Digital's Basement•14mo ago

status-sharding cluster died

Hello, so my bot is up and running fine but every now an then I get this in my console and I have to restart my bot.. LOG [SYSTEM] [ClusterID: 0] died This is my code:

const manager = new ClusterManager(path.join(__dirname, 'bot.js'), {
    mode: 'process',
    token: process.env.token,
    totalShards: 2,
    totalClusters: 1
});


manager.on('clusterCreate', (cluster) => {
    ['unhandledRejection', 'uncaughtException', 'uncaughtExceptionMonitor'].forEach((event) => {
        process.on(event, logger.log);
    });

    logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] Launched`);

    cluster.on('ready', () => logger.ready(`[SYSTEM] [ClusterID: ${cluster.id}] is ready`));

    cluster.on('death', () => logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] died`));

    cluster.on('error', () => logger.error(`[SYSTEM] [ClusterID: ${cluster.id}] errored`));

    cluster.on('spawn', () => logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] has spawned`));
});

manager.spawn();

const manager = new ClusterManager(path.join(__dirname, 'bot.js'), {
    mode: 'process',
    token: process.env.token,
    totalShards: 2,
    totalClusters: 1
});


manager.on('clusterCreate', (cluster) => {
    ['unhandledRejection', 'uncaughtException', 'uncaughtExceptionMonitor'].forEach((event) => {
        process.on(event, logger.log);
    });

    logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] Launched`);

    cluster.on('ready', () => logger.ready(`[SYSTEM] [ClusterID: ${cluster.id}] is ready`));

    cluster.on('death', () => logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] died`));

    cluster.on('error', () => logger.error(`[SYSTEM] [ClusterID: ${cluster.id}] errored`));

    cluster.on('spawn', () => logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] has spawned`));
});

manager.spawn();

How can I prevent it / is it possible to "revive" the dead cluster on cluster death?

26 Replies

Digital•14mo ago

1. your code somehow crashes the cluster by exiting the process 2. to add logs for that, you can setup this (image below) in your main and cluster files, so it tells you the issue instead of dying 3. yes, you can use heartbeat system, by passing in heartbeat: { enabled: true } in manager options

notvexiOP•14mo ago

Cant I do something like this? or will this not work:

cluster.on('death', () => {
        logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] died`);

        logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] Attempting to respawn...`);
        cluster.respawn().then(() => {
            logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] Successfully respawned`);
        }).catch((error) => {
            logger.error(`[SYSTEM] [ClusterID: ${cluster.id}] Failed to respawn: ${error}`);
        });
    });

cluster.on('death', () => {
        logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] died`);

        logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] Attempting to respawn...`);
        cluster.respawn().then(() => {
            logger.log(`[SYSTEM] [ClusterID: ${cluster.id}] Successfully respawned`);
        }).catch((error) => {
            logger.error(`[SYSTEM] [ClusterID: ${cluster.id}] Failed to respawn: ${error}`);
        });
    });

Digital•14mo ago

you can, instead of 3rd note, but i still suggest you to prevent crashing of clusters as said in 2nd note

notvexiOP•14mo ago

well the thing is I had the thing you mentioned in step 2 already implemented and no error was showing. I just saw random in my console today LOG [SYSTEM] [ClusterID: 0] died

Digital•14mo ago

well that will prevent it dying next time it wants to

notvexiOP•14mo ago

so what would you do. would wou still implement process.on("warning")?

Digital•14mo ago

well i screenshoted my code above, so, it doesnt hurt to have

notvexiOP•14mo ago

wait you said in your "main and cluster" files.? I only have this (image) in my main file. I start from index.js (where the image is) and there it calls bot.js the actual bot file

Digital•14mo ago

your bot.js file crashes, just put those 3 in both files, like at the bottom or at the top

notvexiOP•14mo ago

ok ill try that but in any case this will work right?

Digital•14mo ago

yes, but difference between that and those 3 lines from my ss is that my 3 lines will prevent death from even happening

notvexiOP•14mo ago

tru wait I am dumb tho i have those as well?

notvexiOP•14mo ago

Digital•14mo ago

put them fully outside of any function, and copy them over to bot.js file too

notvexiOP•14mo ago

okay ty so with these startup files: index.js: https://sourceb.in/zxi1sRWmfG bot.js: https://sourceb.in/wMxyNcdZEW the bot just crashed after 2d, 14h. This is what I seein the console:

notvexiOP•14mo ago

Digital•14mo ago

hm weird that error is when sharding couldnt find the process to terminate on restart

notvexiOP•14mo ago

any suggestions on what I could do to fix it?

Digital•14mo ago

im not sure, i run the same much bigger codebases on it and i never had such issues what's your Client

notvexiOP•14mo ago

its a custom class from which extends status-sharding https://sourceb.in/pYEitfiILS

Digital•14mo ago

my only guess would be that you tripple check that you dont somewhere import a file that is used by manager into client or vice versa, ei: - this below will fail, or if you import Lunio client in some file that manager uses client.js:

const test = require('../functions');

const test = require('../functions');

functions.js:

const CustomManager = require('status-sharding')

const CustomManager = require('status-sharding')

manager:

const test = require('../functions');

const test = require('../functions');

notvexiOP•14mo ago

okay thanks ill check for that then

notvexiOP•13mo ago

Does "status-sharding" support this? https://discordjs.guide/miscellaneous/cache-customization.html#sweeping-caches

discord.js Guide

Imagine a guide... that explores the many possibilities for your discord.js bot.

notvexiOP•13mo ago

https://discordjs.guide/miscellaneous/cache-customization.html#cache-customization

discord.js Guide

Imagine a guide... that explores the many possibilities for your discord.js bot.

Digital•13mo ago

yes if course, we dont interfere with discord.js new ShardingClient({ .... })

notvexiOP•13mo ago

okay thanks

Gaming

Programming

status-sharding cluster died

Did you find this page helpful?