一种改进的分布式资源共享的互斥控制算法

改进了Andfew S.TanenlJaum等提出的分布式进程互斥算法,主要方法是客户进程给其他所有进程发送请求进入临界区的消息,然后当前使用临界区的进程回答NO消息,并把该进程排入等待队列.当使用临界区的进程退出时,再行唤醒因等待该临界区而被阻塞的进程.每次如此传递,一直到再无进程使用临界区.通过性能比较,该算法能明显减少网络传输的消息数量,具有一定优势.

Improved Mutex Control Aigodthm of Resources Sharing in Distributed Systems

The process mutex algorithm is proposed by Andrew S. Tanenbaum. The algorithm is implemented through a method that a client wishing to get a critical area sends all of other processes a message asking for a critical area, and then the process in the critical area answers a NO message to stop and queue the former process. When it is going to exit, the process now in the critical area must awaken the process in the head of the queue for awaiting the critical area. Like relay race, all the processes use a critical area by turns until no process wants the critical area. Through comparing the algorithm by Andrew S. Tanenbaum with this algorithm in the performance, it can find that it is obvious that this algorithm can decrease the number of messages transmitted in the network. In conclusion, this algorithm has more advantages.