Queen.pm

#
# You may distribute this module under the same terms as perl itself

=pod 

=head1 NAME

  Bio::EnsEMBL::Hive::Queen

=head1 SYNOPSIS

  The Queen of the Hive based job control system

=head1 DESCRIPTION

  The Queen of the Hive based job control system is responsible to 'birthing' the
  correct number of workers of the right type so that they can find jobs to do.
  It will also free up jobs of Workers that died unexpectantly so that other workers
  can claim them to do.

  Hive based processing is a concept based on a more controlled version
  of an autonomous agent type system.  Each worker is not told what to do
  (like a centralized control system - like the current pipeline system)
  but rather queries a central database for jobs (give me jobs).

  Each worker is linked to an analysis_id, registers its self on creation
  into the Hive, creates a RunnableDB instance of the Analysis->module,
  gets $analysis->stats->batch_size jobs from the job table, does its work,
  creates the next layer of job entries by interfacing to
  the DataflowRuleAdaptor to determine the analyses it needs to pass its
  output data to and creates jobs on the next analysis database.
  It repeats this cycle until it has lived its lifetime or until there are no
  more jobs left.
  The lifetime limit is just a safety limit to prevent these from 'infecting'
  a system.

  The Queens job is to simply birth Workers of the correct analysis_id to get the
  work down.  The only other thing the Queen does is free up jobs that were
  claimed by Workers that died unexpectantly so that other workers can take
  over the work.

  The Beekeeper is in charge of interfacing between the Queen and a compute resource
  or 'compute farm'.  Its job is to query Queens if they need any workers and to
  send the requested number of workers to open machines via the runWorker.pl script.
  It is also responsible for interfacing with the Queen to identify worker which died
  unexpectantly.

=head1 CONTACT

  Please contact ehive-users@ebi.ac.uk mailing list with questions/suggestions.

=head1 APPENDIX

  The rest of the documentation details each of the object methods. 
  Internal methods are usually preceded with a _

=cut


package Bio::EnsEMBL::Hive::Queen;

use strict;
use POSIX;
use Bio::EnsEMBL::Utils::Argument;
use Bio::EnsEMBL::Utils::Exception;

use Bio::EnsEMBL::Hive::Utils ('destringify', 'dir_revhash');  # import 'destringify()' and 'dir_revhash()'
use Bio::EnsEMBL::Hive::AnalysisJob;
use Bio::EnsEMBL::Hive::Worker;

use base ('Bio::EnsEMBL::Hive::DBSQL::ObjectAdaptor');


sub default_table_name {
    return 'worker';
}


sub default_insertion_method {
    return 'INSERT';
}


sub object_class {
    return 'Bio::EnsEMBL::Hive::Worker';
}


############################
#
# PUBLIC API
#
############################


=head2 create_new_worker

  Description: Creates an entry in the worker table,
               populates some non-storable attributes
               and returns a Worker object based on that insert.
               This guarantees that each worker registered in this Queen's hive is properly registered.
  Returntype : Bio::EnsEMBL::Hive::Worker
  Caller     : runWorker.pl

=cut

sub create_new_worker {
    my ($self, @args) = @_;

    my ($meadow_type, $meadow_name, $process_id, $exec_host, $resource_class_id, $resource_class_name,
        $no_write, $debug, $worker_log_dir, $hive_log_dir, $job_limit, $life_span, $no_cleanup, $retry_throwing_jobs, $compile_module_once) =

    rearrange([qw(meadow_type meadow_name process_id exec_host resource_class_id resource_class_name
                no_write debug worker_log_dir hive_log_dir job_limit life_span no_cleanup retry_throwing_jobs compile_module_once) ], @args);

    if($resource_class_name) {
        my $rc = $self->db->get_ResourceClassAdaptor->fetch_by_name($resource_class_name)
            or die "resource_class with name='$resource_class_name' could not be fetched from the database";

