Abstract:

Distributed system is a collection of distinct processes that do not share memory or clock. These processes interconnected by a communication network in which each process has its own local memory and other peripherals. The communication between any two processes of the system takes place by message passing over the communication network. These processes in the distributed system run concurrently. The purpose of the distributed system is to provide an efficient and convenient environment for sharing of resources.In this paper, we present an efficient fault-tolerance algorithm for Mutual Exclusion (ME) in distributed systems. This algorithm is considered as an enhancement to Reddy algorithm. The algorithm is based in token-based technique and requires a lot number of messages to enter the critical section (CS) and to detect the system failures. In Reddy algorithm the process must request to enter the CS from all process all the times, we used a technique wherein it can enter the CS by sending less number of messages. Also our algorithm reduces the number of messages for the detection of the token lost. We compare our algorithm with Reddy and we find that our algorithm is more efficient.