3 collectd-perl - Documentation of collectd's C<perl plugin>
10 IncludeDir "/path/to/perl/plugins"
11 BaseName "Collectd::Plugin"
22 The C<perl plugin> embeds a Perl-interpreter into collectd and provides an
23 interface to collectd's plugin system. This makes it possible to write plugins
24 for collectd in Perl. This is a lot more efficient than executing a
25 Perl-script every time you want to read a value with the C<exec plugin> (see
26 L<collectd-exec(5)>) and provides a lot more functionality, too.
32 =item B<LoadPlugin> I<Plugin>
34 Loads the Perl plugin I<Plugin>. This does basically the same as B<use> would
35 do in a Perl program. As a side effect, the first occurrence of this option
36 causes the Perl-interpreter to be initialized.
38 =item B<BaseName> I<Name>
40 Prepends I<Name>B<::> to all plugin names loaded after this option. This is
41 provided for convenience to keep plugin names short.
43 =item E<lt>B<Plugin> I<Name>E<gt> block
45 This block may be used to pass on configuration settings to a Perl plugin. The
46 configuration is converted into a config-item data type which is passed to the
47 registered configuration callback. See below for details about the config-item
48 data type and how to register callbacks.
50 The I<name> identifies the callback. It is used literally and independent of
51 the B<BaseName> setting.
53 =item B<EnableDebugger> I<Package>[=I<option>,...]
55 Run collectd under the control of the Perl source debugger. If I<Package> is
56 not the empty string, control is passed to the debugging, profiling, or
57 tracing module installed as Devel::I<Package>. A comma-separated list of
58 options may be specified after the "=" character. Please note that you may not
59 leave out the I<Package> option even if you specify B<"">. This is the same as
60 using the B<-d:Package> command line option.
62 See L<perldebug> for detailed documentation about debugging Perl.
64 This option does not prevent collectd from daemonizing, so you should start
65 collectd with the B<-f> command line option. Else you will not be able to use
66 the command line driven interface of the debugger.
68 =item B<IncludeDir> I<Dir>
70 Adds I<Dir> to the B<@INC> array. This is the same as using the B<-IDir>
71 command line option or B<use lib Dir> in the source code. Please note that it
72 only has effect on plugins loaded after this option.
76 =head1 WRITING YOUR OWN PLUGINS
78 Writing your own plugins is quite simple. collectd manages plugins by means of
79 B<dispatch functions> which call the appropriate B<callback functions>
80 registered by the plugins. Any plugin basically consists of the implementation
81 of these callback functions and initializing code which registers the
82 functions with collectd. See the section "EXAMPLES" below for a really basic
83 example. The following types of B<callback functions> are known to collectd
84 (all of them are optional):
88 =item configuration functions
90 This type of functions is called during configuration if an appropriate
91 B<Plugin> block has been encountered. It is called once for each B<Plugin>
92 block which matches the name of the callback as provided with the
93 B<plugin_register> method - see below.
97 This type of functions is called once after loading the module and before any
98 calls to the read and write functions. It should be used to initialize the
99 internal state of the plugin (e.E<nbsp>g. open sockets, ...). If the return
100 value evaluates to B<false>, the plugin will be disabled.
104 This type of function is used to collect the actual data. It is called once
105 per interval (see the B<Interval> configuration option of collectd). Usually
106 it will call B<plugin_dispatch_values> to dispatch the values to collectd
107 which will pass them on to all registered B<write functions>. If the return
108 value evaluates to B<false> the plugin will be skipped for an increasing
109 amount of time until it returns B<true> again.
111 =item write functions
113 This type of function is used to write the dispatched values. It is called
114 once for each call to B<plugin_dispatch_values>.
116 =item flush functions
118 This type of function is used to flush internal caches of plugins. It is
119 usually triggered by the user only. Any plugin which caches data before
120 writing it to disk should provide this kind of callback function.
124 This type of function is used to pass messages of plugins or the daemon itself
127 =item notification function
129 This type of function is used to act upon notifications. In general, a
130 notification is a status message that may be associated with a data instance.