        $resource_class_id = $rc->dbID;
    }
    $resource_class_id ||= 1;

    my $sql = q{INSERT INTO worker (born, last_check_in, meadow_type, meadow_name, host, process_id, resource_class_id)
              VALUES (CURRENT_TIMESTAMP, CURRENT_TIMESTAMP, ?, ?, ?, ?, ?)};

    my $sth = $self->prepare($sql);
    $sth->execute($meadow_type, $meadow_name, $exec_host, $process_id, $resource_class_id);
    my $worker_id = $self->dbc->db_handle->last_insert_id(undef, undef, 'worker', 'worker_id')
        or die "Could not create a new worker";
    $sth->finish;

    if($hive_log_dir or $worker_log_dir) {
        my $dir_revhash = dir_revhash($worker_id);
        $worker_log_dir ||= $hive_log_dir .'/'. ($dir_revhash ? "$dir_revhash/" : '') .'worker_id_'.$worker_id;

            # Note: the following die-message will not reach the log files for circular reason!
        system("mkdir -p $worker_log_dir") && die "Could not create '$worker_log_dir' because: $!";

        my $sth_add_log = $self->prepare( "UPDATE worker SET log_dir=? WHERE worker_id=?" );
        $sth_add_log->execute($worker_log_dir, $worker_id);
        $sth_add_log->finish;
    }

    my $worker = $self->fetch_by_dbID($worker_id)
        or die "Could not fetch worker with dbID=$worker_id";

    $worker->init;

    if($job_limit) {
      $worker->job_limit($job_limit);
      $worker->life_span(0);
    }

    $worker->life_span($life_span * 60)                 if($life_span);

    $worker->execute_writes(0)                          if($no_write);

    $worker->perform_cleanup(0)                         if($no_cleanup);

    $worker->debug($debug)                              if($debug);

    $worker->retry_throwing_jobs($retry_throwing_jobs)  if(defined $retry_throwing_jobs);

    $worker->compile_module_once($compile_module_once)  if(defined $compile_module_once);

    return $worker;
}


=head2 specialize_new_worker

  Description: If analysis_id or logic_name is specified it will try to specialize the Worker into this analysis.
               If not specified the Queen will analyze the hive and pick the most suitable analysis.
  Returntype : Bio::EnsEMBL::Hive::Worker
  Caller     : runWorker.pl

=cut

sub specialize_new_worker {
    my ($self, $worker, @args) = @_;

    my ($analysis_id, $logic_name, $job_id, $force) =
        rearrange([qw(analysis_id logic_name job_id force) ], @args);

    if( scalar( grep {defined($_)} ($analysis_id, $logic_name, $job_id) ) > 1) {
        die "At most one of the options {-analysis_id, -logic_name, -job_id} can be set to pre-specialize a Worker";
    }

    my ($job, $analysis, $stats);
    my $analysis_stats_adaptor = $self->db->get_AnalysisStatsAdaptor;

    if($job_id or $analysis_id or $logic_name) {    # probably pre-specialized by running a runWorker.pl by hand

        if($job_id) {
            print "resetting and fetching job for job_id '$job_id'\n";

            my $job_adaptor = $self->db->get_AnalysisJobAdaptor;
            $job_adaptor->reset_job_by_dbID($job_id); 

            $job = $job_adaptor->fetch_by_dbID($job_id)
                or die "job with dbID='$job_id' could not be fetched from the database";

            $analysis_id = $job->analysis_id;
        }

        if($logic_name) {
            $analysis = $self->db->get_AnalysisAdaptor->fetch_by_logic_name($logic_name)
                or die "analysis with name='$logic_name' could not be fetched from the database";

            $analysis_id = $analysis->dbID;

        } elsif($analysis_id) {
            $analysis = $self->db->get_AnalysisAdaptor->fetch_by_dbID($analysis_id)
                or die "analysis with dbID='$analysis_id' could not be fetched from the database";
        }

        if($worker->resource_class_id != $analysis->resource_class_id) {
            die "resource_class of analysis ".$analysis->logic_name." is incompatible with this Worker's resource_class";
        }

