Processes can monitor each other. Processes dedicated to monitoring and restarting other processes are called supervisors. You code for the "happy path" and let

Processes can monitor each other. Processes dedicated to monitoring and restarting other processes are called supervisors. You code for the "happy path" and let a tree of supervisors handle unexpected failures. An uncaught exception will torpedo its process. Processes might also monitor each other for other reasons. For example, a resource pool would monitor processes which check out a resource, in case they die before returning the resource.