Self-stabilizing mechanism for High-Available Stock Trading System

In enterprise computing, distribute computing and client/server computing are useful technologies because most companies have so many servers and various computing resource. Moreover, in military or financing computing parts, real-time computing is also useful technology because they need to catch the enemy quickly and they need to trade stock quickly. Fault tolerance assists stabilized system. Self-repair and self-stabilizing is useful technology for fault-tolerance. However, it is difficult to implement fault-tolarance for the whole systematic enterprise service. In this paper, we will introduce the case of Meritz Securities Company. We were operating the stock trading system to apply a self-stabilizing mechanism in order for 24 by 365. Our system is operated with some special method we called the RCM (reconnection mechanism) and the RRM (rollback recording mechanism), RCM and RRM are useful methods not only system crash, but also network failure.