Skip to content

Latest commit

 

History

History
1634 lines (1329 loc) · 44.7 KB

README.md

File metadata and controls

1634 lines (1329 loc) · 44.7 KB

Collectd module for Puppet

Puppet Forge Build Status

Description

Puppet module for configuring collectd and plugins.

Usage

The simplest use case is to use all of the configurations in the default collectd.conf file shipped with collectd. This can be done by simply including the class:

include ::collectd

Collectd is most useful when configured with customized plugins. This is accomplished by removing the default collectd.conf file and replacing it with a file that includes all alternative configurations. Configure a node with the following class declaration:

class { '::collectd':
  purge           => true,
  recurse         => true,
  purge_config    => true,
  minimum_version => '5.4',
}

Set purge, recurse, and purge_config to true in order to override the default configurations shipped in collectd.conf and use custom configurations stored in conf.d. From here you can set up additional plugins as shown below.

Specifying the version or minimum_version of collectd as shown above reduces the need for two puppet runs to coverge. See Puppet needs two runs to correctly write my conf, why? below.

Hiera example in YAML of passing install_options to the package resource for managing the collectd package. This parameter must be an array.

collectd::package_install_options:
  - '--nogpgcheck'

Simple Plugins

Example of how to load plugins with no additional configuration:

collectd::plugin { 'battery': }

where 'battery' is the name of the plugin. Note, this should only be done in the case of a class for the plugin not existing in this module.

Configurable Plugins

Parameters will vary widely between plugins. See the collectd documentation for each plugin for configurable attributes.

####Class: collectd::plugin::aggregation

collectd::plugin::aggregation::aggregator {
  cpu':
    plugin           => 'cpu',
    type             => 'cpu',
    groupby          => ["Host", "TypeInstance",],
    calculateaverage => true,
}

You can as well configure this plugin with a parameterized class :

class { 'collectd::plugin::aggregation':
  aggregators => {
    cpu' => {
      plugin           => 'cpu',
      type             => 'cpu',
      groupby          => ["Host", "TypeInstance",],
      calculateaverage => true,
    },
  },
}

####Class: collectd::plugin::amqp

class { 'collectd::plugin::amqp':
  amqphost => '127.0.0.1',
  amqpvhost => 'myvirtualhost',
  graphiteprefix => 'collectdmetrics',
  amqppersistent => true,
}

####Class: collectd::plugin::apache

class { 'collectd::plugin::apache':
  instances => {
    'apache80' => {
      'url' => 'http://localhost/mod_status?auto', 'user' => 'collectd', 'password' => 'hoh2Coo6'
    },
    'lighttpd8080' => {
      'url' => 'http://localhost:8080/mod_status?auto'
    }
  },
}

####Class: collectd::plugin::bind

class { 'collectd::plugin::bind':
  url    => 'http://localhost:8053/',
}

####Class: collectd::plugin::ceph

class { 'collectd::plugin::ceph':
  osds          => [ 'osd.0', 'osd.1', 'osd.2'],
}

####Class: collectd::plugin::cgroups

See collectd plugin_cgroups documentation for more details.

class { 'collectd::plugin::cgroups':
  ignore_selected => true,
  cgroups         => ['array', 'of', 'paths']
}

####Class: collectd::plugin::chain

class { 'collectd::plugin::chain':
    chainname     => "PostCache",
    defaulttarget => "write",
    rules         => [
      {
        'match'   => {
          'type'    => 'regex',
          'matches' => {
            'Plugin'         => "^cpu$",
            'PluginInstance' => "^[0-9]+$",
          },
        },
        'targets' => [
          {
            'type'       => "write",
            'attributes' => {
              "Plugin" => "aggregation",
            },
          },
          {
            'type' => "stop",
          },
        ],
      },
    ],
  }

####Class: collectd::plugin::conntrack

class { 'collectd::plugin::conntrack':
}

####Class: collectd::plugin::cpu

  • reportbystate available from collectd version >= 5.5
  • reportbycpu available from collectd version >= 5.5
  • valuespercentage available from collectd version >= 5.5

See collectd plugin_cpu documentation for more details.

class { 'collectd::plugin::cpu':
  reportbystate => true,
  reportbycpu => true,
  valuespercentage => true,
}

####Class: collectd::plugin::cpufreq

class { 'collectd::plugin::cpufreq':
}

####Class: collectd::plugin::csv

class { 'collectd::plugin::csv':
  datadir    => '/etc/collectd/var/lib/collectd/csv',
  storerates => false,
}