131 Usually, a notification is generated by the daemon if a configured threshold
132 has been exceeded (see the section "THRESHOLD CONFIGURATION" in
133 L<collectd.conf(5)> for more details), but any plugin may dispatch
134 notifications as well.
136 =item shutdown functions
138 This type of function is called once before the daemon shuts down. It should
139 be used to clean up the plugin (e.g. close sockets, ...).
143 Any function (except log functions) may set the B<$@> variable to describe
144 errors in more detail. The message will be passed on to the user using
145 collectd's logging mechanism.
147 See the documentation of the B<plugin_register> method in the section
148 "METHODS" below for the number and types of arguments passed to each
149 B<callback function>. This section also explains how to register B<callback
150 functions> with collectd.
152 To enable a plugin, copy it to a place where Perl can find it (i.E<nbsp>e. a
153 directory listed in the B<@INC> array) just as any other Perl plugin and add
154 an appropriate B<LoadPlugin> option to the configuration file. After
155 restarting collectd you're done.
159 The following complex types are used to pass values between the Perl plugin
166 A config-item is one structure which keeps the informations provided in the
167 configuration file. The array of children keeps one entry for each
168 configuration option. Each such entry is another config-item structure, which
169 may nest further if nested blocks are used.
173 values => [ val1, val2, ... ],
174 children => [ { ... }, { ... }, ... ]
179 A data-set is a list of one or more data-sources. Each data-source defines a
180 name, type, min- and max-value and the data-set wraps them up into one
181 structure. The general layout looks like this:
184 name => 'data_source_name',
185 type => DS_TYPE_COUNTER || DS_TYPE_GAUGE,
186 min => value || undef,
187 max => value || undef
192 A value-list is one structure which features an array of values and fields to
193 identify the values, i.E<nbsp>e. time and host, plugin name and
194 plugin-instance as well as a type and type-instance. Since the "type" is not
195 included in the value-list but is passed as an extra argument, the general
196 layout looks like this:
199 values => [123, 0.5],
201 interval => $interval_g,
203 plugin => 'myplugin',
205 plugin_instance => '',
211 A notification is one structure defining the severity, time and message of the
212 status message as well as an identification of a data instance. Also, it
213 includes an optional list of user-defined meta information represented as
217 severity => NOTIF_FAILURE || NOTIF_WARNING || NOTIF_OKAY,
219 message => 'status message',
221 plugin => 'myplugin',
223 plugin_instance => '',
225 meta => [ { name => <name>, value => <value> }, ... ]
230 A match-proc is one structure storing the callbacks of a "match" of the filter
231 chain infrastructure. The general layout looks like this:
234 create => 'my_create',
235 destroy => 'my_destroy',
241 A target-proc is one structure storing the callbacks of a "target" of the
242 filter chain infrastructure. The general layout looks like this:
245 create => 'my_create',
246 destroy => 'my_destroy',
247 invoke => 'my_invoke'
254 The following functions provide the C-interface to Perl-modules. They are
255 exported by the ":plugin" export tag (see the section "EXPORTS" below).
259 =item B<plugin_register> (I<type>, I<name>, I<data>)
261 Registers a callback-function or data-set.
263 I<type> can be one of:
287 I<name> is the name of the callback-function or the type of the data-set,
288 depending on the value of I<type>. (Please note that the type of the data-set
289 is the value passed as I<name> here and has nothing to do with the I<type>
290 argument which simply tells B<plugin_register> what is being registered.)
292 The last argument, I<data>, is either a function name or an array-reference.
293 If I<type> is B<TYPE_DATASET>, then the I<data> argument must be an
294 array-reference which points to an array of hashes. Each hash describes one
295 data-set. For the exact layout see B<Data-Set> above. Please note that
296 there is a large number of predefined data-sets available in the B<types.db>
297 file which are automatically registered with collectd - see L<types.db(5)> for
298 a description of the format of this file.