        $stats = $analysis_stats_adaptor->fetch_by_analysis_id($analysis_id);
        $self->safe_synchronize_AnalysisStats($stats);

        unless($job or $force) {    # do we really need to run this analysis?
            if($self->get_hive_current_load() >= 1.1) {
                die "Hive is overloaded, can't create a worker";
            }
            if($stats->status eq 'BLOCKED') {
                die "Analysis is BLOCKED, can't create workers";
            }
            if($stats->num_required_workers <= 0) {
                die "Analysis does not require workers at the moment";
            }
            if($stats->status eq 'DONE') {
                die "Analysis is DONE, don't need to create workers";
            }
        }

    } else {    # probably scheduled by beekeeper.pl

        $stats = $self->suggest_analysis_to_specialize_by_rc_id($worker->resource_class_id)
            or die "Scheduler failed to pick an analysis for the worker";

        print "Scheduler picked analysis with dbID=".$stats->analysis_id." for the worker\n";

        $analysis_id = $stats->analysis_id;
    }

        # now set it in the $worker:

    $worker->analysis_id( $analysis_id );

    if($job) {
        $worker->_specific_job($job);
    } else {    # count it as autonomous worker sharing the load of that analysis:
        $analysis_stats_adaptor->decrease_required_workers($worker->analysis_id);
    }

        # The following increment used to be done only when no specific task was given to the worker,
        # thereby excluding such "special task" workers from being counted in num_running_workers.
        #
        # However this may be tricky to emulate by triggers that know nothing about "special tasks",
        # so I am (temporarily?) simplifying the accounting algorithm.
        #
    unless( $self->db->hive_use_triggers() ) {
        $analysis_stats_adaptor->increase_running_workers($worker->analysis_id);
    }

    my $sth_update_analysis_id = $self->prepare( "UPDATE worker SET analysis_id=? WHERE worker_id=?" );
    $sth_update_analysis_id->execute($worker->analysis_id, $worker->dbID);
    $sth_update_analysis_id->finish;

    $stats->update_status('WORKING');

    return $worker;
}


sub register_worker_death {
  my ($self, $worker) = @_;

  return unless($worker);

  my $cod = $worker->cause_of_death() || 'UNKNOWN';    # make sure we do not attempt to insert a void

  my $sql = "UPDATE worker SET died=CURRENT_TIMESTAMP, last_check_in=CURRENT_TIMESTAMP";
  $sql .= " ,status='DEAD'";
  $sql .= " ,work_done='" . $worker->work_done . "'";
  $sql .= " ,cause_of_death='$cod'";
  $sql .= " WHERE worker_id='" . $worker->dbID ."'";

  $self->dbc->do( $sql );

  unless( $self->db->hive_use_triggers() ) {
      $worker->analysis->stats->adaptor->decrease_running_workers($worker->analysis_id);
  }

  if($cod eq 'NO_WORK') {
    $self->db->get_AnalysisStatsAdaptor->update_status($worker->analysis->dbID, 'ALL_CLAIMED');
  }
  if($cod eq 'UNKNOWN'
  or $cod eq 'MEMLIMIT'
  or $cod eq 'RUNLIMIT'
  or $cod eq 'KILLED_BY_USER') {
    $self->db->get_AnalysisJobAdaptor->release_undone_jobs_from_worker($worker);
  }
  
  # re-sync the analysis_stats when a worker dies as part of dynamic sync system
  if($self->safe_synchronize_AnalysisStats($worker->analysis->stats)->status ne 'DONE') {
    # since I'm dying I should make sure there is someone to take my place after I'm gone ...
    # above synch still sees me as a 'living worker' so I need to compensate for that
    $self->db->get_AnalysisStatsAdaptor->increase_required_workers($worker->analysis->dbID);
  }