####Class: collectd::plugin::curl

collectd::plugin::curl::page {
  'stock_quotes':
    url      => 'http://finance.google.com/finance?q=NYSE%3AAMD',
    user     => 'foo',
    password => 'bar',
    matches  => [
      {
        'dstype'   => 'GaugeAverage',
        'instance' => 'AMD',
        'regex'    => ']*> *([0-9]*\\.[0-9]+) *',
        'type'     => 'stock_value',
      }],
}

You can as well configure this plugin with a parameterized class :

class { 'collectd::plugin::curl':
  pages => {
    'stock_GM' => {
      url      => 'http://finance.google.com/finance?q=NYSE%3AGM',
      user     => 'foo',
      password => 'bar',
      matches  => [
        {
          'dstype'   => 'GaugeAverage',
          'instance' => 'AMD',
          'regex'    => ']*> *([0-9]*\\.[0-9]+) *',
          'type'     => 'stock_value',
        },
      ],
    },
  },
}

####Class: collectd::plugin::curl_json

collectd::plugin::curl_json {
  'rabbitmq_overview':
    url => 'http://localhost:55672/api/overview',
    instance => 'rabbitmq_overview',
    interval => '300',
    keys => {
      'message_stats/publish' => {
        'type'     => 'gauge',
        'instance' => 'overview',
      },
    }
}

####Class: collectd::plugin::dbi

collectd::plugin::dbi::database{'monitoring_node1':
  driver       => 'mysql',
  driveroption => {
    'host' => 'hostname',
    'username' => 'user',
    'password' => 'password',
    'dbname'   => 'monitoring'
  },
  query    => ['log_delay'],
}
collectd::plugin::dbi::query{'log_delay':
  statement => 'SELECT * FROM log_delay_repli;',
  results   => [{
    type           => 'gauge',
    instanceprefix => 'log_delay',
    instancesfrom  => 'inet_server_port',
    valuesfrom     => 'log_delay',
  }],
}

You can as well configure this plugin as a parameterized class :

class { 'collectd::plugin::dbi':
  package   => 'libdbd-mysql',
  databases => {
    'monitoring_node1' => {
      driver       => 'mysql',
      driveroption => {
        'host' => 'hostname',
        'username' => 'user',
        'password' => 'password',
        'dbname'   => 'monitoring'
      },
      query    => ['log_delay'],
    }
  },
}

####Class: collectd::plugin::df

class { 'collectd::plugin::df':
  mountpoints    => ['/u'],
  fstypes        => ['nfs','tmpfs','autofs','gpfs','proc','devpts'],
  ignoreselected => true,
}

####Class: collectd::plugin::disk

class { 'collectd::plugin::disk':
  disks          => ['/^dm/'],
  ignoreselected => true,
  udevnameattr   => 'DM_NAME',
}

####Class: collectd::plugin::dns

class { 'collectd::plugin::dns':
}
Parameters:

See collectd documentation.

ensure

Optional. String that must be 'present' or 'absent'.

  • Default: 'present'

ignoresource

Optional. String of IP address to ignore.

  • Default: undef

interface

String of interface to use. May be interface identifier such as 'eth0' or 'any'.

  • Default: 'any'

interval

Numeric for interval in seconds. Changing this can be a disaster. Consult the documentation.

  • Default: undef

manage_package

Boolean to determine if system package for collectd's dns plugin should be managed. If set to true, you must specify the package name for any unsupported OS.

  • Default: false

package_name

String for name of package. A value of 'USE_DEFAULTS' will set the value based on the platform. This is necessary when setting manage_package on an unsupported platform.

  • Default: 'USE_DEFAULTS'

selectnumericquerytypes

Boolean for SelectNumericQueryTypes configuration option.

  • Default: true

####Class: collectd::plugin::entropy

class { 'collectd::plugin::entropy':
}

####Class: collectd::plugin::exec

collectd::plugin::exec::cmd {
  'dummy':
    user => nobody,
    group => nogroup,
    exec => ["/bin/echo", "PUTVAL myhost/foo/gauge-flat N:1"],
}

You can also configure this plugin with a parameterized class:

class { 'collectd::plugin::exec':
  commands => {
    'dummy1' => {
      user  => nobody,
      group => nogroup,
      exec  => ["/bin/echo", "PUTVAL myhost/foo/gauge-flat1 N:1"],
    },
    'dummy2' => {
      user  => nobody,
      group => nogroup,
      exec  => ["/bin/echo", "PUTVAL myhost/foo/gauge-flat2 N:1"],
    },
  }
}

