Abstract: Apparatus, methods, and machine-readable articles of manufacture improve reliability and scalability of a messaging system. In some embodiments, the system's configuration data and subscriber information are stored in a central repository of a core node, and replicated to one or more non-core nodes. Data replication may be performed in real-time or near real-time, periodically, sporadically, or otherwise. Nodes may rely on the replicated data to provide the system's subscribers and outside callers with services that do not involve updating of configuration or subscriber data. In this way, the non-core nodes may continue to process calls when the core node is unavailable. For example, the non-core nodes can route calls to the appropriate subscribers without relying on the core. As another example, the non-core nodes can take messages from callers when the core node is unavailable. The messages may be queued for posting and delivery when the core node becomes available.