The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.
#!/usr/bin/perl

eval 'exec /usr/bin/perl  -S $0 ${1+"$@"}'
    if 0; # not running under some shell

use strict;
use warnings FATAL => 'all';

use YATG::SharedStorage;
YATG::SharedStorage->factory(qw(cache results));

use YAML::XS;
use YATG::Config;
use YATG::Callback;

use File::stat;
use Symbol;
use Readonly;
use Time::HiRes qw(time sleep);

use DBI;
use SNMP;
use SNMP::Effective;
die "must not use SNMP::Effective 1.0\n"
    if $SNMP::Effective::VERSION == 1.0;
use Net::Netmask;
use Regexp::Common 'net';

use Sys::Syslog 0.25;
use Log::Dispatch::Syslog;
use Data::Printer;

use POSIX;
use FindBin;
use File::Basename;
use File::Spec::Functions;

my $config = YATG::Config->parse(@ARGV) || die "failed to load config\n";
Readonly my $INTERVAL => $ENV{YATG_INTERVAL} || $config->yatg->{'interval'};
Readonly my $POLLERS  => $config->yatg->{'max_pollers'};
Readonly my $TIMEOUT  => $config->yatg->{'timeout'};
Readonly my $DEBUG    => $ENV{YATG_DEBUG} || $config->yatg->{'debug'};
$ENV{YATG_DEBUG} = $DEBUG; # copy from config for use in backends

# p $config if $ENV{YATG_DEBUG};

my $logger = Log::Dispatch::Syslog->new($config->log_dispatch_syslog);
sub to_log {
    my $msg = shift;
    return unless $msg;

    print STDERR $msg if $DEBUG;
    $msg =~ s/^/[$$] /gm;

    $logger->log(
        level => $config->log_dispatch_syslog->{'min_level'},
        message => $msg
    );
}

Readonly my @modules => qw( Disk RPC Memcached NSCA STDOUT );
Readonly my @basic_oids => (
        'ifDescr'       => '.1.3.6.1.2.1.2.2.1.2',
        'ifAdminStatus' => '.1.3.6.1.2.1.2.2.1.7',
);

my $results = YATG::SharedStorage->results({});
my $cache   = YATG::SharedStorage->cache({
    oid_for  => { @basic_oids },
    leaf_for => { reverse @basic_oids },
});

# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

# POSIX unmasks the sigprocmask properly
my $sigset = POSIX::SigSet->new;
my $action = POSIX::SigAction->new(
    'hup_handler', $sigset, &POSIX::SA_NODEFER
);
POSIX::sigaction(&POSIX::SIGHUP, $action);

chomp(my $command = (qx/ps -o args $$/)[1]);
sub hup_handler {
    to_log("SIGHUP received, restarting...\n");
    exec ($command) or die "Couldn't restart $0: $!\n";
}

# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# rebuild device cache (also on sigHUP)
to_log("YATG Loaded.\n") if $DEBUG;

sub get_hosts {
    # list of all device ips
    my $dbh = DBI->connect(
        @{$config->yatg->{'dbi_connect'}},
        {PrintError => 1, RaiseError => 1, AutoCommit => 1}
    ) or die $DBI::errstr;

    $cache->{devices} = $dbh->selectcol_arrayref(
        $config->yatg->{'dbi_ip_query'});

    # build resolver table for host names for device IPs
    if ($config->yatg->{'dbi_host_query'}) {
        $cache->{'host_for'} = { map {$_->{ip} => $_->{name}} @{
            $dbh->selectall_arrayref(
              $config->yatg->{'dbi_host_query'},
              { Slice => {} },
            )}
        };
    }

    # p $cache->{host_for} if exists $cache->{host_for};

    # build filter list of interfaces for ifIndex items for each device
    if ($config->yatg->{'dbi_interfaces_query'}) {
        foreach my $ip (@{$cache->{devices}}) {
            $cache->{'interfaces_for'}->{$ip} = {
                map { $_ => 1 } @{
                  $dbh->selectcol_arrayref(
                    $config->yatg->{'dbi_interfaces_query'},
                    {}, $ip)
                }
            };