####Class: collectd::plugin::ethstat

class { 'collectd::plugin::ethstat':
  interfaces => [ 'eth0', 'eth1'],
  maps       => [ '"rx_csum_offload_errors" "if_rx_errors" checksum_offload"', '"multicast" "if_multicast"' ],
  mappedonly => false,
}

####Class: collectd::plugin::fhcount

class { 'collectd::plugin::fhcount':
  valuesabsolute   => true,
  valuespercentage => false,
}

####Class: collectd::plugin::filecount

collectd::plugin::filecount::directory {'foodir':
  path          => '/path/to/dir',
  pattern       => '*.conf',
  mtime         => '-5m',
  recursive     => true,
  includehidden => false
}

You can also configure this plugin with a parameterized class:

class { 'collectd::plugin::filecount':
  directories => {
    'foodir' => {
      'path'          => '/path/to/dir',
      'pattern'       => '*.conf',
      'mtime'         => '-5m',
      'recursive'     => true,
      'includehidden' => false
      },
  },
}

For backwards compatibility:

class { 'collectd::plugin::filecount':
  directories => {
    'active'   => '/var/spool/postfix/active',
    'incoming' => '/var/spool/postfix/incoming'
  },
}

Class: collectd::plugin::filter

The filter plugin implements the powerful filter configuration of collectd. For further details have a look on the collectd manpage.

Parameters:
  • ensure ("ensure","absent"): Ob absent it will remove all references of the filter plugins. Note: The Chain config needs to be purged by the chain define.
  • precachechain (String): The Name of the default Pre Chain.
  • postcachechain (String): The Name of the default Post Chain.
Examples:
Overwrite default chains:
class { 'collectd::plugin::filter':
    ensure          => 'present',
    precachechain   => 'PreChain',
    postcachechain  => 'PostChain',
}
Full Example:

This Example will rename the hostname of the mysql plugin.

include collectd::plugin::filter

# define default chains with default target
collectd::plugin::filter::chain { 'PreChain':
    target => 'return'
}
collectd::plugin::filter::chain { 'PostChain':
    target => 'write'
}

# create a third chain,
$chainname = 'MyAweseomeChain'
collectd::plugin::filter::chain { $chainname:
    ensure => present,
    target => 'return'
}

# add a new rule to chain
$rulename = 'MyAweseomeRule'
collectd::plugin::filter::rule { $rulename:
    chain => $chainname,
}

# add a new match rule, match metrics of the mysql plugin
collectd::plugin::filter::match { "Match mysql plugin":
    chain   => $chainname,
    rule    => $rulename,
    plugin  => 'regex',
    options => {
        'Plugin' => '^mysql',
    }
}

#rewrite hostname
collectd::plugin::filter::target{ "overwrite hostname":
    chain   => $chainname,
    rule    => $rulename,
    plugin  => 'set',
    options => {
        'Host' => 'hostname.domain',
    },
}

# hook the configured chain in the prechain
collectd::plugin::filter::target{ "1_prechain_jump_${chainname}":
    chain   => 'PreChain',
    plugin  => 'jump',
    options => {
        'Chain' => $chainname,
    },
}

Define: collectd::plugin::filter::chain

This define will create a new chain, which is required by targets, matches and rules.

Parameters:
  • ensure ("ensure","absent"): When set to absent it will remove the chain with all assigned rules, targets and matches.
  • target ('notification','replace','set','return','stop','write','jump'): Optional. Set default target if no target has been applied. Strongly recommend for default chains.
  • target_options (Hash): If target is specified, pass optional hash to define.
Example:

see collectd::plugin::filter above

Define: collectd::plugin::filter::rule

This define will add a new rule to a specific chain

Parameters:
  • chain (String): Assign to this chain.
Example:

see collectd::plugin::filter above

Define: collectd::plugin::filter::target

This define will add a target to a chain or rule.

Parameters:
  • chain (String): Assign to this chain.
  • plugin ('notification','replace','set','return','stop','write','jump'): The plugin of the target.
  • options (Hash): Optional parameters of the target plugin.
  • rule (String): Optional. Assign to this rule. If not present, target will be applied at the end of chain without rule matching.
Example:

see collectd::plugin::filter above

Define: collectd::plugin::filter::match

This define will add a match rule.

