Evaluation of Lump-sum Update Methods for Nonstop Service System

In many mission-critical systems, the lump-sum update of large amounts of data is performed. On the one hand, with the development of internet business, nonstop online services have become to be provided in many missioncritical systems. So, the lump-sum update has to be performed concurrently with the online entry. However, in the actual mission-critical systems, there are various kinds of lumpsum update operations corresponding to their business. In this paper, we define the lump-sum update models from the point of view of both the actual business process and characteristics of the target data, and show the problem of the conventional update methods. Then, we propose a novel update method for this problem, which utilizes the transaction time database, and show its evaluation results of the efficiency of both the lump-sum update and online entry comparing with the conventional update method. Based on these results, we show the proposal method is effective in the case where the update data is related to each other.