            # a device with no interfaces means allow all interfaces
            delete $cache->{'interfaces_for'}->{$ip}
              if 0 == scalar keys %{ $cache->{'interfaces_for'}->{$ip} };
        }
    }

    # p $cache->{'interfaces_for'} if exists $cache->{'interfaces_for'};
    $dbh->disconnect;
}

# initialise hosts list
&get_hosts;

# work out what OIDs to grab, and how
SNMP::addMibDirs(@{$config->yatg->{'mibdirs'}});
SNMP::loadModules('ALL');
my %modhash = map{lc($_) => $_} @modules;
my @to_load = ();

foreach my $leaf (keys %{$config->yatg->{'oids'}}) {
    my %tokens = map {lc($_) => 1} @{$config->yatg->{'oids'}->{$leaf}};

    # difference or counter
    $cache->{oids}->{$leaf}->{diff} = $tokens{diff} ? 1 : 0;

    # store module to load
    my (%union, %isect);
    foreach my $e (keys %modhash, keys %tokens) { $union{$e}++ && $isect{$e}++ }
    my @store_list = keys %isect;

    $cache->{oids}->{$leaf}->{store_list} = [ map {$modhash{$_}} @store_list ];
    push @to_load, @{$cache->{oids}->{$leaf}->{store_list}};

    # transate leaf to OID
    $cache->{oid_for}->{$leaf} = SNMP::translateObj($leaf)
        or die "Failed to translate leaf $leaf to OID\n";
}

sub get_community {
    my $ip = shift;

    # shortcut if there is only one community supplied
    if (scalar @{$config->yatg->{'communities'}} == 1) {
        $cache->{community_for}->{$ip}
            = $config->yatg->{'communities'}->[0];
        return;
    }

    foreach my $c (@{$config->yatg->{'communities'}}) {
        my $sess = SNMP::Session->new(
            DestHost  => $ip,
            Version   => 1,
            Timeout   => 500000,
            Community => $c,
        );
        my $val = $sess->get('sysUpTime.0');

        # failure, go to next community
        next if !$val or $sess->{ErrorNum} or $sess->{ErrorStr};

        # success, use this community and go to next device
        $cache->{community_for}->{$ip} = $c;
        # to_log("commnity for $ip discovered as $c\n") if $DEBUG;
        last;
    }

    to_log("warning: Failed to find community for [$ip] !\n")
      if !exists $cache->{community_for}->{$ip};
}

sub prime_cache {
    to_log("Generating oid/host map... \n") if $DEBUG;

    foreach my $leaf (keys %{$config->yatg->{'oids'}}) {
        my %tokens = map {lc($_) => 1} @{$config->yatg->{'oids'}->{$leaf}};

        my $oid = $cache->{oid_for}->{$leaf};
        $cache->{leaf_for}->{ $oid } = $leaf;

        # prune IPs for each OID
        my @nets_incl = grep {$_ = Net::Netmask->new2($_)}
                        grep {m#^$RE{net}{IPv4}(?:/\d+)?$#} keys %tokens;
        my @nets_excl = grep {$_ = Net::Netmask->new2($_)}
                        grep {m#^!$RE{net}{IPv4}(?:/\d+)?$#} keys %tokens;

        if (scalar @nets_incl == 0 and scalar @nets_excl == 0) {
            # short-circuit
            map { $cache->{get_for}->{$_}->{ $oid } = 1 }
                @{$cache->{devices}};
        }
        else {
            foreach my $ip (@{$cache->{devices}}) {
                $cache->{get_for}->{$ip}->{ $oid } = 1
                    if  ((scalar grep { $_->match($ip) } @nets_incl) > 0
                        and (scalar grep { $_->match($ip) } @nets_excl) == 0)
                    or  (scalar @nets_incl == 0
                        and (scalar grep { $_->match($ip) } @nets_excl) == 0);
            }
        }

        # indexed oid?
        if ($tokens{ifindex}) {
            $cache->{oids}->{$leaf}->{indexer} = 'iid';
            foreach my $ip (
                grep { exists $cache->{get_for}->{$_}->{ $oid } }
                    @{$cache->{devices}} ) {

                $cache->{get_for}->{$ip}->{ $cache->{oid_for}->{ifDescr} } = 1;
                $cache->{get_for}->{$ip}->{ $cache->{oid_for}->{ifAdminStatus} } = 1;
                $cache->{$ip}->{build_ifindex} = 1;
            }
        }
        else {
            $cache->{oids}->{$leaf}->{indexer} = 0;
        }
    }

    to_log("Getting communities... \n") if $DEBUG;

    # connect to each device and find out its community,
    # shortcut if there is only one community supplied
    if (scalar @{$config->yatg->{'communities'}} == 1) {
        map {$cache->{community_for}->{$_}
                    = $config->yatg->{'communities'}->[0]}
            @{$cache->{devices}};
    }
    else {
        # build lookup table for snmp communities
        if ($config->yatg->{'dbi_community_query'}) {
            my $dbh = DBI->connect(
                @{$config->yatg->{'dbi_connect'}},
                {PrintError => 1, RaiseError => 1, AutoCommit => 1}
            ) or die $DBI::errstr;

            $cache->{'community_for'} = { map {$_->{ip} => $_->{snmp_community}} @{
                $dbh->selectall_arrayref(
                  $config->yatg->{'dbi_community_query'},
                  { Slice => {} },
                )}
            };

            # p $cache->{'community_for'} if exists $cache->{'community_for'};
            $dbh->disconnect;
        }
        else {
            foreach my $ip (@{$cache->{devices}}) {
                # to_log("finding community string for $ip...\n") if $DEBUG;
                &get_community($ip);
            }
        }
    }
}

sub check_newhosts_watch {
    return unless exists $config->{'yatg'}->{'newhosts_watch'};

    my $file = $config->{'yatg'}->{'newhosts_watch'};
    return unless length $file and -e $file;

    my $mtime = stat($file)->mtime;
    return if $mtime < (time - $INTERVAL);

    to_log("newhosts_watch updated - getting new hosts...\n");
    &get_hosts;
    &prime_cache;
}

# load storage module(s)
foreach my $module (@to_load) {
    die "Request to load undefined module\n" if !defined $module;
    eval "require YATG::Store::$module"
        or die "Failed to load $module store module: $@\n";
}

&prime_cache;
to_log(sprintf "Initial cache build took %.3f seconds.\n", (time - $^T));

# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

while (1) {
    my $naptime = $INTERVAL - (time % $INTERVAL);
    to_log("Sleeping for $naptime seconds...\n") if $DEBUG;
    sleep $naptime;

    # see if we need to refresh the cache
    &check_newhosts_watch;

    my $se = SNMP::Effective->new(
        MaxSessions   => $POLLERS,
        MasterTimeout => $TIMEOUT,
    );

    foreach my $ip (@{$cache->{devices}}) {
        next unless scalar (keys %{$cache->{get_for}->{$ip}}) > 0;
        next unless $cache->{community_for}->{$ip};

        $se->add(
            desthost => $ip,
            args      => {
                Community => $cache->{community_for}->{$ip},
                Timeout   => 5000000,
            },
            callback => \&YATG::Callback::snmp_callback,
            walk     => [ keys %{$cache->{get_for}->{$ip}} ],
        );
    }

    my $exec = time;
    $se->execute;
    to_log(sprintf "Execute run took %.3f seconds.\n", (time - $exec));

    my $update = time;
    my $stamp = floor($update - ($update % $INTERVAL));
    (($stamp % $INTERVAL) == 0) or die "yatg: FATAL: time is shifting!\n";

    # send gathered data from this run to storage
    foreach my $mod (keys %$results) {
        &{*{Symbol::qualify_to_ref('store',"YATG::Store::$mod")}}
            ({%$config}, $stamp, $results->{$mod});
    }

    to_log(sprintf "Remote update took %.3f seconds.\n", (time - $update));

    $results = YATG::SharedStorage->results({});
    last if $ENV{YATG_SINGLE_RUN};
}

# ABSTRACT: Fast SNMP data poller daemon, with storage and graphing
# PODNAME: YATG


__END__
=pod

=head1 NAME

YATG - Fast SNMP data poller daemon, with storage and graphing

=head1 VERSION

version 5.20002

=head1 DESCRIPTION

YATG is a daemon (background process) which at intervals wakes up and polls
network devices for SNMP data, and then stores or prints that data. In this
distribution are also included examples for presenting simple CGI web pages
with graphs.