}


sub check_for_dead_workers {    # scans the whole Valley for lost Workers (but ignores unreachagle ones)
    my ($self, $valley, $check_buried_in_haste) = @_;

    warn "GarbageCollector:\tChecking for lost Workers...\n";

    my $queen_worker_list           = $self->fetch_overdue_workers(0);
    my %mt_and_pid_to_worker_status = ();
    my %worker_status_counts        = ();
    my %mt_and_pid_to_lost_worker   = ();

    warn "GarbageCollector:\t[Queen:] we have ".scalar(@$queen_worker_list)." Workers alive.\n";

    foreach my $worker (@$queen_worker_list) {

        my $meadow_type = $worker->meadow_type;
        if(my $meadow = $valley->find_available_meadow_responsible_for_worker($worker)) {
            $mt_and_pid_to_worker_status{$meadow_type} ||= $meadow->status_of_all_our_workers;
        } else {
            $worker_status_counts{$meadow_type}{'UNREACHABLE'}++;

            next;   # Worker is unreachable from this Valley
        }

        my $process_id = $worker->process_id;
        if(my $status = $mt_and_pid_to_worker_status{$meadow_type}{$process_id}) { # can be RUN|PEND|xSUSP
            $worker_status_counts{$meadow_type}{$status}++;
        } else {
            $worker_status_counts{$meadow_type}{'LOST'}++;

            $mt_and_pid_to_lost_worker{$meadow_type}{$process_id} = $worker;
        }
    }

        # just a quick summary report:
    foreach my $meadow_type (keys %worker_status_counts) {
        warn "GarbageCollector:\t[$meadow_type Meadow:]\t".join(', ', map { "$_:$worker_status_counts{$meadow_type}{$_}" } keys %{$worker_status_counts{$meadow_type}})."\n\n";
    }

    while(my ($meadow_type, $pid_to_lost_worker) = each %mt_and_pid_to_lost_worker) {
        my $this_meadow = $valley->available_meadow_hash->{$meadow_type};

        if(my $lost_this_meadow = scalar(keys %$pid_to_lost_worker) ) {
            warn "GarbageCollector:\tDiscovered $lost_this_meadow lost $meadow_type Workers\n";

            my $wpid_to_cod = {};
            if($this_meadow->can('find_out_causes')) {
                $wpid_to_cod = $this_meadow->find_out_causes( keys %$pid_to_lost_worker );
                my $lost_with_known_cod = scalar(keys %$wpid_to_cod);
                warn "GarbageCollector:\tFound why $lost_with_known_cod of $meadow_type Workers died\n";
            } else {
                warn "GarbageCollector:\t$meadow_type meadow does not support post-mortem examination\n";
            }

            warn "GarbageCollector:\tReleasing the jobs\n";
            while(my ($process_id, $worker) = each %$pid_to_lost_worker) {
                $worker->cause_of_death( $wpid_to_cod->{$process_id} || 'UNKNOWN');
                $self->register_worker_death($worker);
            }
        }
    }

        # the following bit is completely Meadow-agnostic and only restores database integrity:
    if($check_buried_in_haste) {
        warn "GarbageCollector:\tChecking for Workers buried in haste...\n";
        my $buried_in_haste_list = $self->fetch_all_dead_workers_with_jobs();
        if(my $bih_number = scalar(@$buried_in_haste_list)) {
            warn "GarbageCollector:\tfound $bih_number jobs, reclaiming.\n\n";
            if($bih_number) {
                my $job_adaptor = $self->db->get_AnalysisJobAdaptor();
                foreach my $worker (@$buried_in_haste_list) {
                    $job_adaptor->release_undone_jobs_from_worker($worker);
                }
            }
        } else {
            warn "GarbageCollector:\tfound none\n";
        }
    }
}


    # a new version that both checks in and updates the status
sub check_in_worker {
    my ($self, $worker) = @_;

    $self->dbc->do("UPDATE worker SET last_check_in=CURRENT_TIMESTAMP, status='".$worker->status."', work_done='".$worker->work_done."' WHERE worker_id='".$worker->dbID."'");
}


=head2 reset_job_by_dbID_and_sync

  Arg [1]: int $job_id
  Example: 
    my $job = $queen->reset_job_by_dbID_and_sync($job_id);
  Description: 
    For the specified job_id it will fetch just that job, 
    reset it completely as if it has never run, and return it.  
    Specifying a specific job bypasses the safety checks, 
    thus multiple workers could be running the 
    same job simultaneously (use only for debugging).
  Returntype : none
  Exceptions :
  Caller     : beekeeper.pl

