: true
        username: 'username'

Test your commit with vagrant https://github.com/kalinux/vagrant-puppet-prometheus.git

Known issues

In version 0.1.14 of this module the alertmanager was configured to run as the service alert_manager. This has been changed in version 0.2.00 to be alertmanager.

Do not use version 1.0.0 of Prometheus: https://groups.google.com/forum/#!topic/prometheus-developers/vuSIxxUDff8 ; it is not compatible with this module!

Even if the module has templates for several linux distributions, only RedHat family distributions were tested.

This module has unit tests for CentOS/RHEL, Debian, Ubuntu and Archlinux. Acceptance tests are executed for CentOS, Debian and Ubuntu. Other operating systems may work but are untested.

Development

This project contains tests for rspec-puppet.

Quickstart to run all linter and unit tests:

bundle install --path .vendor/ --without system_tests --without development --without release
bundle exec rake test

Authors

puppet-prometheus is maintained by Vox Pupuli, it was written by brutus333.

Packages

No packages published

Languages

  • Puppet 62.0%
  • Ruby 33.3%
  • Shell 3.0%
  • HTML 1.6%
  • Dockerfile 0.1%