documentcloud-cloud-crowd 0.2.1 → 0.2.2

data/cloud-crowd.gemspec

@@ -1,7 +1,7 @@
 Gem::Specification.new do |s|
   s.name      = 'cloud-crowd'
-  s.version   = '0.2.1'         # Keep version in sync with cloud-cloud.rb
-  s.date      = '2009-09-18'
+  s.version   = '0.2.2'         # Keep version in sync with cloud-cloud.rb
+  s.date      = '2009-09-23'
 
   s.homepage    = "http://wiki.github.com/documentcloud/cloud-crowd"
   s.summary     = "Parallel Processing for the Rest of Us"

data/config/config.example.yml

@@ -29,10 +29,14 @@
 :s3_bucket:           [your CloudCrowd bucket]
 :s3_authentication:   no
 
-# If you're using the 'filesystem' storage, perhaps with an NFS share or
-# something similar, all files will be saved inside of the 'local_storage_path'.
-# The default value if left unspecified is '/tmp/cloud_crowd_storage'.
-:local_storage_path:  /tmp/cloud_crowd_storage
+# The following settings configure local paths. 'local_storage_path' is the 
+# directory in which all files will be saved if you're using the 'filesystem'
+# storage. 'log_path' and 'pid_path' are the directories in which daemonized 
+# servers and nodes will store their process ids and log files. The default 
+# values are listed.
+# :local_storage_path:  /tmp/cloud_crowd_storage
+# :log_path:            log
+# :pid_path:            tmp/pids
 
 # Use HTTP Basic Auth for all requests? (Includes all internal worker requests 
 # to the central server). If yes, specify the login and password that all 

data/lib/cloud-crowd.rb

@@ -43,13 +43,19 @@ module CloudCrowd
   autoload :WorkUnit,     'cloud_crowd/models'
   
   # Keep this version in sync with the gemspec.
-  VERSION        = '0.2.1'
+  VERSION        = '0.2.2'
   
   # Increment the schema version when there's a backwards incompatible change.
   SCHEMA_VERSION = 3
   
   # Root directory of the CloudCrowd gem.
   ROOT           = File.expand_path(File.dirname(__FILE__) + '/..')
+  
+  # Default folder to log daemonized servers and nodes into.
+  LOG_PATH       = 'log'
+  
+  # Default folder to contain the pids of daemonized servers and nodes.
+  PID_PATH       = 'tmp/pids'
     
   # A Job is processing if its WorkUnits are in the queue to be handled by nodes.
   PROCESSING     = 1
@@ -107,6 +113,18 @@ module CloudCrowd
       @central_server ||= RestClient::Resource.new(CloudCrowd.config[:central_server], CloudCrowd.client_options)
     end
     
+    # The path that daemonized servers and nodes will log to.
+    def log_path(log_file=nil)
+      @log_path ||= config[:log_path] || LOG_PATH
+      log_file ? File.join(@log_path, log_file) : @log_path
+    end
+    
+    # The path in which daemonized servers and nodes will store their pids.
+    def pid_path(pid_file=nil)
+      @pid_path ||= config[:pid_path] || PID_PATH
+      pid_file ? File.join(@pid_path, pid_file) : @pid_path
+    end
+    
     # The standard RestClient options for the central server talking to nodes,
     # as well as the other way around. There's a timeout of 5 seconds to open
     # a connection, and a timeout of 30 to finish reading it.

data/lib/cloud_crowd/action.rb

@@ -103,10 +103,13 @@ module CloudCrowd
       @input = JSON.parse(@input)
     end
     
+    def input_is_url?
+      !URI.parse(@input).scheme.nil? rescue false
+    end
+    
     # If the input is a URL, download the file before beginning processing.
     def download_input
-      input_is_url = !!URI.parse(@input) rescue false
-      return unless input_is_url
+      return unless input_is_url?
       Dir.chdir(@work_directory) do
         @input_path = File.join(@work_directory, safe_filename(@input))
         @file_name = File.basename(@input_path, File.extname(@input_path))

data/lib/cloud_crowd/command_line.rb

@@ -24,6 +24,9 @@ Commands:
   node          Start up a worker node (only one node per machine, please)
   console       Launch a CloudCrowd console, connected to the central database
   load_schema   Load the schema into the database specified by database.yml
+  
+  server -d [start | stop | restart]    Servers and nodes can be launched as
+  node -d [start | stop | restart]      daemons, then stopped or restarted.
 