=cut

sub reset_job_by_dbID_and_sync {
    my ($self, $job_id) = @_;

    my $job_adaptor = $self->db->get_AnalysisJobAdaptor;
    $job_adaptor->reset_job_by_dbID($job_id); 

    my $job = $job_adaptor->fetch_by_dbID($job_id); 
    my $stats = $self->db->get_AnalysisStatsAdaptor->fetch_by_analysis_id($job->analysis_id);
    $self->synchronize_AnalysisStats($stats);
}


######################################
#
# Public API interface for beekeeper
#
######################################


    # Note: asking for Queen->fetch_overdue_workers(0) essentially means
    #       "fetch all workers known to the Queen not to be officially dead"
    #
sub fetch_overdue_workers {
    my ($self,$overdue_secs) = @_;

    $overdue_secs = 3600 unless(defined($overdue_secs));

    my $constraint = "status!='DEAD' AND ".
                    ( ($self->dbc->driver eq 'sqlite')
                        ? "(strftime('%s','now')-strftime('%s',last_check_in))>$overdue_secs"
                        : "(UNIX_TIMESTAMP()-UNIX_TIMESTAMP(last_check_in))>$overdue_secs");
    return $self->fetch_all( $constraint );
}


sub fetch_all_dead_workers_with_jobs {
    my $self = shift;

    return $self->fetch_all( "JOIN job j USING(worker_id) WHERE j.status NOT IN ('DONE', 'READY', 'FAILED', 'PASSED_ON') GROUP BY worker_id" );
}


=head2 synchronize_hive

  Arg [1]    : $filter_analysis (optional)
  Example    : $queen->synchronize_hive();
  Description: Runs through all analyses in the system and synchronizes
              the analysis_stats summary with the states in the job 
              and worker tables.  Then follows by checking all the blocking rules
              and blocks/unblocks analyses as needed.
  Exceptions : none
  Caller     : general

=cut

sub synchronize_hive {
  my $self          = shift;
  my $filter_analysis = shift; # optional parameter

  my $start_time = time();

  my $list_of_analyses = $filter_analysis ? [$filter_analysis] : $self->db->get_AnalysisAdaptor->fetch_all;

  print STDERR "\nSynchronizing the hive (".scalar(@$list_of_analyses)." analyses this time):\n";
  foreach my $analysis (@$list_of_analyses) {
    $self->synchronize_AnalysisStats($analysis->stats);
    print STDERR ( ($analysis->stats()->status eq 'BLOCKED') ? 'x' : 'o');
  }
  print STDERR "\n";

  print STDERR ''.((time() - $start_time))." seconds to synchronize_hive\n\n";
}


=head2 safe_synchronize_AnalysisStats

  Arg [1]    : Bio::EnsEMBL::Hive::AnalysisStats object
  Example    : $self->safe_synchronize_AnalysisStats($stats);
  Description: Prewrapper around synchronize_AnalysisStats that does
               checks and grabs sync_lock before proceeding with sync.
               Used by distributed worker sync system to avoid contention.
  Exceptions : none
  Caller     : general

=cut

sub safe_synchronize_AnalysisStats {
  my $self = shift;
  my $stats = shift;

  return $stats unless($stats->analysis_id);
  return $stats if($stats->status eq 'SYNCHING');
  return $stats if($stats->status eq 'DONE');
  return $stats if($stats->sync_lock);
  return $stats if(($stats->status eq 'WORKING') and
                   ($stats->seconds_since_last_update < 3*60));

  # OK try to claim the sync_lock
  my $sql = "UPDATE analysis_stats SET status='SYNCHING', sync_lock=1 ".
            "WHERE sync_lock=0 and analysis_id=" . $stats->analysis_id;
  #print("$sql\n");
  my $row_count = $self->dbc->do($sql);  
  return $stats unless($row_count == 1);        # return the un-updated status if locked
  #printf("got sync_lock on analysis_stats(%d)\n", $stats->analysis_id);
  
