In a multi-machine environment, update, or delete data is taken out of the number of updates, and then the number of updates is the same when performing the operation, and the difference is returned, not processing.