Parameters:
  • chain (String): Assign to this chain.
  • rule (String): Assign to this rule.
  • plugin ('regex','timediff','value','empty_counter','hashed'): The plugin of the match.
  • options (Hash): Optional parameters of the match plugin.
Example:

see collectd::plugin::filter above

####Class: collectd::plugin::genericjmx

include collectd::plugin::genericjmx

collectd::plugin::genericjmx::mbean {
  'garbage_collector':
    object_name     => 'java.lang:type=GarbageCollector,*',
    instance_prefix => 'gc-',
    instance_from   => 'name',
    values          => [
      {
        mbean_type => 'invocations',
        table      => false,
        attribute  => 'CollectionCount',
      },
      {
        mbean_type      => 'total_time_in_ms',
        instance_prefix => 'collection_time',
        table           => false,
        attribute       => 'CollectionTime',
      },
    ];
}

collectd::plugin::genericjmx::connection {
  'java_app':
    host            => $fqdn,
    service_url     => 'service:jmx:rmi:///jndi/rmi://localhost:3637/jmxrmi',
    collect         => [ 'memory-heap', 'memory-nonheap','garbage_collector' ],
}

####Class: collectd::plugin::interface

class { 'collectd::plugin::interface':
  interfaces     => ['lo'],
  ignoreselected => true
}

####Class: collectd::plugin::irq

class { 'collectd::plugin::irq':
  irqs           => ['7', '23'],
  ignoreselected => true,
}

####Class: collectd::plugin::ipmi

class { 'collectd::plugin::ipmi':
  ignore_selected           => true,
  sensors                   => ['temperature'],
  notify_sensor_add         => true,
  notify_sensor_remove      => true,
  notify_sensor_not_present => true,
}

####Class: collectd::plugin::iptables

class { 'collectd::plugin::iptables':
  chains  => {
    'nat'    => 'In_SSH',
    'filter' => 'HTTP'
  },
}

####Class: collectd::plugin::java

jvmarg options must be declared if declaring loadplugin, as the JVM must be initialized prior to loading collectd java plugins.

class { 'collectd::plugin::java':
  jvmarg      => ['arg1', 'arg2'],
  loadplugin  => {"plugin.name" => ["option line 1", "option line 2"]}
}

####Class: collectd::plugin::load

class { 'collectd::plugin::load':
}

####Class: collectd::plugin::logfile

class { 'collectd::plugin::logfile':
  log_level => 'warning',
  log_file => '/var/log/collected.log'
}

####Class: collectd::plugin::virt

The interface_format parameter was introduced in collectd 5.0 and will therefore be ignored (with a warning) when specified with older versions.

class { 'collectd::plugin::virt':
  connection       => 'qemu:///system',
  interface_format => 'address'
}

####Class: collectd::plugin::lvm

class { 'collectd::plugin::lvm': }

####Class: collectd::plugin::memcached

class { 'collectd::plugin::memcached':
  host => '192.168.122.1',
  port => 11211,
}

####Class: collectd::plugin::memory

class { 'collectd::plugin::memory':
}

####Class: collectd::plugin::mysql

collectd::plugin::mysql::database { 'betadase':
  host        => 'localhost',
  username    => 'stahmna',
  password    => 'secret',
  port        => '3306',
  masterstats => true,
}

####Class: collectd::plugin::mongodb

class { 'collectd::plugin::mongodb':
  db_user => 'admin',
  db_pass => 'adminpass',
}
Parameters:
  • ensure (String): String that must be 'present' or 'absent'. Default: 'present'
  • interval (String): Number of seconds that collectd pauses between data collection. Default: undef
  • db_host (String): String that holds the IP of the MongoDB server. Default: '127.0.0.1'
  • db_user (String): Required. String that specifies the user name of an account that can log into MongoDB
  • db_user (String): Required. String that specifies the password of an account that can log into MongoDB
  • configured_dbs (String): Optional. Array of Strings that lists the databases that should be monitored in addition to the "admin"
  • db_port (String): Required if the configured_dbs parameter is set. Unused otherwise. Integer that specifies with port MongoDB listens on.
class { 'collectd::plugin::mongodb':
  db_host        => '127.0.0.1',
  db_user        => 'foo',
  db_pass        => 'bar',
  db_port        => '27017',
  configured_dbs => ['database', 'names'],
  collectd_dir   => '/collectd/module/path',
}

####Class: collectd::plugin::netlink

class { 'collectd::plugin::netlink':
  interfaces        => ['eth0', 'eth1'],
  verboseinterfaces => ['ppp0'],
  qdiscs            => ['"eth0" "pfifo_fast-1:0"', '"ppp0"'],
  classes           => ['"ppp0" "htb-1:10"'],
  filters           => ['"ppp0" "u32-1:0"'],
  ignoreselected    => false,
}