 Options:
     EOS
@@ -31,11 +34,12 @@ Options:
     # Creating a CloudCrowd::CommandLine runs from the contents of ARGV.
     def initialize
       parse_options
-      command = ARGV.shift
+      command     = ARGV.shift
+      subcommand  = ARGV.shift
       case command
       when 'console'      then run_console
-      when 'server'       then run_server
-      when 'node'         then run_node
+      when 'server'       then run_server(subcommand)
+      when 'node'         then run_node(subcommand)
       when 'load_schema'  then run_load_schema
       when 'install'      then run_install
       else                     usage
@@ -53,29 +57,77 @@ Options:
       IRB.start
     end
     
-    # Convenience command for quickly spinning up the central server. More 
-    # sophisticated deployments, load-balancing across multiple app servers, 
-    # should use the config.ru rackup file directly. This method will start
-    # a single Thin server, if Thin is installed, otherwise the rackup defaults 
-    # (Mongrel, falling back to WEBrick). The equivalent of Rails' script/server.
-    def run_server
+    # `crowd server` can either 'start', 'stop', or 'restart'.
+    def run_server(subcommand)
       ensure_config
-      @options[:port] ||= 9173
-      require 'rubygems'
+      load_code
+      subcommand ||= 'start'
+      case subcommand
+      when 'start'    then start_server
+      when 'stop'     then stop_server
+      when 'restart'  then restart_server
+      end
+    end
+    
+    # Convenience command for quickly spinning up the central server. More
+    # sophisticated deployments, load-balancing across multiple app servers,
+    # should use the config.ru rackup file directly. This method will start
+    # a single Thin server.
+    def start_server
+      port        = @options[:port] || 9173
+      daemonize   = @options[:daemonize] ? '-d' : ''
+      log_path    = CloudCrowd.log_path('server.log')
+      pid_path    = CloudCrowd.pid_path('server.pid')
       rackup_path = File.expand_path("#{@options[:config_path]}/config.ru")
-      if Gem.available? 'thin'
-        exec "thin -e #{@options[:environment]} -p #{@options[:port]} -R #{rackup_path} start"
-      else
-        exec "rackup -E #{@options[:environment]} -p #{@options[:port]} #{rackup_path}"
+      FileUtils.mkdir_p(CloudCrowd.log_path) if @options[:daemonize] && !File.exists?(CloudCrowd.log_path)
+      puts "Starting CloudCrowd Central Server on port #{port}..."
+      exec "thin -e #{@options[:environment]} -p #{port} #{daemonize} --tag cloud-crowd-server --log #{log_path} --pid #{pid_path} -R #{rackup_path} start"
+    end
+    
+    # Stop the daemonized central server, if it exists.
+    def stop_server
+      Thin::Server.kill(CloudCrowd.pid_path('server.pid'), 0)
+    end
+    
+    # Restart the daemonized central server.
+    def restart_server
+      stop_server
+      sleep 1
+      start_server
+    end
+    
+    # `crowd node` can either 'start', 'stop', or 'restart'.
+    def run_node(subcommand)
+      ensure_config
+      load_code
+      subcommand ||= 'start'
+      case subcommand
+      when 'start'    then start_node
+      when 'stop'     then stop_node
+      when 'restart'  then restart_node
       end
     end
     
     # Launch a Node. Please only run a single node per machine. The Node process
     # will be long-lived, although its workers will come and go.
-    def run_node
+    def start_node
       ENV['RACK_ENV'] = @options['environment']
       load_code
-      Node.new(@options[:port])
+      port = @options[:port] || Node::DEFAULT_PORT
+      puts "Starting CloudCrowd Node on port #{port}..."
+      Node.new(port, @options[:daemonize])
+    end
+    
+    # If the daemonized Node is running, stop it.
+    def stop_node
+      Thin::Server.kill CloudCrowd.pid_path('node.pid')
+    end
+    
+    # Restart the daemonized Node, if it exists.
+    def restart_node
+      stop_node
+      sleep 1
+      start_node
     end
     
