How to implement service discovery and load balancing in Node.js without Consul or Kubernetes

Distributed Node.js applications typically rely on external infrastructure like Consul, Kubernetes, or service mesh sidecars to handle service discovery and load balancing, but an alternative approach embeds the registry directly into the application framework. Using Moleculer, a Node.js microservices framework active since 2017, each process maintains a full local copy of the service catalogue and communicates changes peer-to-peer over a message bus such as NATS. When a service broker starts, it broadcasts its available actions and metadata to all other brokers, which update their local registries accordingly; regular heartbeats keep the catalogue current, and a clean shutdown sends an explicit disconnect packet. Load balancing decisions are made client-side using configurable strategies such as Round Robin or custom zone-preference logic, with no central registry server or additional network hop required. The approach is demonstrated with live scenarios, including a node killed mid-traffic via SIGKILL, to show exactly how the system detects and responds to failures.