300 B<Note>: Using B<plugin_register> to register a data-set is deprecated. Add
301 the new type to a custom L<types.db(5)> file instead. This functionality might
302 be removed in a future version of collectd.
304 If the I<type> argument is any of the other types (B<TYPE_INIT>, B<TYPE_READ>,
305 ...) then I<data> is expected to be a function name. If the name is not
306 prefixed with the plugin's package name collectd will add it automatically.
307 The interface slightly differs from the C interface (which expects a function
308 pointer instead) because Perl does not support to share references to
309 subroutines between threads.
311 These functions are called in the various stages of the daemon (see the
312 section "WRITING YOUR OWN PLUGINS" above) and are passed the following
319 The only argument passed is I<config-item>. See above for the layout of this
328 No arguments are passed.
332 The arguments passed are I<type>, I<data-set>, and I<value-list>. I<type> is a
333 string. For the layout of I<data-set> and I<value-list> see above.
337 The arguments passed are I<timeout> and I<identifier>. I<timeout> indicates
338 that only data older than I<timeout> seconds is to be flushed. I<identifier>
339 specifies which values are to be flushed.
343 The arguments are I<log-level> and I<message>. The log level is small for
344 important messages and high for less important messages. The least important
345 level is B<LOG_DEBUG>, the most important level is B<LOG_ERR>. In between there
346 are (from least to most important): B<LOG_INFO>, B<LOG_NOTICE>, and
347 B<LOG_WARNING>. I<message> is simply a string B<without> a newline at the end.
351 The only argument passed is I<notification>. See above for the layout of this
356 =item B<plugin_unregister> (I<type>, I<plugin>)
358 Removes a callback or data-set from collectd's internal list of
359 functionsE<nbsp>/ datasets.
361 =item B<plugin_dispatch_values> (I<value-list>)
363 Submits a I<value-list> to the daemon. If the data-set identified by
364 I<value-list>->{I<type>}
365 is found (and the number of values matches the number of data-sources) then the
366 type, data-set and value-list is passed to all write-callbacks that are
367 registered with the daemon.
369 B<Note>: Prior to version 4.4 of collectd, the data-set type used to be passed
370 as the first argument to B<plugin_register>. This syntax is still supported
371 for backwards compatibility but has been deprecated and will be removed in
372 some future version of collectd.
374 =item B<plugin_write> ([B<plugins> => I<...>][, B<datasets> => I<...>],
375 B<valuelists> => I<...>)
377 Calls the write function of the given I<plugins> with the provided I<data
378 sets> and I<value lists>. In contrast to B<plugin_dispatch_values>, it does
379 not update collectd's internal cache and bypasses the filter mechanism (see
380 L<collectd.conf(5)> for details). If the B<plugins> argument has been omitted,
381 the values will be dispatched to all registered write plugins. If the
382 B<datasets> argument has been omitted, the required data sets are looked up
383 according to the C<type> member in the appropriate value list. The value of
384 all three arguments may either be a single scalar or a reference to an array.
385 If the B<datasets> argument has been specified, the number of data sets has to
386 equal the number of specified value lists.
388 =item B<plugin_flush> ([B<timeout> => I<timeout>][, B<plugins> => I<...>][,
389 B<identifiers> => I<...>])
391 Flush one or more plugins. I<timeout> and the specified I<identifiers> are
392 passed on to the registered flush-callbacks. If omitted, the timeout defaults
393 to C<-1>. The identifier defaults to the undefined value. If the B<plugins>
394 argument has been specified, only named plugins will be flushed. The value of
395 the B<plugins> and B<identifiers> arguments may either be a string or a
396 reference to an array of strings.