####Class: collectd::plugin::network

collectd::plugin::network::server{'hostname':
  port => 25826,
}

collectd::plugin::network::listener{'hostname':
  port => 25826,
}

You can as well configure this plugin with a parameterized class :

class { 'collectd::plugin::network':
  timetolive    => '70',
  maxpacketsize => '42',
  forward       => false,
  reportstats   => true,
  servers       => { 'hostname' => {
    'port'          => '25826',
    'interface'     => 'eth0',
    'securitylevel' => '',
    'username'      => 'foo',
    'password'      => 'bar',},
  },
  listeners     => { 'hostname' => {
    'port'          => '25826',
    'interface'     => 'eth0',
    'securitylevel' => '',
    'authfile'      => '/etc/collectd/passwd',},
  },
}

####Class: collectd::plugin::nfs

class { 'collectd::plugin::nfs':
}

####Class: collectd::plugin::nginx

class { 'collectd::plugin::nginx':
  url      => 'https://localhost:8433',
  user     => 'stats',
  password => 'uleePi4A',
}

####Class: collectd::plugin::ntpd

class { 'collectd::plugin::ntpd':
  host           => 'localhost',
  port           => 123,
  reverselookups => false,
  includeunitid  => false,
}

####Class: collectd::plugin::openldap

class { 'collectd::plugin::openldap':
  instances => {
    'foo' => {
      'url' => 'ldap://localhost/'
    },
    'bar' => {
      'url' => 'ldaps://localhost/'
    }
  },
}

####Class: collectd::plugin::openvpn

  • statusfile (String or Array) Status file(s) to collect data from. (Default /etc/openvpn/openvpn-status.log)
  • improvednamingschema (Bool) When enabled, the filename of the status file will be used as plugin instance and the client's "common name" will be used as type instance. This is required when reading multiple status files. (Default: false)
  • collectcompression Sets whether or not statistics about the compression used by OpenVPN should be collected. This information is only available in single mode. (Default true)
  • collectindividualusers Sets whether or not traffic information is collected for each connected client individually. If set to false, currently no traffic data is collected at all because aggregating this data in a save manner is tricky. (Default true)
  • collectusercount When enabled, the number of currently connected clients or users is collected. This is especially interesting when CollectIndividualUsers is disabled, but can be configured independently from that option. (Default false)

Watch multiple statusfiles:

class { 'collectd::plugin::openvpn':
  statusfile             => [ '/etc/openvpn/openvpn-status-tcp.log', '/etc/openvpn/openvpn-status-udp.log' ],
  collectindividualusers => false,
  collectusercount       => true,
}

Watch the single default statusfile:

class { 'collectd::plugin::openvpn':
  collectindividualusers => false,
  collectusercount       => true,
}

####Class: collectd::plugin::perl

This class has no parameters and will load the actual perl plugin. It will be automatically included if any perl::plugin is defined.

#####Example:

include collectd::plugin::perl

####Define: collectd::plugin::perl::plugin

This define will load a new perl plugin.

#####Parameters:

  • module (String): name of perl module to load (mandatory)
  • enable_debugger (False or String): whether to load the perl debugger. See collectd-perl manpage for more details.
  • include_dir (String or Array): directories to add to @INC
  • provider ("package","cpan","file" or false): method to get the plugin code
  • source (String): this parameter is consumed by the provider to infer the source of the plugin code
  • destination (String or false): path to plugin code if providerisfile`. Ignored otherwise.
  • order (String containing numbers): order in which the plugin should be loaded. Defaults to "00"
  • config (Hash): plugin configuration in form of a hash. This will be converted to a suitable structure understood by liboconfig which is the collectd configuration parser. Defaults to {}

#####Examples:

######Using a preinstalled plugin:

collectd::plugin::perl::plugin { 'foo':
    module          => 'Collectd::Plugins::Foo',
    enable_debugger => "",
    include_dir     => '/usr/lib/collectd/perl5/lib',
}

######Using a plugin from a file from source:

collectd::plugin::perl::plugin { 'baz':
    module      => 'Collectd::Plugins::Baz',
    provider    => 'file',
    source      => 'puppet:///modules/myorg/baz_collectd.pm',
    destination => '/path/to/my/perl5/modules'
}

######Using a plugin from cpan (requires the puppet cpan module):

collectd::plugin::perl::plugin {
  'openafs_vos':
    module        => 'Collectd::Plugins::OpenAFS::VOS',
    provider      => 'cpan',
    source        => 'Collectd::Plugins::OpenAFS',
    config        => {'VosBin' => '/usr/afsws/etc/vos'},
}

######Using a plugin from package source:

collectd::plugin::perl::plugin {
  'bar':
    module        => 'Collectd::Plugins::Bar',
    provider      => 'package',
    source        => 'perl-Collectd-Plugins-Bar',
    config        => {'foo' => 'bar'},
}

