[BUG] Repeated "disconnect_reason": "Too many reconnect attempts: 428|close" causing constant restarts

[tomiheneldra]

Describe the bug
After scanning QR and successfully connecting, the bot runs normally for several minutes to hours. Then, it suddenly disconnects with the error "disconnect_reason": "Too many reconnect attempts: 428|close" and enters a restart loop, reconnecting and quickly disconnecting again with the same 428 error. This has been occurring persistently for the past day.

To Reproduce
Steps to reproduce the behavior:

1. Start the client using Baileys with saved credentials or fresh QR scan.
2. The connection establishes successfully and the bot functions normally.
3. After an unpredictable period (minutes to hours), the connection drops.
4. The client attempts to reconnect automatically but fails repeatedly with error code 428.
5. The cycle of reconnect -> 428 error -> disconnect continues indefinitely.

Expected behavior
The client should either:

• Reconnect successfully and resume normal operation after a temporary disconnect, or
• Fail clearly without entering an infinite restart loop.

A 428 error ("Too many reconnect attempts") should not trigger continuously after a single successful connection session.

Environment :

• Is this on a server? Yes
• What do your connectOptions look like?

const client = makeWASocket({
        logger: pino({
            level: "silent"
        }),
        version: waVersion, // [2, 3000, 1030831524]
        auth: state,
        browser: [ "Chrome", "Windows", "124.0.6367.207" ],
        defaultQueryTimeoutMs: 60000,
        syncFullHistory: false,
        shouldSyncHistoryMessage: () => false,
        online: false,
        markOnlineOnConnect: false,
        connectTimeoutMs: 60000,
        keepAliveIntervalMs: 60000,
        qrTimeout: 60000,
        retryRequestDelayMs: 1000 + Math.floor(Math.random() * 2000), // delay random
        generateHighQualityLinkPreview: false,
        getMessage: async (key) => {
            const msg = store.loadMessage(key.remoteJid, key.id);
            return msg?.message || { conversation: '' };
        },
        patchMessageBeforeSending: async (msg, recipientJids) => {
            await client.uploadPreKeysToServerIfRequired();
            return msg;
        }
        }
    });

Additional context :

• Baileys version: 7.0.0-rc.9 (latest from npm)
• This issue started occurring approximately 24 hours ago. Prior to this, the same code/configuration ran stably for extended periods.
• The error appears to be triggered by WhatsApp's server-side enforcement (428 is likely a custom WhatsApp protocol error), not directly by local network issues.

[tomiheneldra]

Additional Technical Details & Connection Handling Logic:

I have implemented a custom reconnection handler that manages disconnect events. However, the current 428 error appears to bypass my normal reconnection logic and triggers an infinite restart loop.

My current connection handling logic:

if (connection === "close") {
    let reason = new Boom(lastDisconnect?.error)?.output.statusCode;

    // Reset counter for non-network related disconnects
    if (![DisconnectReason.connectionLost, DisconnectReason.connectionClosed].includes(reason)) {
        reconnectAttempts = 0;
    }
    
    if (reason === DisconnectReason.badSession) {
        console.log(chalk.red('Bad Session File, Please Delete Session and Scan Again'));
        process.exit();
    } else if (reason === DisconnectReason.connectionClosed || reason === DisconnectReason.connectionLost) {
        reconnectAttempts++;
        console.log(`Connection ${reason === DisconnectReason.connectionLost ? 'Lost' : 'Closed'} from Server (attempt ${reconnectAttempts}/${MAX_RECONNECT_ATTEMPTS}), reconnecting...`);

        if (reconnectAttempts > MAX_RECONNECT_ATTEMPTS) {
            // Log the error and exit
            const errorLog = {
                timestamp: new Date().toISOString(),
                waktu_indonesia: new Date().toLocaleString('id-ID', { timeZone: 'Asia/Jakarta' }),
                disconnect_reason: `Too many reconnect attempts: ${reason}|${connection} (attempts: ${reconnectAttempts})`,
                pesan: "Bot stopped automatically due to reconnect loop (possible network/server issue)"
            };
            // ... logging implementation
            process.exit(1);
        } else {
            startTomi(); // Normal reconnect
        }
    } else if (reason === DisconnectReason.connectionReplaced) {
        console.log("Connection Replaced, Another New Session Opened, Please Close Current Session First");
        process.exit();
    } else if (reason === DisconnectReason.loggedOut) {
        console.log("Device Logged Out, Please Delete Session folder and Scan Again.");
        process.exit();
    } else if (reason === DisconnectReason.restartRequired) {
        console.log("Restart Required, Restarting...");
        startTomi();
    } else if (reason === DisconnectReason.timedOut) {
        console.log("Connection TimedOut, Reconnecting...");
        startTomi();
    } else {
        // Catch-all for unknown errors
        const errorLog = {
            timestamp: new Date().toISOString(),
            waktu_indonesia: new Date().toLocaleString('id-ID', { timeZone: 'Asia/Jakarta' }),
            disconnect_reason: `Unknown DisconnectReason: ${reason}|${connection}`,
            pesan: "Bot stopped automatically due to undetected error (possibly outdated WA version or other bug)"
        };
        // ... logging implementation
        process.exit(1);
    }
}

Update: I'm using WhatsApp Web version [2.3000.1031800002] from wappconnect.io. This issue started exactly 2 days ago with no changes to my code.