     # Load in the database schema to the database specified in 'database.yml'.
@@ -117,7 +169,8 @@ Options:
     def parse_options
       @options = {
         :environment  => 'production',
-        :config_path  => ENV['CLOUD_CROWD_CONFIG'] || '.'
+        :config_path  => ENV['CLOUD_CROWD_CONFIG'] || '.',
+        :daemonize    => false
       }
       @option_parser = OptionParser.new do |opts|
         opts.on('-c', '--config PATH', 'path to configuration directory') do |conf_path|
@@ -129,6 +182,9 @@ Options:
         opts.on('-e', '--environment ENV', 'server environment (sinatra)') do |env|
           @options[:environment] = env
         end
+        opts.on('-d', '--daemonize', 'run as a background daemon') do |daemonize|
+          @options[:daemonize] = daemonize
+        end
         opts.on_tail('-v', '--version', 'show version') do
           require "#{CC_ROOT}/lib/cloud-crowd"
           puts "CloudCrowd version #{VERSION}"

data/lib/cloud_crowd/models/job.rb

@@ -114,7 +114,9 @@ module CloudCrowd
     def percent_complete
       return 99  if merging?
       return 100 if complete?
-      (work_units.complete.count / work_units.count.to_f * 100).round
+      unit_count = work_units.count
+      return 100 if unit_count <= 0
+      (work_units.complete.count / unit_count.to_f * 100).round
     end
     
     # How long has this Job taken?

data/lib/cloud_crowd/models/node_record.rb

@@ -48,7 +48,7 @@ module CloudCrowd
     
     # What Actions is this Node able to run?
     def actions
-      enabled_actions.split(',')
+      @actions ||= enabled_actions.split(',')
     end
     
     # Is this Node too busy for more work? Determined by number of workers, or 

data/lib/cloud_crowd/models/work_unit.rb

@@ -17,24 +17,27 @@ module CloudCrowd
     # Reserved WorkUnits have been marked for distribution by a central server process.
     named_scope :reserved,  {:conditions => {:reservation => $$}, :order => 'updated_at asc'}
     
-    # Attempt to send a list of work_units to nodes with available capacity.
+    # Attempt to send a list of WorkUnits to nodes with available capacity.
     # A single central server process stops the same WorkUnit from being
     # distributed to multiple nodes by reserving it first. The algorithm used
     # should be lock-free.
+    #
+    # We loop over the WorkUnits reserved by this process and try to match them
+    # to Nodes that are capable of handling the Action. WorkUnits get removed 
+    # from the availability list when they are successfully sent, and Nodes get
+    # removed when they are busy or have the action in question disabled.
     def self.distribute_to_nodes
       return unless WorkUnit.reserve_available
       work_units = WorkUnit.reserved
       available_nodes = NodeRecord.available
-      until work_units.empty? do
-        node = available_nodes.shift
-        unit = work_units.first
-        break unless node && unit
-        next unless node.actions.include? unit.action
-        sent = node.send_work_unit(unit)
-        if sent
-          work_units.shift
-          available_nodes.push(node) unless node.busy?
+      while node = available_nodes.shift and unit = work_units.shift do
+        if node.actions.include? unit.action
+          if node.send_work_unit(unit)
+            available_nodes.push(node) unless node.busy?
+            next
+          end
         end
+        work_units.push(unit)
       end
     ensure
       WorkUnit.cancel_reservations
@@ -106,7 +109,7 @@ module CloudCrowd
         :output         => output,
         :time           => time_taken
       })
-      self.job.check_for_completion
+      job && job.check_for_completion
     end
     
     # Ever tried. Ever failed. No matter. Try again. Fail again. Fail better.

data/lib/cloud_crowd/node.rb

@@ -27,7 +27,7 @@ module CloudCrowd
     # The response sent back when this node is overloaded.
     OVERLOADED_MESSAGE  = 'Node Overloaded'
     
-    attr_reader :asset_store, :enabled_actions, :host, :port, :server
+    attr_reader :enabled_actions, :host, :port, :central
             
     set :root, ROOT
     set :authorization_realm, "CloudCrowd"
@@ -53,19 +53,20 @@ module CloudCrowd
     # Returns a 503 if this Node is overloaded.
     post '/work' do
       throw :halt, [503, OVERLOADED_MESSAGE] if @overloaded
-      pid = fork { Worker.new(self, JSON.parse(params[:work_unit])).run }
+      unit = JSON.parse(params[:work_unit])
+      pid = fork { Worker.new(self, unit).run }
       Process.detach(pid)
       json :pid => pid
     end
     
