Threaderror invoking singleton (different behaviour than MRI)

The following occurs within a rails project:

From within environment.rb, place the following:

MANAGER_PREFS = {
:development => {
:threads => 10,
:root_dir => "/jboss/deploy",
:util_dir => "/usr2/bea/bin"
},
:test => { },
:production => { }
}

if ENV['RAILS_ENV'] != nil
require "lib/manager"
MANAGER=Manager.instance
end

Here's the snippet of Manager which is causing theissue:

1. manager
   require 'singleton'
   require 'socket'
   require 'thread'
   require 'threadpool'
   require 'logger'
   require 'tasker'
2. The manager uses the MANAGER_PREFS variable defined in the
3. config/environment.rb

1. The manager has two parts:
2. + the first periodically checks for servers which
3. are managed and checks to see if they are running
4. + the second looks in the tasks table for tasks to be
5. performed and performs the tasks
   class Manager < Thread
   include Singleton

def initialize()
@logger = RAILS_DEFAULT_LOGGER

1. $DEBUG = true
   pool_size = MANAGER_PREFS[ENV['RAILS_ENV'].to_sym][:threads] || 10
   @pool = ThreadPool.new(pool_size,@logger)
   @continue = true
   @tasker = Tasker.instance
   super do
   cycle = 0
   @logger.debug "Manager is sleeping for 60 seconds to give system time to stabilize"
   sleep 60 # give time to stabilize
   @logger.debug "Manager is now awake"
   while @continue do
   @logger.debug "Manager doing stuff: #{cycle}"
2. set true by default
   monitor_active = false # fixme
   tasker_active = true
   begin
   settings = GlobalSetting.find_first
   @logger.debug "Manager: #{settings.inspect}"
   monitor_active = settings.monitor_active
   tasker_active = settings.tasker_active
   rescue Exception => e
   @logger.info "Manager: Exception occurred reading global settings: #{e.message}"
   @logger.info "Manager: Exception Backtrace:\n #{e.backtrace.inspect}"
   end
   check_managed_servers if ((cycle == 0) && (monitor_active))
   @logger.debug "Manager stage 1"
   perform_tasks if tasker_active
   @logger.debug "Manager sleeping"
   sleep 10
   cycle = (cycle + 1) % 6
   end
   @logger.debug "Manager Finished"
   end
   end
   end

Here's the stack trace:
/usr2/common/jruby/lib/ruby/1.8/singleton.rb:95:in `new': must be called with a block (ThreadError)
from /usr2/common/jruby/lib/ruby/1.8/singleton.rb:95:in `instance'
from /usr2/common/jruby/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in `instance'
from ./script/../config/../config/environment.rb:106
from /usr2/common/jruby/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in `require'
from /usr2/common/jruby/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in `require'
from /usr2/common/jruby/lib/ruby/gems/1.8/gems/rails-1.2.3/lib/commands/servers/webrick.rb:52
from /usr2/common/jruby/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in `require'
from /usr2/common/jruby/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in `require'
from /usr2/common/jruby/lib/ruby/gems/1.8/gems/rails-1.2.3/lib/commands/server.rb:39
from /usr2/common/jruby/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in `require'
from /usr2/common/jruby/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in `require'
from script/server:2