####Class: collectd::plugin::ping

collectd::plugin::ping {
  'example':
    hosts => ['example.com'],
}

####Class: collectd::plugin::postgresql

collectd::plugin::postgresql::database{'monitoring_node1':
  name     => 'monitoring',
  port     => '5433',
  instance => 'node1',
  host     => 'localhost',
  user     => 'collectd',
  password => 'collectd',
  query    => 'log_delay',
}
collectd::plugin::postgresql::query{'log_delay':
  statement => 'SELECT * FROM log_delay_repli;',
  results   => [{
    type           => 'gauge',
    instanceprefix => 'log_delay',
    instancesfrom  => 'inet_server_port',
    valuesfrom     => 'log_delay',
  }],
}
collectd::plugin::postgresql::writer{'sqlstore':
  statement  => 'SELECT collectd_insert($1, $2, $3, $4, $5, $6, $7, $8, $9);',
  storerates => 'true',
}

You can as well configure this plugin as a parameterized class :

class { 'collectd::plugin::postgresql':
  databases => {
    'postgres' => {
      'host'     => '/var/run/postgresql/',
      'user'     => 'postgres',
      'password' => 'postgres',
      'sslmode'  => 'disable',
      'query'    => [ 'query_plans', 'queries', 'table_states', 'disk_io' ],
    },
    'devdb' => {
      'host'     => 'host.example.com',
      'port'     => '5432',
      'user'     => 'postgres',
      'password' => 'secret',
      'sslmode'  => 'prefer',
    }
  }
}

####Class: collectd::plugin::processes

You can either specify processes / process matches at once:

class { 'collectd::plugin::processes':
  processes => ['process1', 'process2'],
  process_matches => [
    { name => 'process-all', regex => 'process.*' }
  ],
}

Or define single processes / process matches:

collectd::plugin::processes::process { 'collectd' : }
collectd::plugin::processes::processmatch { 'elasticsearch' :
  regex => '.*java.*org.elasticsearch.bootstrap.Elasticsearch'
}

####Class: collectd::plugin::protocols

  • values is an array of Protocol names, Protocol:ValueName pairs, or a regex
  • see /proc/net/netstat and /proc/net/snmp for a list of Protocol targets

See [collectd.conf documentation] (https://collectd.org/documentation/manpages/collectd.conf.5.shtml#plugin_protocols) for details

class { 'collectd::plugin::protocols':
  values => ['/^Tcp:*/', '/^Udp:*/', 'Icmp:InErrors' ],
  ignoreselected => false,
}

####Class: collectd::plugin::python

  • modulepaths is an array of paths where will be Collectd looking for Python modules, Puppet will ensure that each of specified directories exists and it is owned by root (and chmod 0750). If you don't specify any modulepaths a default value for given distribution will be used.
  • modules a Hash containing configuration of Python modules, where the key is the module name
  • globals Unlike most other plugins, this one should set Globals true. This will cause collectd to export the name of all objects in the Python interpreter for all plugins to see. If you don't do this or your platform does not support it, the embedded interpreter will start anyway but you won't be able to load certain Python modules, e.g. "time".
  • interactive when true it will launch an interactive Python interpreter that reads from and writes to the terminal (default: false)
  • logtraces if a Python script throws an exception it will be logged by collectd with the name of the exception and the message (default: false)

See collectd-python documentation for more details.

NOTE: Since v3.4.0 the syntax of this plugin has changed. Make sure to update your existing configuration. Now you can specify multiple Python modules at once:

class { 'collectd::plugin::python':
  modulepaths => ['/usr/share/collectd/python'],
  modules     => {
    'elasticsearch' => {
      'script_source' => 'puppet:///modules/myorg/elasticsearch_collectd_python.py',
      'config'        => {'Cluster' => 'elasticsearch'},
    },
    'another-module' => {
      'config'        => {'Verbose' => 'true'},
    }
  }
  logtraces   => true,
  interactive => false
}

When script_source provided, a file called {module}.py will be created in $modulepath/$module.py.