     # When creating a node, specify the port it should run on.
-    def initialize(port=DEFAULT_PORT)
+    def initialize(port=nil, daemon=false)
       require 'json'
-      @server           = CloudCrowd.central_server
+      @central          = CloudCrowd.central_server
       @host             = Socket.gethostname
       @enabled_actions  = CloudCrowd.actions.keys
-      @asset_store      = AssetStore.new
       @port             = port || DEFAULT_PORT
+      @daemon           = daemon
       @overloaded       = false
       @max_load         = CloudCrowd.config[:max_load]
       @min_memory       = CloudCrowd.config[:min_free_memory]
@@ -75,10 +76,17 @@ module CloudCrowd
     # Starting up a Node registers with the central server and begins to listen
     # for incoming WorkUnits.
     def start
+      FileUtils.mkdir_p(CloudCrowd.log_path) if @daemon && !File.exists?(CloudCrowd.log_path)
+      @server          = Thin::Server.new('0.0.0.0', @port, self, :signals => false)
+      @server.tag      = 'cloud-crowd-node'
+      @server.pid_file = CloudCrowd.pid_path('node.pid')
+      @server.log_file = CloudCrowd.log_path('node.log')
+      @server.daemonize if @daemon
       trap_signals
-      start_server
-      monitor_system if @max_load || @min_memory
+      asset_store
+      @server_thread   = Thread.new { @server.start }
       check_in(true)
+      monitor_system if @max_load || @min_memory
       @server_thread.join
     end
     
@@ -86,21 +94,26 @@ module CloudCrowd
     # configuration of this Node. If it can't check-in, there's no point in 
     # starting.
     def check_in(critical=false)
-      @server["/node/#{@host}"].put(
+      @central["/node/#{@host}"].put(
         :port             => @port,
         :busy             => @overloaded,
         :max_workers      => CloudCrowd.config[:max_workers],
         :enabled_actions  => @enabled_actions.join(',')
       )
     rescue Errno::ECONNREFUSED
-      puts "Failed to connect to the central server (#{@server.to_s})."
+      puts "Failed to connect to the central server (#{@central.to_s})."
       raise SystemExit if critical
     end
     
     # Before exiting, the Node checks out with the central server, releasing all
     # of its WorkUnits for other Nodes to handle
     def check_out
-      @server["/node/#{@host}"].delete
+      @central["/node/#{@host}"].delete
+    end
+    
+    # Lazy-initialize the asset_store, preferably after the Node has launched.
+    def asset_store
+      @asset_store ||= AssetStore.new
     end
     
     # Is the node overloaded? If configured, checks if the load average is 
@@ -133,13 +146,6 @@ module CloudCrowd
     
     private
     
-    # Launch the Node's Thin server in a separate thread because it blocks.
-    def start_server
-      @server_thread = Thread.new do
-        Thin::Server.start('0.0.0.0', @port, self, :signals => false)
-      end
-    end
-    
     # Launch a monitoring thread that periodically checks the node's load 
     # average and the amount of free memory remaining. If we transition out of 
     # the overloaded state, let central know.
@@ -156,6 +162,7 @@ module CloudCrowd
     
     # Trap exit signals in order to shut down cleanly.
     def trap_signals
+      Signal.trap('QUIT') { shut_down }
       Signal.trap('INT')  { shut_down }
       Signal.trap('KILL') { shut_down }
       Signal.trap('TERM') { shut_down }
@@ -165,7 +172,7 @@ module CloudCrowd
     def shut_down
       @monitor_thread.kill if @monitor_thread
       check_out
-      @server_thread.kill
+      @server_thread.kill if @server_thread
       Process.exit
     end
     

