Managed Instance

Getting Started: Managed Instance

A Managed Instance lets you run galactic clusters on separate machines or containers, all coordinated centrally by a galactic Bridge. This setup is ideal for scaling your Discord bot horizontally across multiple servers, VPS, or cloud environments.

Unlike a standalone instance, the managed instance does not launch clusters directly. Instead, it connects to a galactic Bridge server, which centrally manages the assignment of shards and clusters. Each managed instance acts as a "worker", following the Bridge's instructions for what shards and clusters to run.

When multiple managed instances connect to the same Bridge, the Bridge automatically balances the workload—running your bot in a distributed and resilient way.

---

1. Setup Bridge

---

2. Connect a Managed Instance

Use the ManagedInstance class to connect a machine to your Bridge. This process can be repeated on any number of hosts.

import {ManagedInstance} from "galactic.ts";

// Connects this instance to the bridge on localhost:3000
const machine = new ManagedInstance(
    `${__dirname}/bot.js`, // Paths to your bot entry script
    'localhost', // Bridge host
    3000, // Bridge port
    1, // Identifier for this instance
    { key: "value" }, // Optional metadata for this instance, that the bridge can use
    [], // Additional arguments to pass to the bot script
    false // Dev-Mode
);

machine.start(); // Start the managed instance

• The instance contacts the bridge, receives cluster assignments, and launches clusters based on the bridge's instructions.

---

3. Implement Your Bot Logic

The bot logic is the same as for the standalone setup. Integrate Cluster into your extended Discord.js client as demonstrated below.

Create bot.ts

---

Summary

• The Bridge centrally manages all sharding and clustering decisions.
• Each managed instance listens for assignments and can run on different hosts, VMs, or containers.
• With multiple managed instances, galactic automatically distributes your bot’s workload—maximizing scalability and resilience.

---