The principal goal of this project is to investigate methods for implementing both load sharing and fault tolerance in distributed computing systems. Load sharing methods involve transferring jobs from heavily loaded nodes in the network to more lightly loaded nodes to give better service to the users of the system. Several schemes for load sharing have been proposed and analyzed. Fault tolerance methods involve placing copies of jobs at multiple sites in the network, so that if a site fails, one of the copies of its jobs can be executed. This project will investigate the performance of load sharing schemes in networks in which failures may occur, and the performance of combined load sharing and fault tolerance methods.