data/lib/cloud_crowd/worker.rb

@@ -30,7 +30,7 @@ module CloudCrowd
     def complete_work_unit(result)
       keep_trying_to "complete work unit" do
         data = base_params.merge({:status => 'succeeded', :output => result})
-        @node.server["/work/#{data[:id]}"].put(data)
+        @node.central["/work/#{data[:id]}"].put(data)
         log "finished #{display_work_unit} in #{data[:time]} seconds"
       end
     end
@@ -39,7 +39,7 @@ module CloudCrowd
     def fail_work_unit(exception)
       keep_trying_to "mark work unit as failed" do
         data = base_params.merge({:status => 'failed', :output => {'output' => exception.message}.to_json})
-        @node.server["/work/#{data[:id]}"].put(data)
+        @node.central["/work/#{data[:id]}"].put(data)
         log "failed #{display_work_unit} in #{data[:time]} seconds\n#{exception.message}\n#{exception.backtrace}"
       end
     end
@@ -70,35 +70,37 @@ module CloudCrowd
     # failures. We capture the thread so that we can kill it from the outside,
     # when exiting.
     def run_work_unit
-      @worker_thread = Thread.new do
-        begin
-          result = nil
-          action_class = CloudCrowd.actions[@unit['action']]
-          action = action_class.new(@status, @unit['input'], enhanced_unit_options, @node.asset_store)
-          Dir.chdir(action.work_directory) do
-            result = case @status
-            when PROCESSING then action.process
-            when SPLITTING  then action.split
-            when MERGING    then action.merge
-            else raise Error::StatusUnspecified, "work units must specify their status"
-            end
+      begin
+        result = nil
+        action_class = CloudCrowd.actions[@unit['action']]
+        action = action_class.new(@status, @unit['input'], enhanced_unit_options, @node.asset_store)
+        Dir.chdir(action.work_directory) do
+          result = case @status
+          when PROCESSING then action.process
+          when SPLITTING  then action.split
+          when MERGING    then action.merge
+          else raise Error::StatusUnspecified, "work units must specify their status"
           end
-          complete_work_unit({'output' => result}.to_json)
-        rescue Exception => e
-          fail_work_unit(e)
-        ensure
-          action.cleanup_work_directory if action
         end
+        complete_work_unit({'output' => result}.to_json)
+      rescue Exception => e
+        fail_work_unit(e)
+      ensure
+        action.cleanup_work_directory if action
       end
-      @worker_thread.join
     end
     
+    # Run this worker inside of a fork. Attempts to exit cleanly.
     # Wraps run_work_unit to benchmark the execution time, if requested.
     def run
       trap_signals
       log "starting #{display_work_unit}"
-      return run_work_unit unless @unit['options']['benchmark']
-      log("ran #{display_work_unit} in " + Benchmark.measure { run_work_unit }.to_s)
+      if @unit['options']['benchmark']
+        log("ran #{display_work_unit} in " + Benchmark.measure { run_work_unit }.to_s)
+      else
+        run_work_unit
+      end
+      Process.exit!
     end
     
     # There are some potentially important attributes of the WorkUnit that we'd 
@@ -133,22 +135,13 @@ module CloudCrowd
       puts "Worker ##{@pid}: #{message}" unless ENV['RACK_ENV'] == 'test'
     end
     
-    # When signaled to exit, make sure that the Worker shuts down cleanly.
+    # When signaled to exit, make sure that the Worker shuts down without firing
+    # the Node's at_exit callbacks.
     def trap_signals
-      Signal.trap('INT') { shut_down }
-      Signal.trap('KILL') { shut_down }
-      Signal.trap('TERM') { shut_down }
-    end
-    
-    # Force the Worker to quit, even if it's in the middle of processing.
-    # If it had a checked-out WorkUnit, the Node should have released it on
-    # the central server already.
-    def shut_down
-      if @worker_thread
-        @worker_thread.kill
-        @worker_thread.kill! if @worker_thread.alive?
-      end
-      Process.exit
+      Signal.trap('QUIT') { Process.exit! }
+      Signal.trap('INT')  { Process.exit! }
+      Signal.trap('KILL') { Process.exit! }
+      Signal.trap('TERM') { Process.exit! }
     end
     
   end

data/public/css/admin_console.css

@@ -14,13 +14,13 @@ body {
   height: 110px;
   position: absolute;
   top: 0; left: 0; right: 0;
-  background: url(/images/header_back.png);
+  background: url(../images/header_back.png);
 }
   #logo {
     position: absolute;
     left: 37px; top: 9px;
     width: 236px; height: 91px;
-    background: url(/images/logo.png);
+    background: url(../images/logo.png);
   }
   