398 =item B<plugin_flush_one> (I<timeout>, I<plugin>)
400 This is identical to using "plugin_flush (timeout =E<gt> I<timeout>, plugins
403 B<Note>: Starting with version 4.5 of collectd, B<plugin_flush_one> has been
404 deprecated and will be removed in some future version of collectd. Use
405 B<plugin_flush> instead.
407 =item B<plugin_flush_all> (I<timeout>)
409 This is identical to using "plugin_flush (timeout =E<gt> I<timeout>)".
411 B<Note>: Starting with version 4.5 of collectd, B<plugin_flush_all> has been
412 deprecated and will be removed in some future version of collectd. Use
413 B<plugin_flush> instead.
415 =item B<plugin_dispatch_notification> (I<notification>)
417 Submits a I<notification> to the daemon which will then pass it to all
418 notification-callbacks that are registered.
420 =item B<plugin_log> (I<log-level>, I<message>)
422 Submits a I<message> of level I<log-level> to collectd's logging mechanism.
423 The message is passed to all log-callbacks that are registered with collectd.
425 =item B<ERROR>, B<WARNING>, B<NOTICE>, B<INFO>, B<DEBUG> (I<message>)
427 Wrappers around B<plugin_log>, using B<LOG_ERR>, B<LOG_WARNING>,
428 B<LOG_NOTICE>, B<LOG_INFO> and B<LOG_DEBUG> respectively as I<log-level>.
432 The following function provides the filter chain C-interface to Perl-modules.
433 It is exported by the ":filter_chain" export tag (see the section "EXPORTS"
438 =item B<fc_register> (I<type>, I<name>, I<proc>)
440 Registers filter chain callbacks with collectd.
442 I<type> may be any of:
452 I<name> is the name of the match or target. By this name, the callbacks are
453 identified in the configuration file when specifying a B<Match> or B<Target>
454 block (see L<collectd.conf(5)> for details).
456 I<proc> is a hash reference. The hash includes up to three callbacks: an
457 optional constructor (B<create>) and destructor (B<destroy>) and a mandatory
458 B<match> or B<invoke> callback. B<match> is called whenever processing an
459 appropriate match, while B<invoke> is called whenever processing an
460 appropriate target (see the section "FILTER CONFIGURATION" in
461 L<collectd.conf(5)> for details). Just like any other callbacks, filter chain
462 callbacks are identified by the function name rather than a function pointer
463 because Perl does not support to share references to subroutines between
464 threads. The following arguments are passed to the callbacks:
470 The arguments passed are I<config-item> and I<user-data>. See above for the
471 layout of the config-item data-type. I<user-data> is a reference to a scalar
472 value that may be used to store any information specific to this particular
473 instance. The daemon does not care about this information at all. It's for the
478 The only argument passed is I<user-data> which is a reference to the user data
479 initialized in the B<create> callback. This callback may be used to cleanup
480 instance-specific information and settings.
484 The arguments passed are I<data-set>, I<value-list>, I<meta> and I<user-data>.
485 See above for the layout of the data-set and value-list data-types. I<meta> is
486 a pointer to an array of meta information, just like the B<meta> member of the
487 notification data-type (see above). I<user-data> is a reference to the user
488 data initialized in the B<create> callback.
494 =head1 GLOBAL VARIABLES
500 As the name suggests this variable keeps the hostname of the system collectd
501 is running on. The value might be influenced by the B<Hostname> or
502 B<FQDNLookup> configuration options (see L<collectd.conf(5)> for details).
506 This variable keeps the interval in seconds in which the read functions are
507 queried (see the B<Interval> configuration option).
