hadoop jobtracker machine list parser

mapred_machines.rb

require 'nokogiri'
require 'net/http'
require 'timeout'
require 'uri'

job_tracker_machine_list_uri = URI.parse 'http://jobtracker:50030/machines.jsp?type=active'

response = Timeout::timeout 30 do
  Net::HTTP.start job_tracker_machine_list_uri.host, job_tracker_machine_list_uri.port do |http|
    http.request Net::HTTP::Get.new job_tracker_machine_list_uri.request_uri
  end
end

if response.kind_of? Net::HTTPSuccess
  document = Nokogiri::HTML response.body
  fields   = document.xpath('//table//thead//th//b').flat_map do |field_entry|
    field_entry.children.map(&:text).reject do |value|
      value == 'Task Trackers'
    end.map do |value|
      value.sub('# ', '').gsub(' ', '_').downcase.to_sym
    end
  end

  task_trackers = document.xpath('//table//tbody//tr').map do |table_row|
    row_cells = table_row.children.map do |row_cell|
      row_cell.children.map do |cell_entry|
        case cell_entry.name
        when 'a'
          cell_entry.attributes['href'].value
        else
          cell_entry.text
        end.chomp
      end
    end

    Hash[fields.zip(row_cells.flatten)]
  end
end

# example output:
# {:name=>"http://dn-9be23803:50060/",
#  :host=>"dn-9be23803",
#  :running_tasks=>"8",
#  :max_map_tasks=>"8",
#  :max_reduce_tasks=>"6",
#  :task_failures=>"41",
#  :directory_failures=>"0",
#  :node_health_status=>"N/A",
#  :seconds_since_node_last_healthy=>"0",
#  :total_tasks_since_start=>"85628",
#  :succeeded_tasks_since_start=>"83337",
#  :total_tasks_last_day=>"0",
#  :succeeded_tasks_last_day=>"0",
#  :total_tasks_last_hour=>"0",
#  :succeeded_tasks_last_hour=>"0",
#  :seconds_since_heartbeat=>"0"}