Autonomic Computing (AC) aims at solving the problem of managing the rapidly-growing complexity of Information Technology systems, by creating self-managing systems. In this thesis, we have surveyed the progress of the AC field, and studied the requirements, models and architectures of AC. The commonly recognized AC requirements are four properties - self-configuring, self-healing, self-optimizing, and self-protecting. The recommended software architecture is the MAPE-K model containing four modules, namely - monitor, analyze, plan and execute, as well as the knowledge repository. In the modern software marketplace, Java Management Extensions (JMX) has facilitated one function of the AC requirements - monitoring. Using JMX, we implemented a package that attempts to assist programming for AC features including socket management, logging, and recovery of distributed computation. In the experiments, we have not only realized the powerful Java capabilities that are unknown to many educators, we also illustrated the feasibility of learning AC in senior computer science courses.