 #disconnected {
@@ -37,7 +37,7 @@ body {
   #disconnected .server_error {
     float: left;
     width: 16px; height: 16px;
-    background: url(/images/server_error.png);
+    background: url(../images/server_error.png);
     opacity: 0.7;
     margin-right: 3px;
   }
@@ -64,7 +64,7 @@ body {
     height: 75px;
     border: 1px solid #5c5c5c;
     -moz-border-radius: 10px; -webkit-border-radius: 10px; border-radius: 10px;
-    background: transparent url(/images/queue_fill.png) repeat-x 0px -1px;
+    background: transparent url(../images/queue_fill.png) repeat-x 0px -1px;
   }
     #queue.no_jobs #queue_fill {
       opacity: 0.3;
@@ -126,11 +126,11 @@ body {
   }
     #sidebar_top {
       top: 0px;
-      background: url(/images/sidebar_top.png);
+      background: url(../images/sidebar_top.png);
     }
     #sidebar_bottom {
       bottom: 0px;
-      background: url(/images/sidebar_bottom.png);
+      background: url(../images/sidebar_bottom.png);
     }
   #sidebar_header {
     position: absolute;
@@ -164,10 +164,10 @@ body {
       #nodes .node {
         font-size: 11px;
         line-height: 22px;
-        background-image: url(/images/server.png);
+        background-image: url(../images/server.png);
       }
         #nodes .node.busy {
-          background-image: url(/images/server_busy.png);
+          background-image: url(../images/server_busy.png);
         }
           #nodes .node.busy span.busy {
             font-size: 9px;
@@ -178,7 +178,7 @@ body {
         font-size: 10px;
         line-height: 18px;
         cursor: pointer;
-        background-image: url(/images/bullet_green.png);
+        background-image: url(../images/bullet_green.png);
       }
         #nodes .worker:hover {
           border: 1px solid #aaa;
@@ -190,8 +190,7 @@ body {
   position: absolute;
   width: 231px; height: 79px;
   margin: -9px 0 0 -20px;
-  background: url(/images/worker_info.png);
-  overflow: hidden;
+  background: url(../images/worker_info.png);
   cursor: pointer;
 }
   #worker_info_inner {
@@ -199,9 +198,10 @@ body {
     line-height: 15px;
     color: #333;
     text-shadow: 0px 1px 1px #eee;
+    overflow: hidden;
   }
     #worker_info.loading #worker_info_inner {
-      background: url(/images/worker_info_loading.gif) no-repeat right bottom;
+      background: url(../images/worker_info_loading.gif) no-repeat right bottom;
       width: 45px; height: 9px;
     }
     #worker_info.awake #worker_details, 

data/public/js/admin_console.js

@@ -18,7 +18,7 @@ window.Console = {
   DISPLAY_STATUS_MAP : ['unknown', 'processing', 'succeeded', 'failed', 'splitting', 'merging'],    
   
   // Images to preload
-  PRELOAD_IMAGES : ['/images/server_error.png'],
+  PRELOAD_IMAGES : ['images/server_error.png'],
   
   // All options for drawing the system graphs.
   GRAPH_OPTIONS : {
@@ -53,7 +53,7 @@ window.Console = {
   // Request the lastest status of all jobs and workers, re-render or update
   // the DOM to reflect.
   getStatus : function() {
-    $.ajax({url : '/status', dataType : 'json', success : function(resp) {
+    $.ajax({url : 'status', dataType : 'json', success : function(resp) {
       Console._jobs           = resp.jobs;
       Console._nodes          = resp.nodes;
       Console._workUnitCount  = resp.work_unit_count;
@@ -167,7 +167,7 @@ window.Console = {
     var info = Console._workerInfo;
     var row = $(this);
     info.addClass('loading');
-    $.get('/worker/' + row.attr('rel'), null, Console.renderWorkerInfo, 'json');
+    $.get('worker/' + row.attr('rel'), null, Console.renderWorkerInfo, 'json');
     info.css({top : row.offset().top, left : 325});
     info.fadeIn(Console.ANIMATION_SPEED);
     $(document).bind('click', Console.hideWorkerInfo);

data/test/unit/test_action.rb

@@ -9,7 +9,7 @@ end
 
 class ActionTest < Test::Unit::TestCase
   
-  context "A CloudCrowd Job" do
+  context "A CloudCrowd::Action" do
         
     setup do
       @store = CloudCrowd::AssetStore.new
@@ -41,9 +41,30 @@ class ActionTest < Test::Unit::TestCase
     end
     
