Indomitable
A lightweight (the actual ship is heavy though), performant, and powerful sharder for Discord.JS
The ShipGirl Project; ⓒ Azur Lane
Features
-
Fast
-
Lightweight
-
Reliable
-
ESM & CommonJS supported
-
Shard Concurrency Support
-
Easy to use promised based IPC
-
No dependencies
-
Very cute (and lazy)
Used in prod by the ff:
Chip (https://chipbot.gg/)
Kashima (https://kashima.saya.moe/)
Or add your own!
Installation
- Stable
npm install indomitable
- Dev
npm install https://github.com/Deivu/Indomitable.git#master
Documentation
🔗 https://indomitable.shipgirl.moe/
Don't forget to read my "Notes" below!
Example Usage
Basic usage
const { Indomitable } = require("indomitable");
const { Client } = require("discord.js");
const token = "your_token";
const manager = new Indomitable({ client: Client, token }).on(
"error",
console.error
);
manager.spawn();Basic usage with more Indomitable Options
const { Indomitable } = require("indomitable");
const { Client } = require("discord.js");
const token = "your_token";
const options = {
// Processes to run
clusterCount: 2,
// Websocket shards to run
shardCount: 8,
// Discord.js options
clientOptions: {
intents: [1 << 0], // Bitwise for GUILD intent only
},
// Auto restart processes that have been killed
// This defaults to false by default unless you specify it
autoRestart: true,
// Discord.js client
client: Client,
// Your bot token
token,
};
const manager = new Indomitable(options).on("error", console.error);
manager.spawn();Broadcasteval as one way to get data across shards
// Saya's note:
// Not recommended as every broadcastEval uses eval() internally
// Consider learning the ipc system of this library in future to get data across your clusters
client.shard
.broadcastEval((client) => client.guilds.cache.size)
.then(console.log);Example of a very basic ipc communication (non repliable)
// Primary Process
indomitable.on("message", (message) => {
if (message.content.op === "something") {
doSomething();
}
});
// ClientWorker Process (your client most likely)
client.shard.send({ content: { op: "something" } }).catch(console.error);Example of a very basic ipc communication (repliable)
// Primary Process
indomitable.on("message", (message) => {
if (message.content.op === "something") {
if (!message.repliable) return; // check if the message is repliable just incase, though it won't error even it is not
const someValue = doSomething();
message.reply(someValue);
}
});
// ClientWorker Process (your client most likely)
client.shard
.send({ content: { op: "something" }, repliable: true })
.then(console.log)
.catch(console.error);You could also do it reversely (main process asking data from clusters instead of clusters asking main process)
// Primary Process
// send to specific cluster
indomitable.send(0, { content: { op: "nya" } }).catch(console.error);
// send to specific cluster with repliable
indomitable
.send(0, { content: { op: "something" }, repliable: true })
.then(console.log)
.catch(console.error);
// broadcast to all clusters
indomitable.broadcast({ content: { op: "meow" } }).catch(console.error);
// broadcast to all clusters with repliable is possible as well
indomitable
.broadcast({ content: { op: "meow" }, repliable: true })
.then(console.log)
.catch(console.error);
// ClientWorker Process (your client most likely)
client.shard.on("message", (message) => {
if (message.content.op === "something") {
if (!message.repliable) return;
const someValue = doSomething();
message.reply(someValue);
}
if (message.content.op === "nya") {
doSomething();
}
if (message.content.op === "meow") {
if (!message.repliable) return;
message.reply("nya");
}
});Reconfiguring Indomitable on the go, to launch more clusters or more shard, or more on both
// Reconfigure to launch more shards based on Discord Recommendation without spawning more clusters
indomitable
.reconfigure()
.then(() => console.log("Done, Indomitable is reconfigured"));
// Reconfigure to launch more clusters based on your value, but leave the shards based on Discord's recommendation
indomitable
.reconfigure({ clusters: 8 })
.then(() => console.log("Done, Indomitable is reconfigured"));
// Reconfigure to launch more clusters or shards based on your values
indomitable
.reconfigure({ clusters: 8, shards: 8 })
.then(() => console.log("Done, Indomitable is reconfigured"));
// Do not run restart() or restartAll() while this is running. It will cause your cluster / clusters to restart twice.
if (!indomitable.isBusy) indomitable.restartAll();
// Do not use reconfigure for just restarting all clusters sequentially, still use .restartAll() for that
indomitable.restartAll();Enabling concurrency handling (across all your shards) (Read Notes Below)
const { Indomitable } = require("indomitable");
const { Client } = require("discord.js");
const token = "your_token";
const options = {
// Processes to run
clusterCount: 2,
// Websocket shards to run
shardCount: 8,
// Discord.js options
clientOptions: {
intents: [1 << 0], // Bitwise for GUILD intent only
},
// Auto restart processes that have been killed
// This defaults to false by default unless you specify it
autoRestart: true,
// Enable max concurrency handling
handleConcurrency: true,
// Your Discord.js client
client: Client,
// Your bot token
token,
};
const manager = new Indomitable(options).on("error", console.error);
manager.spawn();-
Indomitable's concurrency handling will work in any shard # and cluster # regardless
-
Pair this with waitForReady disabled, and you will get amazingly fast boot times, specially for those who have access to big bot sharding.
-
Reduces your identify calls by following global max_concurrency identifies resulting into better connection quality
Notes
-
You don't need to call
client.login('token');yourself, Indomitable will call it for you. -
Extended clients that extend from discord.js client will work, as long as you use
client.login('token');to get your bot running -
Your Discord.JS Client ShardClientUtil is replaced with Indomitable's ShardClientUtil. Refer to our docs for documentation 🔗 https://deivu.github.io/Indomitable/classes/client_ShardClientUtil.ShardClientUtil.html
Other Links
Support (#Development)
Indomitable Options
| Option | Type | Description | Required | Default |
|---|---|---|---|---|
| clusterCount | number or 'auto' | How many clusters we should spawn | No | 'auto' |
| shardCount | number or 'auto' | How many websocket shards we should make | No | 'auto' |
| clientOptions | Object | Discord.JS Client Options | No | {} |
| clusterSettings | Object | Options for the forked process | No | {} |
| ipcTimeout | number | Timeout before we fail a request | No | 30000 |
| spawnTimeout | number | Timeout before we fail a cluster spawn | No | 60000 (multiplied by clusterShardCount internally) |
| spawnDelay | number | Time to wait before spawning a new cluster | No | 5000 |
| autoRestart | boolean | If you want to auto restart the shards that have been killed unintentionally | No | false |
| waitForReady | boolean | If you want to wait for cluster ready before spawning another cluster | No | true |
| handleConcurrency | boolean | Whether you want to handle shard concurrency properly | No | false |
| client | Client | Your Discord.JS non modified OR modified client | Yes | None |
| token | string | The token of your bot | Yes | None |
