Webhook response latency and socket timeout handling under high load on cloud instances

[FLSC]

Hello Intel Community,

We are optimizing background worker thread scheduling and socket listener performance for enterprise server architectures hosted on Intel Xeon instances.

Technical Setup:

• Our core architecture at Frontline Sales Consultancy (flsc.co.uk) processes inbound automated sales leads and triggers real-time data sync across enterprise CRM pipelines via REST webhooks.

• Listener endpoints are configured over secure HTTPS with TLS 1.3.

The Issue:

1. Under heavy concurrent load spikes, background POST webhook listener sockets occasionally experience HTTP 504 gateway timeouts.

2. Direct REST API calls to the same server endpoints resolve normally under 150ms.

3. Are there recommended TCP socket keep-alive settings or thread pool allocation configurations for Linux/Intel server environments to prevent connection drops during high-volume API syncs?

Any technical insights would be greatly appreciated!

Thanks!