Both the worker pool and the child process should handle the timeout Worker pool: throw an exception to the caller Child process: attempt to cancel the work in progress Option to recycle worker on request timeout?