     should "be able to count the number of words in this file" do
-      assert @action.process == 149
+      assert @action.process == 212
+    end
+    
+    should "raise an exception when backticks fail" do
+      def @action.process; `utter failure 2>&1`; end
+      assert_raise(CloudCrowd::Error::CommandFailed) { @action.process }
     end
             
   end
+
+
+  context "A CloudCrowd::Action without URL input" do
+        
+    setup do
+      @store = CloudCrowd::AssetStore.new
+      @args = [CloudCrowd::PROCESSING, 'inputstring', {'job_id' => 1, 'work_unit_id' => 1}, @store]
+      @action = CloudCrowd.actions['word_count'].new(*@args)
+    end
+    
+    should "should not interpret the input data as an url" do
+      assert_equal 'inputstring', @action.input
+      assert_nil @action.input_path
+    end
+    
+  end
   
 end

data/test/unit/test_job.rb

@@ -31,6 +31,11 @@ class JobTest < Test::Unit::TestCase
       assert @job.percent_complete == 100
       assert @job.outputs == "[\"hello\"]"
     end
+
+    should "not throw a FloatDomainError, even when WorkUnits have vanished" do
+      @job.work_units.destroy_all
+      assert @job.percent_complete == 100
+    end
     
     should "be able to create a job from a JSON request" do
       job = CloudCrowd::Job.create_from_request(JSON.parse(<<-EOS

data/test/unit/test_node.rb

@@ -9,7 +9,7 @@ class NodeUnitTest < Test::Unit::TestCase
     end
     
     should "instantiate correctly" do
-      assert @node.server.to_s == "http://localhost:9173"
+      assert @node.central.to_s == "http://localhost:9173"
       assert @node.port == 11011
       assert @node.host == Socket.gethostname
       assert @node.enabled_actions.length > 2
@@ -17,8 +17,7 @@ class NodeUnitTest < Test::Unit::TestCase
     end
     
     should "trap signals and launch a server at start" do
-      Signal.expects(:trap).times(3)
-      Thin::Server.expects(:start)
+      Thin::Server.any_instance.expects(:start)
       @node.expects(:check_in)
       @node.start
     end

data/views/operations_center.erb

@@ -3,12 +3,12 @@
 <head>
   <meta http-equiv="content-type" content="text/html; charset=UTF-8" />
   <title>Operations Center | CloudCrowd</title>
-  <link href="/css/reset.css" media="screen" rel="stylesheet" type="text/css" />
-  <link href="/css/admin_console.css" media="screen" rel="stylesheet" type="text/css" />
-  <script src="/js/jquery.js" type="text/javascript"></script>
-  <!--[if IE]><script src="/js/excanvas.js" type="text/javascript"></script><![endif]-->
-  <script src="/js/flot.js" type="text/javascript"></script>
-  <script src="/js/admin_console.js" type="text/javascript"></script>
+  <link href="css/reset.css" media="screen" rel="stylesheet" type="text/css" />
+  <link href="css/admin_console.css" media="screen" rel="stylesheet" type="text/css" />
+  <script src="js/jquery.js" type="text/javascript"></script>
+  <!--[if IE]><script src="js/excanvas.js" type="text/javascript"></script><![endif]-->
+  <script src="js/flot.js" type="text/javascript"></script>
+  <script src="js/admin_console.js" type="text/javascript"></script>
 </head>
 
 <body>
@@ -68,8 +68,8 @@
       <div id="worker_info_inner" class="small_caps">
         <div id="worker_details">
           <div id="worker_status">status: <span class="status"></span></div>
-          <div id="worker_action">action: <span class="action"></span></div>
-          <div id="worker_job_id">job #<span class="job_id"></span> / work unit #<span class="work_unit_id"></span></div>
+          <div id="worker_action">action:&nbsp;<span class="action"></span></div>
+          <div id="worker_job_id">job #<span class="job_id"></span> / unit #<span class="work_unit_id"></span></div>
         </div>
         <div id="worker_sleeping">
           worker exiting&hellip;

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: documentcloud-cloud-crowd
 version: !ruby/object:Gem::Version 
-  version: 0.2.1
+  version: 0.2.2
 platform: ruby
 authors: 
 - Jeremy Ashkenas
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-09-18 00:00:00 -07:00
+date: 2009-09-23 00:00:00 -07:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency