package Monitoring::Plugin; use Monitoring::Plugin::Functions qw(:codes %ERRORS %STATUS_TEXT @STATUS_CODES); use Params::Validate qw(:all); use 5.006; use strict; use warnings; use Carp; use base qw(Class::Accessor::Fast); Monitoring::Plugin->mk_accessors(qw( shortname perfdata messages opts threshold )); use Exporter; our @ISA = qw(Exporter); our @EXPORT = (@STATUS_CODES); our @EXPORT_OK = qw(%ERRORS %STATUS_TEXT); # CPAN stupidly won't index this module without a literal $VERSION here, # so we're forced to duplicate it explicitly # Make sure you update $Monitoring::Plugin::Functions::VERSION too our $VERSION = "0.39"; sub new { my $class = shift; # my %args = @_; my %args = validate( @_, { shortname => 0, usage => 0, version => 0, url => 0, plugin => 0, blurb => 0, extra => 0, license => 0, timeout => 0 }, ); my $shortname = Monitoring::Plugin::Functions::get_shortname(\%args); delete $args{shortname} if (exists $args{shortname}); my $self = { shortname => $shortname, perfdata => [], # to be added later messages => { warning => [], critical => [], ok => [] }, opts => undef, # see below threshold => undef, # defined later }; bless $self, $class; if (exists $args{usage}) { require Monitoring::Plugin::Getopt; $self->opts( new Monitoring::Plugin::Getopt(%args) ); } return $self; } sub add_perfdata { my ($self, %args) = @_; require Monitoring::Plugin::Performance; my $perf = Monitoring::Plugin::Performance->new(%args); push @{$self->perfdata}, $perf; } sub all_perfoutput { my $self = shift; return join(" ", map {$_->perfoutput} (@{$self->perfdata})); } sub set_thresholds { my $self = shift; require Monitoring::Plugin::Threshold; return $self->threshold( Monitoring::Plugin::Threshold->set_thresholds(@_)); } # MP::Functions wrappers sub plugin_exit { my $self = shift; Monitoring::Plugin::Functions::plugin_exit(@_, { plugin => $self }); } sub plugin_die { my $self = shift; Monitoring::Plugin::Functions::plugin_die(@_, { plugin => $self }); } sub nagios_exit { my $self = shift; Monitoring::Plugin::Functions::plugin_exit(@_, { plugin => $self }); } sub nagios_die { my $self = shift; Monitoring::Plugin::Functions::plugin_die(@_, { plugin => $self }); } sub die { my $self = shift; Monitoring::Plugin::Functions::plugin_die(@_, { plugin => $self }); } sub max_state { Monitoring::Plugin::Functions::max_state(@_); } sub max_state_alt { Monitoring::Plugin::Functions::max_state_alt(@_); } # top level interface to Monitoring::Plugin::Threshold sub check_threshold { my $self = shift; my %args; if ( $#_ == 0 && (! ref $_[0] || ref $_[0] eq "ARRAY" )) { # one positional param %args = (check => shift); } else { %args = validate ( @_, { # named params check => 1, warning => 0, critical => 0, } ); } # in order of preference, get warning and critical from # 1. explicit arguments to check_threshold # 2. previously explicitly set threshold object # 3. implicit options from Getopts object if ( exists $args{warning} || exists $args{critical} ) { $self->set_thresholds( warning => $args{warning}, critical => $args{critical}, ); } elsif ( defined $self->threshold ) { # noop } elsif ( defined $self->opts ) { $self->set_thresholds( warning => $self->opts->warning, critical => $self->opts->critical, ); } else { return UNKNOWN; } return $self->threshold->get_status($args{check}); } # top level interface to my Monitoring::Plugin::Getopt object sub add_arg { my $self = shift; $self->opts->arg(@_) if $self->_check_for_opts; } sub getopts { my $self = shift; $self->opts->getopts(@_) if $self->_check_for_opts; } sub _check_for_opts { my $self = shift; croak "You have to supply a 'usage' param to Monitoring::Plugin::new() if you want to use Getopts from your Monitoring::Plugin object." unless ref $self->opts() eq 'Monitoring::Plugin::Getopt'; return $self; } # ------------------------------------------------------------------------- # MP::Functions::check_messages helpers and wrappers sub add_message { my $self = shift; my ($code, @messages) = @_; croak "Invalid error code '$code'" unless defined($ERRORS{uc $code}) || defined($STATUS_TEXT{$code}); # Store messages using strings rather than numeric codes $code = $STATUS_TEXT{$code} if $STATUS_TEXT{$code}; $code = lc $code; croak "Error code '$code' not supported by add_message" if $code eq 'unknown' || $code eq 'dependent'; $self->messages($code, []) unless $self->messages->{$code}; push @{$self->messages->{$code}}, @messages; } sub check_messages { my $self = shift; my %args = @_; # Add object messages to any passed in as args for my $code (qw(critical warning ok)) { my $messages = $self->messages->{$code} || []; if ($args{$code}) { unless (ref $args{$code} eq 'ARRAY') { if ($code eq 'ok') { $args{$code} = [ $args{$code} ]; } else { croak "Invalid argument '$code'" } } push @{$args{$code}}, @$messages; } else { $args{$code} = $messages; } } Monitoring::Plugin::Functions::check_messages(%args); } # ------------------------------------------------------------------------- 1; #vim:et:sw=4 __END__ =head1 NAME Monitoring::Plugin - A family of perl modules to streamline writing Naemon, Nagios, Icinga or Shinken (and compatible) plugins. =head1 SYNOPSIS # Constants OK, WARNING, CRITICAL, and UNKNOWN are exported by default # See also Monitoring::Plugin::Functions for a functional interface use Monitoring::Plugin; # Constructor $np = Monitoring::Plugin->new; # OR $np = Monitoring::Plugin->new( shortname => "PAGESIZE" ); # OR # use Monitoring::Plugin::Getopt to process the @ARGV command line options: # --verbose, --help, --usage, --timeout and --host are defined automatically. $np = Monitoring::Plugin->new( usage => "Usage: %s [ -v|--verbose ] [-H ] [-t ] " . "[ -c|--critical= ] [ -w|--warning= ]", ); # add valid command line options and build them into your usage/help documentation. $np->add_arg( spec => 'warning|w=s', help => '-w, --warning=INTEGER:INTEGER . See ' . 'https://www.monitoring-plugins.org/doc/guidelines.html#THRESHOLDFORMAT ' . 'for the threshold format. ', ); # Parse @ARGV and process standard arguments (e.g. usage, help, version) $np->getopts; # Exit/return value methods - plugin_exit( CODE, MESSAGE ), # plugin_die( MESSAGE, [CODE]) $page = retrieve_page($page1) or $np->plugin_exit( UNKNOWN, "Could not retrieve page" ); # Return code: 3; # output: PAGESIZE UNKNOWN - Could not retrieve page test_page($page) or $np->plugin_exit( CRITICAL, "Bad page found" ); # plugin_die() is just like plugin_exit(), but return code defaults # to UNKNOWN $page = retrieve_page($page2) or $np->plugin_die( "Could not retrieve page" ); # Return code: 3; # output: PAGESIZE UNKNOWN - Could not retrieve page # Threshold methods $code = $np->check_threshold( check => $value, warning => $warning_threshold, critical => $critical_threshold, ); $np->plugin_exit( $code, "Threshold check failed" ) if $code != OK; # Message methods (EXPERIMENTAL AND SUBJECT TO CHANGE) - # add_message( CODE, $message ); check_messages() for (@collection) { if (m/Error/) { $np->add_message( CRITICAL, $_ ); } else { $np->add_message( OK, $_ ); } } ($code, $message) = $np->check_messages(); plugin_exit( $code, $message ); # If any items in collection matched m/Error/, returns CRITICAL and # the joined set of Error messages; otherwise returns OK and the # joined set of ok messages # Perfdata methods $np->add_perfdata( label => "size", value => $value, uom => "kB", threshold => $threshold, ); $np->add_perfdata( label => "time", ... ); $np->plugin_exit( OK, "page size at http://... was ${value}kB" ); # Return code: 0; # output: PAGESIZE OK - page size at http://... was 36kB \ # | size=36kB;10:25;25: time=... =head1 DESCRIPTION Monitoring::Plugin and its associated Monitoring::Plugin::* modules are a family of perl modules to streamline writing Monitoring plugins. The main end user modules are Monitoring::Plugin, providing an object-oriented interface to the entire Monitoring::Plugin::* collection, and Monitoring::Plugin::Functions, providing a simpler functional interface to a useful subset of the available functionality. The purpose of the collection is to make it as simple as possible for developers to create plugins that conform the Monitoring Plugin guidelines (https://www.monitoring-plugins.org/doc/guidelines.html). =head2 EXPORTS Nagios status code constants are exported by default: OK WARNING CRITICAL UNKNOWN DEPENDENT The following variables are also exported on request: =over 4 =item %ERRORS A hash mapping error strings ("CRITICAL", "UNKNOWN", etc.) to the corresponding status code. =item %STATUS_TEXT A hash mapping status code constants (OK, WARNING, CRITICAL, etc.) to the corresponding error string ("OK", "WARNING, "CRITICAL", etc.) i.e. the reverse of %ERRORS. =back =head2 CONSTRUCTOR Monitoring::Plugin->new; Monitoring::Plugin->new( shortname => 'PAGESIZE' ); Monitoring::Plugin->new( usage => "Usage: %s [ -v|--verbose ] [-H ] [-t ] [ -c|--critical= ] [ -w|--warning= ] ", version => $VERSION, blurb => $blurb, extra => $extra, url => $url, license => $license, plugin => basename $0, timeout => 15, ); Instantiates a new Monitoring::Plugin object. Accepts the following named arguments: =over 4 =item shortname The 'shortname' for this plugin, used as the first token in the plugin output by the various exit methods. Default: uc basename $0. =item usage ("Usage: %s --foo --bar") Passing a value for the usage() argument makes Monitoring::Plugin instantiate its own C object so you can start doing command line argument processing. See L for more about "usage" and the following options: =item version =item url =item blurb =item license =item extra =item plugin =item timeout =back =head2 GETTER/SETTER The following internal variables can be retrieved or set by calling a method with the respective name. Expect for C, don't change values unless you know what you're doing. Examples: use Data::Dumper; print Dumper($plugin->perfdata); $plugin->shortname('DifferentName'); =over =item shortname =item perfdata =item messages =item opts =item threshold =back =head2 OPTION HANDLING METHODS C provides these methods for accessing the functionality in C. =over 4 =item add_arg Examples: # Define --hello argument (named parameters) $plugin->add_arg( spec => 'hello=s', help => "--hello\n Hello string", required => 1, ); # Define --hello argument (positional parameters) # Parameter order is 'spec', 'help', 'default', 'required?' $plugin->add_arg('hello=s', "--hello\n Hello string", undef, 1); See L for more details. =item getopts() Parses and processes the command line options you've defined, automatically doing the right thing with help/usage/version arguments. See L for more details. =item opts() Assuming you've instantiated it by passing 'usage' to new(), opts() returns the Monitoring::Plugin object's C object, with which you can do lots of great things. E.g. if ( $plugin->opts->verbose ) { print "yah yah YAH YAH YAH!!!"; } # start counting down to timeout alarm $plugin->opts->timeout; your_long_check_step_that_might_time_out(); # access any of your custom command line options, # assuming you've done these steps above: # $plugin->add_arg('my_argument=s', '--my_argument [STRING]'); # $plugin->getopts; print $plugin->opts->my_argument; Again, see L. =back =head2 EXIT METHODS =over 4 =item plugin_exit( , $message ) Exit with return code CODE, and a standard nagios message of the form "SHORTNAME CODE - $message". =item nagios_exit( , $message ) Alias for plugin_exit(). Deprecated. =item plugin_die( $message, [] ) Same as plugin_exit(), except that CODE is optional, defaulting to UNKNOWN. NOTE: exceptions are not raised by default to calling code. Set C<$_use_die> flag if this functionality is required (see test code). =item nagios_die( $message, [] ) Alias for plugin_die(). Deprecated. =item die( $message, [] ) Alias for plugin_die(). Deprecated. =item max_state, max_state_alt These are wrapper function for Monitoring::Plugin::Functions::max_state and Monitoring::Plugin::Functions::max_state_alt. =back =head2 THRESHOLD METHODS These provide a top level interface to the C module; for more details, see L and L. =over 4 =item check_threshold( $value ) =item check_threshold( check => $value, warning => $warn, critical => $crit ) Evaluates $value against the thresholds and returns OK, CRITICAL, or WARNING constant. The thresholds may be: 1. explicitly set by passing 'warning' and/or 'critical' parameters to C, or, 2. explicitly set by calling C before C, or, 3. implicitly set by command-line parameters -w, -c, --critical or --warning, if you have run C<< $plugin->getopts() >>. You can specify $value as an array of values and each will be checked against the thresholds. The return value is ready to pass to C , e . g ., $p->plugin_exit( return_code => $p->check_threshold($result), message => " sample result was $result" ); =item set_thresholds(warning => "10:25", critical => "~:25") Sets the acceptable ranges and creates the plugin's Monitoring::Plugins::Threshold object. See https://www.monitoring-plugins.org/doc/guidelines.html#THRESHOLDFORMAT for details and examples of the threshold format. =item threshold() Returns the object's C object, if it has been defined by calling set_thresholds(). You can pass a new Threshold object to it to replace the old one too, but you shouldn't need to do that from a plugin script. =back =head2 