Or define single module:

collectd::plugin::python::module {'zk-collectd':
  script_source => 'puppet:///modules/myorg/zk-collectd.py',
  config        => {
    'Hosts' => "localhost:2181"
  }
}

Each plugin might use different modulepath, however make sure that all paths are included in collectd::plugin::python variable modulepaths. If no modulepath is specified, OS default will be used.

collectd::plugin::python::module {'my-module':
  modulepath    => '/var/share/collectd',
  script_source => 'puppet:///modules/myorg/my-module.py',
  config        => {
    'Key' => "value"
  }
}

####Class: collectd::plugin::redis

class { 'collectd::plugin::redis':
  nodes => {
    'node1' => {
      'host'     => 'localhost',
    },
    'node2' => {
      'host'     => 'node2.example.com',
      'port'     => '6380',
      'timeout'  => 3000,
    },
    'node3' => {
      'host'    => 'node3.example.com',
      'queries' => {
          'dbsize' => {
              'type'  => 'count',
              'query' => 'DBSIZE',
          },
      },
    },
  }
}

####Class: collectd::plugin::rabbitmq

Please note the rabbitmq plugin provides a types.db.custom. You will need to add this to collectd::config::typesdb via hiera or in a manifest. Failure to set the types.db.custom content will result in no metrics from the rabbitmq plugin.

set typesdb to include the collectd-rabbitmq types.db.custom

collectd::config::typesdb:
  - /usr/share/collectd/types.db
  - /usr/share/collect-rabbitmq/types.db.custom
class { '::collectd::plugin::rabbitmq':
  config           => {
    'Username' => 'admin',
    'Password' => $admin_pass,
    'Scheme'   => 'https',
    'Port'     => '15671',
    'Host'     => $::fqdn,
    'Realm'    => '"RabbitMQ Management"',
  },
}

####Class: collectd::plugin::rrdcached

class { 'collectd::plugin::rrdcached':
  daemonaddress => 'unix:/var/run/rrdcached.sock',
  datadir       => '/var/lib/rrdcached/db/collectd',
}

####Class: collectd::plugin::rrdtool

class { 'collectd::plugin::rrdtool':
  datadir           => '/var/lib/collectd/rrd',
  createfilesasync  => false,
  rrarows           => 1200,
  rratimespan       => [3600, 86400, 604800, 2678400, 31622400],
  xff               => 0.1,
  cacheflush        => 900,
  cachetimeout      => 120,
  writespersecond   => 50
}

####Class: collectd::plugin::sensors

class {'collectd::plugin::sensors':
  ignoreselected => false,
}

####Class: collectd::plugin::snmp

class {'collectd::plugin::snmp':
  data  =>  {
    amavis_incoming_messages => {
      'Type'     => 'counter',
      'Table'    => false,
      'Instance' => 'amavis.inMsgs',
      'Values'   => ['AMAVIS-MIB::inMsgs.0']
    }
  },
  hosts => {
    debianvm => {
      'Address'   => '127.0.0.1',
      'Version'   => 2,
      'Community' => 'public',
      'Collect'   => ['amavis_incoming_messages'],
      'Interval'  => 10
    }
  },
}

####Class: collectd::plugin::statsd

class { 'collectd::plugin::statsd':
  host            => '0.0.0.0',
  port            => 8125,
  deletecounters  => false,
  deletetimers    => false,
  deletegauges    => false,
  deletesets      => false,
  timerpercentile => ['50','90'],
}

####Class: collectd::plugin::swap

class { 'collectd::plugin::swap':
  reportbydevice => false,
  reportbytes    => true
}

####Class: collectd::plugin::syslog

class { 'collectd::plugin::syslog':
  log_level => 'warning'
}

####Class: collectd::plugin::target_v5upgrade

class { 'collectd::plugin::target_v5upgrade':
}

####Class: collectd::plugin::tcpconns

class { 'collectd::plugin::tcpconns':
  localports  => ['25', '12026'],
  remoteports => ['25'],
  listening   => false,
}

####Class: collectd::plugin::tail

collectd::plugin::tail::file { 'exim-log':
  filename => '/var/log/exim4/mainlog',
  instance => 'exim',
  matches  => [
    {
      regex    => 'S=([1-9][0-9]*)',
      dstype   => 'CounterAdd',
      type     => 'ipt_bytes',
      instance => 'total',
    },
    {
      regex    => '\\<R=local_user\\>',
      dstype   => 'CounterInc',
      type     => 'counter',
      instance => 'local_user',
    }
  ]
}

####Class: collectd::plugin::unixsock

class {'collectd::plugin::unixsock':
  socketfile   => '/var/run/collectd-sock',
  socketgroup  => 'nagios',
  socketperms  => '0770',
  deletesocket => false,
}

####Class: collectd::plugin::uptime

class {'collectd::plugin::uptime':
}

####Class: collectd::plugin::users

class {'collectd::plugin::users':
}

####Class: collectd::plugin::varnish

class { 'collectd::plugin::varnish':
  instances => {
    'instanceName' => {
      'CollectCache' => 'true',
      'CollectBackend' => 'true',
      'CollectConnections' => 'true',
      'CollectSHM' => 'true',
      'CollectESI' => 'false',
      'CollectFetch' => 'true',
      'CollectHCB' => 'false',
      'CollectTotals' => 'true',
      'CollectWorkers' => 'true',
    }
  },
}

