vector

This puppet module installs, configures, and runs the Vector observability tool.

Description

The vector module installs, configures, and runs the Vector observability tool. This module is designed for Redhat and Fedora type systems, and also assumes it uses systemd for managing services.

Setup

Setup Requirements

This module requires puppet-stdlib.

Beginning with vector

This module will not configure yum or apt repositories, those should be configured outside of this module.

Usage

There are two ways to configure topologies:

1. Using class parameters
   • These can be set with Hiera!
2. Using defined types
   • Especially useful if your modules need to inject their own vector components
   • vector::source to configure a vector source
   • vector::transform to configure a vector transform
   • vector::sink to configure a vector sink
   • vector::configfile to configure a set of sources, transforms, and sinks

You can even use a mix of both of these strategies.

Examples

Using vector::configfile

require vector

vector::configfile { 'vector':
  data => {
    'sources' => {
        'logfiles' => {
            'type' => 'file',
            'include' => ['/var/log/**/*.log'],
        },
    },
    'sinks' => {
        'kafka' => {
            'type' => 'kafka',
            'inputs' => ['logfiles'],
            'bootstrap_servers' => 'localhost:9092',
            'encoding' => {
                'codec' => 'json',
            },
            'topic' => 'logs',
        },
    }
  },
}

Using vector::source, vector::transform, and vector::sink Note: 'type' is required for all of these defined types and 'inputs' is required for transform and sink types. Addition configuration parameters for these should be put in the 'parameters' hash.

require vector

vector::source { 'logfile_input':
    type       => 'file',
    parameters => {
        'include' => ['/var/log/**/*.log'],
    },
}

vector::transform { 'logfile_transform':
    type       => 'remap',
    inputs     => ['*'],
    parameters => {
        'source' => '.foo = "bar"',
    },
}

vector::sink { 'logfile_kafka':
    type       => 'kafka',
    inputs     => ['logfile_transform'],
    parameters => {
        'bootstrap_servers' => 'localhost:9092',
        'topic'             => 'logs',
        'encoding'          => {
            'codec' => 'json',
        },
    }
}

Using class parameters

class { 'vector':
  data_dir => '/data/vector',
  sources  => {
    'logfiles' => {
      'type'       => 'file',
      'parameters' => {
        'include'   => ['/var/log/**/*.log'],
        'read_from' => 'beginning',
      },
    },
    'syslogs'  => {
      'type'       => 'syslog',
      'parameters' => {
        'mode'    => 'tcp',
        'address' => '0.0.0.0:514',
      },
    },
  },
  sinks    => {
    'elasticsearch' => {
      'type'       => 'elasticsearch',
      'inputs'     => ['logfiles', 'syslogs'],
      'parameters' => {
        'endpoints' => 'elastic1:9200',
        'pipeline'  => 'logs',
      },
    },
  },
}

Using Hiera

vector::data_dir: '/data/vector'
vector::sources:
  logfiles:
    type: file
    parameters:
      include: ['/var/log/**/*.log']
      read_from: beginning
  syslogs:
    type: syslog
    parameters:
      mode: tcp
      address: '0.0.0.0:514'
vector::sinks:
  elasticsearch:
    type: elasticsearch
    inputs: ['logfiles','syslogs']
    parameters:
      endpoints: 'elastic1:9200'
      pipeline: logs

Reference

See REFERENCE.md