Futures Trading Platform, Load Balancing And Disaster Recovery And Realization

Nowadays, electrical trading is used in almost all futures exchanges throughout the world. High performance and high reliability are two important characteristics in the trading system.One popular method to achieve high performance is to use distributed system, while load balancing of every server is a major issue in it. Combining today's load balancing tech with the actual situation of time-bargain system platform, we bring forward an adaptive load balancing algorithm based on service in this article.Several state value is choosed as a measure of system load first, which are closely related with system performance. Then, they are assigned with some weight. The system load value is calculated with the average state value of recent. When the server receives client's logining request, the server gets the server list first, then gets and sorts the system load value of the servers, then returns the suitable server or return failure on the basis of the user type(VIP, ordinary users). This algorithm can help to implement load balancing of the whole platform efficiently.Another important tech ensuring the normal operation of bargain system is the disaster recovery. Disaster recovery helps the system respond on time when any software or hardware in it went wrong, so it ensures the normal operation of the system.The combination of disaster recovery and load balancing can really achieve high performance and high stability. They two are complementary to each other and neither is dispensable.Combining today's disaster recovery tech with the actual situation of time-bargain system platform, we designed an integrated disaster recovery project on the system platform in this article. In the software level, processes are monitored by each other, so they can be automatically resotored;In the hardware level, each server has a backup server, and different measures are designed to handle different abnormities. Most of these disaster management measures are done automatically.This project can adapt high demand of reliability and can help to take failure free operation during trading, preventing failure on single point.