Currently, software engineering is becoming even more complex due to distributed computing. In this new context, portability is one of the key issues and hence a cluster-aware Java Virtual Machine (JVM) that can transparently execute Java applications in a distributed fashion on nodes of a cluster, while providing the programmer with the single system image of a classical JVM, is really desirable. This way multi-threaded server applications can take advantage of cluster resources without increasing their programming complexity. However, such kind of JVM is not easy to design. Moreover, one of the most challenging tasks in its design is the development of an efficient, scalable and automatic dynamic memory manager. Inside this manager, one important module is the automatic recycling mechanism or garbage collector. This collector is a module with very intensive processing demands that must concurrently run with user’s application. It can consume a very significant portion of the total execution time spent inside JVM in uniprocessor systems, and its overhead increases in distributed garbage collection because of the update of changing references in different nodes. Hence, the garbage collector is a very critical part in distributed designs of JVMs, both for performance and energy. In this paper our contribution to automatic distributed garbage collection is two-fold. First, we have analyzed the barrier mechanism design space for the study of tracing-based distributed garbage collectors. Second, we have evaluated the impact of the most significative barrier strategies as main bottlenecks in global performance. Our preliminary results show that the choice of the specific technique used in barrier mechanisms produces significant differences both in performance and internodes messaging overhead.