511 Any changes to these variables will be globally visible in collectd.
515 By default no symbols are exported. However, the following export tags are
516 available (B<:all> will export all of them):
524 =item B<plugin_register> ()
526 =item B<plugin_unregister> ()
528 =item B<plugin_dispatch_values> ()
530 =item B<plugin_flush> ()
532 =item B<plugin_flush_one> ()
534 =item B<plugin_flush_all> ()
536 =item B<plugin_dispatch_notification> ()
538 =item B<plugin_log> ()
556 =item B<TYPE_SHUTDOWN>
560 =item B<TYPE_DATASET>
568 =item B<DS_TYPE_COUNTER>
570 =item B<DS_TYPE_GAUGE>
600 =item B<:filter_chain>
606 =item B<FC_MATCH_NO_MATCH>
608 =item B<FC_MATCH_MATCHES>
610 =item B<FC_TARGET_CONTINUE>
612 =item B<FC_TARGET_STOP>
614 =item B<FC_TARGET_RETURN>
632 =item B<NOTIF_FAILURE>
634 =item B<NOTIF_WARNING>
654 Any Perl plugin will start similar to:
656 package Collectd::Plugins::FooBar;
661 use Collectd qw( :all );
663 A very simple read function might look like:
667 my $vl = { plugin => 'foobar' };
668 $vl->{'values'} = [ rand(42) ];
669 plugin_dispatch_values ('gauge', $vl);
673 A very simple write function might look like:
677 my ($type, $ds, $vl) = @_;
678 for (my $i = 0; $i < scalar (@$ds); ++$i) {
679 print "$vl->{'plugin'} ($vl->{'type'}): $vl->{'values'}->[$i]\n";
684 A very simple match callback might look like:
688 my ($ds, $vl, $meta, $user_data) = @_;
689 if (matches($ds, $vl)) {
690 return FC_MATCH_MATCHES;
692 return FC_MATCH_NO_MATCH;
696 To register those functions with collectd:
698 plugin_register (TYPE_READ, "foobar", "foobar_read");
699 plugin_register (TYPE_WRITE, "foobar", "foobar_write");
701 fc_register (FC_MATCH, "foobar", "foobar_match");
703 See the section "DATA TYPES" above for a complete documentation of the data
704 types used by the read, write and match functions.
712 Please feel free to send in new plugins to collectd's mailinglist at
713 E<lt>collectdE<nbsp>atE<nbsp>verplant.orgE<gt> for review and, possibly,
714 inclusion in the main distribution. In the latter case, we will take care of
715 keeping the plugin up to date and adapting it to new versions of collectd.
717 Before submitting your plugin, please take a look at
718 L<http://collectd.org/dev-info.shtml>.
728 collectd is heavily multi-threaded. Each collectd thread accessing the perl
729 plugin will be mapped to a Perl interpreter thread (see L<threads(3perl)>).
730 Any such thread will be created and destroyed transparently and on-the-fly.
732 Hence, any plugin has to be thread-safe if it provides several entry points
733 from collectd (i.E<nbsp>e. if it registers more than one callback or if a
734 registered callback may be called more than once in parallel). Please note
735 that no data is shared between threads by default. You have to use the
736 B<threads::shared> module to do so.
740 Each function name registered with collectd has to be available before the
741 first thread has been created (i.E<nbsp>e. basically at compile time). This
742 basically means that hacks (yes, I really consider this to be a hack) like
743 C<*foo = \&bar; plugin_register (TYPE_READ, "plugin", "foo");> most likely
744 will not work. This is due to the fact that the symbol table is not shared
745 across different threads.
749 Each plugin is usually only loaded once and kept in memory for performance
750 reasons. Therefore, END blocks are only executed once when collectd shuts
751 down. You should not rely on END blocks anyway - use B<shutdown functions>
756 The perl plugin exports the internal API of collectd which is considered
757 unstable and subject to change at any time. We try hard to not break backwards
758 compatibility in the Perl API during the life cycle of one major release.
759 However, this cannot be guaranteed at all times. Watch out for warnings
760 dispatched by the perl plugin after upgrades.
770 Currently, it is not possible to flush a single Perl plugin only. You can
771 either flush all Perl plugins or none at all and you have to use C<perl> as
772 plugin name when doing so.
784 L<threads::shared(3perl)>,
789 The C<perl plugin> has been written by Sebastian Harl
790 E<lt>shE<nbsp>atE<nbsp>tokkee.orgE<gt>.
792 This manpage has been written by Florian Forster
793 E<lt>octoE<nbsp>atE<nbsp>verplant.orgE<gt> and Sebastian Harl
794 E<lt>shE<nbsp>atE<nbsp>tokkee.orgE<gt>.