      # since we managed to obtain the lock, let's go and perform the sync:
  $self->synchronize_AnalysisStats($stats);

  return $stats;
}


=head2 synchronize_AnalysisStats

  Arg [1]    : Bio::EnsEMBL::Hive::AnalysisStats object
  Example    : $self->synchronize($analysisStats);
  Description: Queries the job and worker tables to get summary counts
               and rebuilds the AnalysisStats object.  Then updates the
               analysis_stats table with the new summary info
  Returntype : newly synced Bio::EnsEMBL::Hive::AnalysisStats object
  Exceptions : none
  Caller     : general

=cut

sub synchronize_AnalysisStats {
  my $self = shift;
  my $analysisStats = shift;

  return $analysisStats unless($analysisStats);
  return $analysisStats unless($analysisStats->analysis_id);

  $analysisStats->refresh(); ## Need to get the new hive_capacity for dynamic analyses
  my $hive_capacity = $analysisStats->hive_capacity;

  if($self->db->hive_use_triggers()) {

            my $job_count = $analysisStats->ready_job_count();
            my $required_workers = $hive_capacity && POSIX::ceil( $job_count / $analysisStats->get_or_estimate_batch_size() );

                # adjust_stats_for_living_workers:
            if($hive_capacity > 0) {
                my $unfulfilled_capacity = $hive_capacity - $analysisStats->num_running_workers();

                if($unfulfilled_capacity < $required_workers ) {
                    $required_workers = (0 < $unfulfilled_capacity) ? $unfulfilled_capacity : 0;
                }
            }
            $analysisStats->num_required_workers( $required_workers );

  } else {
      $analysisStats->total_job_count(0);
      $analysisStats->semaphored_job_count(0);
      $analysisStats->ready_job_count(0);
      $analysisStats->done_job_count(0);
      $analysisStats->failed_job_count(0);
      $analysisStats->num_required_workers(0);

            # ask for analysis_id to force MySQL to use existing index on (analysis_id, status)
      my $sql = "SELECT analysis_id, status, count(*) FROM job WHERE analysis_id=? GROUP BY analysis_id, status";
      my $sth = $self->prepare($sql);
      $sth->execute($analysisStats->analysis_id);

      my $done_here       = 0;
      my $done_elsewhere  = 0;
      my $total_job_count = 0;
      while (my ($dummy_analysis_id, $status, $job_count)=$sth->fetchrow_array()) {
    # print STDERR "$status: $job_count\n";

        $total_job_count += $job_count;

        if($status eq 'READY') {
            $analysisStats->ready_job_count($job_count);

            my $required_workers = $hive_capacity && POSIX::ceil( $job_count / $analysisStats->get_or_estimate_batch_size() );

                # adjust_stats_for_living_workers:
            if($hive_capacity > 0) {
                my $unfulfilled_capacity = $hive_capacity - $self->count_running_workers( $analysisStats->analysis_id() );

                if($unfulfilled_capacity < $required_workers ) {
                    $required_workers = (0 < $unfulfilled_capacity) ? $unfulfilled_capacity : 0;
                }
            }
            $analysisStats->num_required_workers( $required_workers );

        } elsif($status eq 'SEMAPHORED') {
            $analysisStats->semaphored_job_count($job_count);
        } elsif($status eq 'DONE') {
            $done_here = $job_count;
        } elsif($status eq 'PASSED_ON') {
            $done_elsewhere = $job_count;
        } elsif ($status eq 'FAILED') {
            $analysisStats->failed_job_count($job_count);
        }
      } # /while
      $sth->finish;

      $analysisStats->total_job_count( $total_job_count );
      $analysisStats->done_job_count( $done_here + $done_elsewhere );
  } # /unless $self->{'_hive_use_triggers'}

  $analysisStats->check_blocking_control_rules();

  if($analysisStats->status ne 'BLOCKED') {
    $analysisStats->determine_status();
  }

