Question

I am trying to create an oplog watcher in ruby. So far ive come up with a small script below.

require 'rubygems'
require 'mongo'
db = Mongo::Connection.new("localhost", 5151).db("local")
coll = db.collection('oplog.$main')

loop do
cursor = Mongo::Cursor.new(coll, :tailable => true)
    while not cursor.closed?
        if doc = cursor.next_document
            puts doc
        else
            sleep 1
        end
    end
end

The problem with this is, after 5 or 6 seconds when it has spit out a lot of data it times out and i get an error

C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/connection.rb
:807:in `check_response_flags': Query response returned CURSOR_NOT_FOUND. Either an invalid c
ursor was specified, or the cursor may have timed out on the server. (Mongo::OperationFailure
)
        from C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/
connection.rb:800:in `receive_response_header'
        from C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/
connection.rb:768:in `receive'
        from C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/
connection.rb:493:in `receive_message'
        from C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/
connection.rb:491:in `synchronize'
        from C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/
connection.rb:491:in `receive_message'
        from C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/
cursor.rb:494:in `send_get_more'
        from C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/
cursor.rb:456:in `refresh'
        from C:/RailsInstaller/Ruby1.8.7/lib/ruby/gems/1.8/gems/mongo-1.4.0/lib/../lib/mongo/
cursor.rb:124:in `next_document'
        from n.rb:7
        from n.rb:6:in `loop'
        from n.rb:6

What i dont understand is when i m able to see the actual data how can it suddenly say cursor not found. Im pretty new to ruby and any ideas on what direction i must take will be useful for me.

Answer 1

The solution is that i need to have an exception handling mechanism to capture the exception which is thrown when the cursor reads the last document in a relatively small oplog with an higher number of writes per second. Since the cursor reaches the end of the oplog it would throw an exception that there are no more records.

require 'rubygems'
require 'mongo'
db = Mongo::Connection.new("localhost",5151).db("local")
coll = db.collection('oplog.$main')
loop do
cursor = Mongo::Cursor.new(coll, :timeout => false, :tailable => true)
    while not cursor.closed?
    begin
      if doc = cursor.next_document   
          puts "Timestamp"
          puts  doc["ts"]
          puts "Record"
          puts  doc["o"]
          puts "Affected Collection"
          puts doc["ns"]
      end
    rescue
        puts ""
        break
    end
  end
end

This now works as the exception is been handled. Thanks to the mongodb-user google group for pointing this out to me.