Key questions:

1. Is this a compatibility issue between Baileys 7.0.0-rc.9 and WhatsApp Web 2.3000.1031800002?
2. Should I fix my reconnect logic (add exponential backoff for 428) or is this a Baileys bug?
3. Recommended action: downgrade Baileys, change WhatsApp Web version, or modify code?

Current situation: Error 428 occurs even on first reconnect after clean connection, suggesting either WhatsApp server changes or Baileys RC9 regression.

[tomiheneldra]

Update: Current Setup Details

Baileys Configuration:

const {
    state,
    saveCreds
} = await useMultiFileAuthState(`./${session}`)

const {
    version,
    isLatest
} = await fetchLatestBaileysVersion();
    
console.log(`using WA v${version.join(".")}, isLatest: ${isLatest}`);

Console Output Shows:

using WA v2.3000.1027934701, isLatest: true
tomystarkdiamond connected 
tomystarkdiamond version: 7.0.0-rc.9
Connection Closed from Server (attempt 1/2), reconnecting...
using WA v2.3000.1027934701, isLatest: true
tomystarkdiamond connected
tomystarkdiamond version: 7.0.0-rc.9
Connection Closed from Server (attempt 2/2), reconnecting...

Important Observation:
After exactly 2 reconnection attempts (hitting my MAX_RECONNECT_ATTEMPTS limit), the bot reconnects normally and continues operating without issues until the next disconnect event.

This suggests:

1. The 428 error might be a temporary server-side rate limit that resolves after brief pauses
2. My current MAX_RECONNECT_ATTEMPTS=2 creates the right "cooldown" period unintentionally
3. WhatsApp servers might be rejecting immediate reconnections but accepting them after short delays

[unetwork2010]

const sock = makeWASocket({
auth: state,
printQRInTerminal: false,
logger: P({ level: 'silent' }),
keepAliveIntervalMs: 20000,
connectTimeoutMs: 60000,
defaultQueryTimeoutMs: 60000,
retryRequestDelayMs: 5000,
shouldIgnoreJid: jid =>
isJidBroadcast(jid) ||
isJidNewsletter(jid) ||
isJidMetaAI(jid) ||
isJidGroup(jid)
});

this solved this issue for me...

[tomiheneldra]

Update: ISSUE RESOLVED

Status: ✅ Fixed

I implemented the fix suggested by @unetwork2010 in the comments, and my bot has now been running smoothly for approximately 4 hours without any reconnect loops or 428 errors. The connection remains stable and functional.

What worked:
The solution provided by @unetwork2010 successfully addressed the underlying cause of the 428 error loop.

To the maintainers: Thank you for your attention to this issue and for providing a platform where community members can contribute solutions.

Special thanks to @unetwork2010: Your fix worked perfectly! I really appreciate you taking the time to analyze the issue and share a working solution. 🙏

This issue can now be closed as resolved.

[mustafa-mahmoud-eduncy]

Hello, I have applied the following settings, yet the problem persists and the connection drops within 10 minutes to an hour, with a new error message appearing:
{ "error":" Request Time-out", "message":"QR refs attempts ended", "statusCode":408 }

[benclanet]

Update: ISSUE RESOLVED

Status: ✅ Fixed

I implemented the fix suggested by @unetwork2010 in the comments, and my bot has now been running smoothly for approximately 4 hours without any reconnect loops or 428 errors. The connection remains stable and functional.

What worked: The solution provided by @unetwork2010 successfully addressed the underlying cause of the 428 error loop.

To the maintainers: Thank you for your attention to this issue and for providing a platform where community members can contribute solutions.

Special thanks to @unetwork2010: Your fix worked perfectly! I really appreciate you taking the time to analyze the issue and share a working solution. 🙏

This issue can now be closed as resolved.

Hey @tomiheneldra, could you please explain what exactly made it work? I understand that you used @unetwork2010 's solution but I'd be curious to know what options really made the difference here.
Also, have you had any errors since?

Thank you !

[tomiheneldra]

Hey @tomiheneldra, could you please explain what exactly made it work? I understand that you used @unetwork2010 's solution but I'd be curious to know what options really made the difference here. Also, have you had any errors since?

Thank you !

Sure thing! Could you share the client config you're using? Also, which version of baileys are you on right now?

[unetwork2010]

@mustafa-mahmoud-eduncy @benclanet
Just
Try to remove browser line ..

[DiegoRojasVera]

Hey everyone! 👋

Is anyone else experiencing issues with their bot getting logged out and being unable to reconnect? 🔌

The bot works fine for a while, then suddenly disconnects with a Too many reconnect attempts (428) error or just stops reconnecting altogether — even after restarting the process.

Some things I've already tried:

• Deleting the session files and re-scanning the QR
• Restarting the Node.js process
• Using experimentalStore: true in the provider config

Still getting random disconnections in production. Has anyone found a stable fix for this? Any help is appreciated! 🙏

Using: @bot-whatsapp/provider-baileys (BuilderBot)

[egama]

up