  # $analysisStats->sync_lock(0); ## do we perhaps need it here?
  $analysisStats->update;  #update and release sync_lock

  return $analysisStats;
}


=head2 get_num_failed_analyses

  Arg [1]    : Bio::EnsEMBL::Hive::AnalysisStats object (optional)
  Example    : if( $self->get_num_failed_analyses( $my_analysis )) { do_something; }
  Example    : my $num_failed_analyses = $self->get_num_failed_analyses();
  Description: Reports all failed analyses and returns
                either the number of total failed (if no $filter_analysis was provided)
                or 1/0, depending on whether $filter_analysis failed or not.
  Returntype : int
  Exceptions : none
  Caller     : general

=cut

sub get_num_failed_analyses {
    my ($self, $filter_analysis) = @_;

    my $failed_analyses = $self->db->get_AnalysisAdaptor->fetch_all_failed_analyses();

    my $filter_analysis_failed = 0;

    foreach my $failed_analysis (@$failed_analyses) {
        print "\t##########################################################\n";
        print "\t# Too many jobs in analysis '".$failed_analysis->logic_name."' FAILED #\n";
        print "\t##########################################################\n\n";
        if($filter_analysis and ($filter_analysis->dbID == $failed_analysis)) {
            $filter_analysis_failed = 1;
        }
    }

    return $filter_analysis ? $filter_analysis_failed : scalar(@$failed_analyses);
}


sub get_hive_current_load {
    my $self = shift;
    my $sql = qq{
        SELECT sum(1/hive_capacity)
        FROM worker w
        JOIN analysis_stats USING(analysis_id)
        WHERE w.status!='DEAD'
        AND hive_capacity>0
    };
    my $sth = $self->prepare($sql);
    $sth->execute();
    my ($load)=$sth->fetchrow_array();
    $sth->finish;
    return ($load || 0);
}


sub count_running_workers {
    my ($self, $analysis_id) = @_;

    my $sql = qq{
            SELECT count(*)
            FROM worker
            WHERE status!='DEAD'
        } . ($analysis_id ? " AND analysis_id='$analysis_id'" : '');

    my $sth = $self->prepare($sql);
    $sth->execute();
    (my $running_workers_count)=$sth->fetchrow_array();
    $sth->finish();

    return $running_workers_count || 0;
}


=head2 schedule_workers

  Arg[1]     : Bio::EnsEMBL::Hive::Analysis object (optional)
  Example    : $count = $queen->schedule_workers();
  Description: Runs through the analyses in the system which are waiting
               for workers to be created for them.  Calculates the maximum
               number of workers needed to fill the current needs of the system
               If Arg[1] is defined, does it only for the given analysis.
  Exceptions : none
  Caller     : beekeepers and other external processes

=cut

sub schedule_workers {
  my ($self, $filter_analysis, $orig_pending_by_rc_name, $available_submit_limit) = @_;

  my @suitable_analyses         = $filter_analysis
                                    ? ( $filter_analysis->stats )
                                    : @{ $self->db->get_AnalysisStatsAdaptor->fetch_all_by_suitability_rc_id() };

  my $analysis_id2rc_id         = $self->db->get_AnalysisAdaptor->fetch_HASHED_FROM_analysis_id_TO_resource_class_id();
  my $rc_id2name                = $self->db->get_ResourceClassAdaptor->fetch_HASHED_FROM_resource_class_id_TO_name();
        # combined mapping
  my %analysis_id2rc_name       = map { $_ => $rc_id2name->{ $analysis_id2rc_id->{ $_ }} } keys %$analysis_id2rc_id;

  return {} unless(@suitable_analyses);

  my %pending_by_rc_name        = %{ $orig_pending_by_rc_name || {} };  # NB: make our own copy to be able to modify it
  my $total_workers_to_run      = 0;
  my %workers_to_run_by_rc_name = ();
  my $available_load            = 1.0 - $self->get_hive_current_load();


  foreach my $analysis_stats (@suitable_analyses) {
    last if ($available_load <= 0.0);
    last if (defined($available_submit_limit) and !$available_submit_limit);

