Handling High-Throughput Webhook Bursts with Redis Streams and Laravel 11

Modern SaaS platforms rely heavily on third-party webhooks—from payment notifications (Stripe, Razorpay) to messaging events (Meta WhatsApp, Twilio). The challenge with webhooks is that traffic patterns are completely asynchronous and bursty. A single promotional broadcast or seasonal flash sale can trigger thousands of webhooks per second directed at your server. If your API handles each webhook with a standard database transaction, your MySQL connection pool can become exhausted within seconds, resulting in `504 Gateway Timeout` errors. ### 1. The Bottleneck: Direct Database Writes A typical naive controller does the following: 1. Receives HTTP POST request. 2. Begins MySQL transaction. 3. Updates customer record or order status. 4. Writes an audit log. 5. Returns HTTP 200. Under high concurrency, table locks and connection pooling limits cause request queues to pile up in Nginx and PHP-FPM. ### 2. The Solution: Ingestion-Worker Decoupling with Redis Streams In our production systems at Solidrix, we decouple ingestion from processing using Redis Streams: - The API endpoint only validates HMAC signatures and pushes the raw event into a Redis Stream. - Response time is consistently under 10 milliseconds. - Dedicated Laravel daemon queue workers read batches from the stream using consumer groups, maintaining steady, controlled throughput to MySQL. ```php // Ultra-fast Redis Stream ingestion public function ingestWebhook(Request $request) { // Verify signature $this->validateSignature($request); // Append to Redis Stream Redis::xadd('stream:webhooks', '*', [ 'event' => $request->input('event'), 'payload' => json_encode($request->all()), 'timestamp' => microtime(true) ]); return response()->json(['status' => 'queued'], 200); } ``` This pattern provides maximum resiliency, zero dropped events, and smooth resource utilization across our backend infrastructure.