YATG is flexible, efficient and powerful. It can poll a large number of
devices with thousands of ports in just a few seconds. The configuration is
very simple, and the defaults sane (it's designed for sysadmins, after all).

You can use YATG both for historical logging, such as traffic counters on
ports, as well as short-term monitoring which might feed into, say, Nagios.
Wherever possible, data is translated to human-friendly formats for storage,
such as using Leaf Names instead of OIDs, translated values (C<up>, C<down>,
etc) and device port names rather than SNMP Interface Indexes.

=head1 How Does It Work?

At startup, C<yatg_updater> loads its configuration from local files and a
database, performs some basic SNMP connections to build a cache about device
capabilities and so on, and then goes to sleep.

Periodically, as determined by the configuration, C<yatg_updater> wakes up and
polls all devices, then stores results, again according to instructions in the
configuration.

If you have only the essential dependencies installed (see below) then you can
only output results to STDOUT. With other modules, you have more options such
as local or remote disk, or memcached based storage.

C<yatg_updater> will re-load all its configuration if given a HUP signal. If
you run the daemon persistently (for example with C<daemontools>) then a cron
job once a day is a good way to refresh the configuration. There is reference
to this in one of the bundled example files.

=head1 What's in this distribution

=over 4

=item C<yatg_updater>

This is the main application, designed to be run persistently. It does not
accept any input and only produces output when in debugging mode. It is a
smart wrapper for the L<SNMP::Effective> module.

=item YATG::Store family of modules

These are modules which take the SNMP poll results and store them to either
local Disk, a Memcached server, the disk on a remote networked server, or
Nagios via NSCA.

=item YATG::Retrieve family of modules

These are modules which read stored results back to you, for a given time
window. The data can be retrieved from local Disk, a Memcached server, or the
disk on a remote networked server.

=item C<yatg_trim>

If using the Disk Store backend for results, eventually you'll want to save
space by deleting old data. This script understands the backend file format
and, given a duration, removes that amount of historical data from the file.

=item RPC::Serialized handlers

If storing and/or retrieving on a remote networked server, it should run an
instance of L<RPC::Serialized>, and these are the RPC Handlers for that server
(see that module's documentation for further details).

=item CGI

For the special case of viewing graphs of disk-based poll results for switch
port traffic counters, there is are two CGI scripts. One is a wrapper which
presents an HTML page embedded with PNG images created from the other script.

=item Examples

The C<examples/> folder includes a copy of each of the files you should need
for a complete deployment of YATG. Obviously some of them contain dummy data.

=back

=head1 Where to go from here

To begin with, you probably want to see how to configure C<yatg_updater> in
L<YATG::Config>.

Alongside that, there are examples of all the files you should need to
install, in the C<examples/> folder of this distribution.

Each of the Store and Retrieve modules might have additional Perl module
dependencies (i.e. from CPAN) - see the relevant docs for more details.

=head1 LOGGING and TESTING

This module uses "Log::Dispatch::Syslog" for logging, and by default will log
timing data to your system's syslog service. More information is provided in
the L<YATG::Config> documentation.

To run in debug mode, where timing data is output to standard out rather than
syslog, set the environment variable C<YATG_DEBUG> to a true value.

To run the poller just once, set the C<YATG_SINGLE_RUN> environment variable
to a true value. This is great for development. It makes C<yatg_updater> load
its configuration, generate the device hints cache, sleep and then run just
one poll cycle before exiting.

To override the interval between polling runs, set the C<YATG_INTERVAL>
environment variable to a number of seconds.

For example:

 YATG_DEBUG=1 YATG_SINGLE_RUN=1 /usr/bin/yatg_updater /etc/yatg.yml

=head1 SEE ALSO

=over 4

=item L<SNMP::Effective>

This system uses SNMP::Effective at its core do the polling.

=item L<RPC::Serialized>

Store polled data on another server using RPC::Serialized.

=back

=head1 AUTHOR

Oliver Gorwits <oliver@cpan.org>

=head1 COPYRIGHT AND LICENSE

This software is copyright (c) 2015 by University of Oxford.

This is free software; you can redistribute it and/or modify it under
the same terms as the Perl 5 programming language system itself.

=cut