        #digging deeper under the surface so need to sync
    if(($analysis_stats->status eq 'LOADING') or ($analysis_stats->status eq 'BLOCKED') or ($analysis_stats->status eq 'ALL_CLAIMED')) {
      $self->synchronize_AnalysisStats($analysis_stats);
    }

    next if($analysis_stats->status eq 'BLOCKED');
    next if($analysis_stats->num_required_workers == 0);

        # FIXME: the following call *sometimes* returns a stale number greater than the number of workers actually needed for an analysis; -sync fixes it
    my $workers_this_analysis = $analysis_stats->num_required_workers;

    if(defined($available_submit_limit)) {                              # submit_limit total capping, if available
        if($workers_this_analysis > $available_submit_limit) {
            $workers_this_analysis = $available_submit_limit;
        }
        $available_submit_limit -= $workers_this_analysis;
    }

    if((my $hive_capacity = $analysis_stats->hive_capacity) > 0) {      # per-analysis hive_capacity capping, if available
        my $remaining_capacity_for_this_analysis = int($available_load * $hive_capacity);

        if($workers_this_analysis > $remaining_capacity_for_this_analysis) {
            $workers_this_analysis = $remaining_capacity_for_this_analysis;
        }

        $available_load -= 1.0*$workers_this_analysis/$hive_capacity;
    }

    my $curr_rc_name    = $analysis_id2rc_name{ $analysis_stats->analysis_id };

    if($pending_by_rc_name{ $curr_rc_name }) {                              # per-rc_name capping by pending processes, if available
        my $pending_this_analysis = ($pending_by_rc_name{ $curr_rc_name } < $workers_this_analysis) ? $pending_by_rc_name{ $curr_rc_name } : $workers_this_analysis;

        print "Scheduler detected $pending_this_analysis pending workers with resource_class_name=$curr_rc_name\n";
        $workers_this_analysis                  -= $pending_this_analysis;
        $pending_by_rc_name{ $curr_rc_name }    -= $pending_this_analysis;
    }

    next unless($workers_this_analysis);    # do not autovivify the hash by a zero

    $total_workers_to_run += $workers_this_analysis;
    $workers_to_run_by_rc_name{ $curr_rc_name } += $workers_this_analysis;
    $analysis_stats->print_stats();
    printf("Scheduler suggests adding %d more workers of resource_class_name=%s for analysis_id=%d [%1.3f hive_load remaining]\n", $workers_this_analysis, $curr_rc_name, $analysis_stats->analysis_id, $available_load);
  }

  printf("Scheduler suggests adding a total of %d workers [%1.5f hive_load remaining]\n", $total_workers_to_run, $available_load);
  return \%workers_to_run_by_rc_name;
}


sub schedule_workers_resync_if_necessary {
    my ($self, $valley, $analysis) = @_;

    my $meadow              = $valley->get_current_meadow();

    my $pending_by_rc_name  = ($meadow->can('count_pending_workers_by_rc_name') and $meadow->config_get('PendingAdjust')) ? $meadow->count_pending_workers_by_rc_name() : {};
    my $submit_limit        = $meadow->config_get('SubmitWorkersMax');
    my $meadow_limit        = ($meadow->can('count_running_workers') and defined($meadow->config_get('TotalRunningWorkersMax'))) ? $meadow->config_get('TotalRunningWorkersMax') - $meadow->count_running_workers : undef;

    my $available_submit_limit = ($submit_limit and $meadow_limit)
                                    ? (($submit_limit<$meadow_limit) ? $submit_limit : $meadow_limit)
                                    : (defined($submit_limit) ? $submit_limit : $meadow_limit);

    my $workers_to_run_by_rc_name = $self->schedule_workers($analysis, $pending_by_rc_name, $available_submit_limit);

    unless( keys %$workers_to_run_by_rc_name or $self->get_hive_current_load() or $self->count_running_workers() ) {
        print "\nScheduler: nothing is running and nothing to do (according to analysis_stats) => executing garbage collection and sync\n" ;

        $self->check_for_dead_workers($valley, 1);
        $self->synchronize_hive($analysis);

        $workers_to_run_by_rc_name = $self->schedule_workers($analysis, $pending_by_rc_name, $available_submit_limit);
    }