####Class: collectd::plugin::vmem

class { 'collectd::plugin::vmem':
  verbose => true,
}

####Class: collectd::plugin::write_graphite

The write_graphite plugin writes data to Graphite, an open-source metrics storage and graphing project.

collectd::plugin::write_graphite::carbon {'my_graphite':
  graphitehost   => 'graphite.example.org',
  graphiteport   => 2003,
  graphiteprefix => '',
  protocol       => 'tcp'
}

You can define multiple Graphite backends where will be metrics send. Each backend should have unique title:

collectd::plugin::write_graphite::carbon {'secondary_graphite':
  graphitehost      => 'graphite.example.org',
  graphiteport      => 2004,
  graphiteprefix    => '',
  protocol          => 'udp',
  escapecharacter   => '_',
  alwaysappendds    => true,
  storerates        => true,
  separateinstances => false,
  logsenderrors     => true
}

####Class: collectd::plugin::write_http

class { 'collectd::plugin::write_http':
  urls => {
    'collect1.example.org' => { 'format' => 'JSON' },
    'collect2.example.org' => {},
  }
}

####Class: collectd::plugin::write_kafka

class { 'collectd::plugin::write_kafka':
  kafka_host => 'localhost',
  kafka_port => 9092,
  topics     => {
    'mytopic'      => { 'format' => 'JSON' },
  }
}

####Class: collectd::plugin::write_network

Deprecated

class { 'collectd::plugin::write_network':
  servers => {
    'collect1.example.org' => { 'serverport' => '25826' },
    'collect2.example.org' => { 'serverport' => '25826' }
  }
}

####Class: collectd::plugin::write_riemann

class { 'collectd::plugin::write_riemann':
  riemann_host => 'riemann.example.org',
  riemann_port => 5555,
  tags         => ['foo'],
  attributes   => {'bar' => 'baz'},
}

####Class: collectd::plugin::write_sensu

class { 'collectd::plugin::write_sensu':
  sensu_host => 'sensu.example.org',
  sensu_port => 3030,
}

####Class: collectd::plugin::write_tsdb

class { 'collectd::plugin::write_tsdb':
  host             => 'tsdb.example.org',
  port             => 4242,
  host_tags        => ['environment=production', 'colocation=AWS'],
  store_rates      => false,
  always_append_ds => false,
}

####Class: collectd::plugin::zfs_arc

class { 'collectd::plugin::zfs_arc':
}

types.db

Collectd needs to know how to handle each collected datapoint. For this it uses a database file called types.db

Those files can be created using the collectd::typesdb and collectd::type define resources.

$db = '/etc/collectd/types.db'
collectd::typesdb { $db: }

collectd::type { "response_size-${db}":
  target  => $db,
  ds_type => 'ABSOLUTE',
  min     => 0,
  max     => 10000000,
  ds_name => 'value',
}

class { 'collectd':
  typesdb      => [
    '/usr/share/collectd/types.db',
    $typesdb,
  ],
}

Other software may need to read the Collectd types database files. To allow non-root users to read from a collectd::typesdb file like so:

$db = '/etc/collectd/types.db'
collectd::typesdb { $db:
  mode => '0644',
}

##Limitations

See metadata.json for supported platforms

##Known issues

###Puppet needs two runs to correctly write my conf, why?

Some plugins will need two runs of Puppet to fully generate the configuration for collectd. See this issue. This can be avoided by specifying a minimum version ($minimum_version) for the collectd class. e.g. Setting this to 1.2.3 will make this module assume on the first run (when the fact responsible to provide the collectd version is not yet available) that your systems are running collectd 1.2.3 and generate the configuration accordingly.

##Development

Running tests

This project contains tests for rspec-puppet.

Quickstart:

gem install bundler
bundle install
bundle exec rake lint
bundle exec rake validate
bundle exec rake spec SPEC_OPTS='--format documentation'

Version scoping

Some plugins or some options in plugins are only available for recent versions of collectd.

This module shall not use unsupported configuration directives. Look at templates/loadplugin.conf.erb for a hands-on example.

Please make use of the search by branch/tags on the collectd github to see when a function has been first released.

Reading the collectd.conf.pod file is good, validating the presence of the code in the .c files is even better.

Authors

Puppet-collectd is maintained by VoxPupuli. Before moving to VoxPupuli it was written and maintained by TheCAT in the pdxcat github org.