MESSAGE METHODS EXPERIMENTAL AND SUBJECT TO CHANGE add_messages and check_messages are higher-level convenience methods to add and then check a set of messages, returning an appropriate return code and/or result message. They are equivalent to maintaining a set of @critical, @warning, and and @ok message arrays (add_message), and then doing a final if test (check_messages) like this: if (@critical) { plugin_exit( CRITICAL, join(' ', @critical) ); } elsif (@warning) { plugin_exit( WARNING, join(' ', @warning) ); } else { plugin_exit( OK, join(' ', @ok) ); } =over 4 =item add_message( , $message ) Add a message with CODE status to the object. May be called multiple times. The messages added are checked by check_messages, following. Only CRITICAL, WARNING, and OK are accepted as valid codes. =item check_messages() Check the current set of messages and return an appropriate nagios return code and/or a result message. In scalar context, returns only a return code; in list context returns both a return code and an output message, suitable for passing directly to plugin_exit() e.g. $code = $np->check_messages; ($code, $message) = $np->check_messages; check_messages returns CRITICAL if any critical messages are found, WARNING if any warning messages are found, and OK otherwise. The message returned in list context defaults to the joined set of error messages; this may be customised using the arguments below. check_messages accepts the following named arguments (none are required): =over 4 =item join => SCALAR A string used to join the relevant array to generate the message string returned in list context i.e. if the 'critical' array @crit is non-empty, check_messages would return: join( $join, @crit ) as the result message. Default: ' ' (space). =item join_all => SCALAR By default, only one set of messages are joined and returned in the result message i.e. if the result is CRITICAL, only the 'critical' messages are included in the result; if WARNING, only the 'warning' messages are included; if OK, the 'ok' messages are included (if supplied) i.e. the default is to return an 'errors-only' type message. If join_all is supplied, however, it will be used as a string to join the resultant critical, warning, and ok messages together i.e. all messages are joined and returned. =item critical => ARRAYREF Additional critical messages to supplement any passed in via add_message(). =item warning => ARRAYREF Additional warning messages to supplement any passed in via add_message(). =item ok => ARRAYREF | SCALAR Additional ok messages to supplement any passed in via add_message(). =back =back =head2 PERFORMANCE DATA METHODS =over 4 =item add_perfdata( label => "size", value => $value, uom => "kB", threshold => $threshold ) Add a set of performance data to the object. May be called multiple times. The performance data is included in the standard plugin output messages by the various exit methods. See the Monitoring::Plugin::Performance documentation for more information on performance data and the various field definitions, as well as the relevant section of the Monitoring Plugin guidelines (https://www.monitoring-plugins.org/doc/guidelines.html#AEN202). =back =head1 EXAMPLES "Enough talk! Show me some examples!" See the file 'check_stuff.pl' in the 't' directory included with the Monitoring::Plugin distribution for a complete working example of a plugin script. =head1 VERSIONING The Monitoring::Plugin::* modules are currently experimental and so the interfaces may change up until Monitoring::Plugin hits version 1.0, although every attempt will be made to keep them as backwards compatible as possible. =head1 SEE ALSO See L for a simple functional interface to a subset of the available Monitoring::Plugin functionality. See also L, L, L, L, and L. The Monitoring Plugin project page is at http://monitoring-plugins.org. =head1 BUGS Please report bugs in these modules to the Monitoring Plugin development team: devel@monitoring-plugins.org. =head1 AUTHOR Maintained by the Monitoring Plugin development team - https://www.monitoring-plugins.org. Originally by Ton Voon, Eton.voon@altinity.comE. =head1 COPYRIGHT AND LICENSE Copyright (C) 2014 by Monitoring Plugin Team Copyright (C) 2006-2014 by Nagios Plugin Development Team This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself, either Perl version 5.8.4 or, at your option, any later version of Perl 5 you may have available. =cut