Configure a server-side idle timeout
We have observed cases when a customer's application has a high number of idle clients connected, but isn't actively sending commands. In such scenarios, you can exhaust all 65,000 connections with a high number of idle clients. To avoid such scenarios, configure the timeout setting appropriately on the server via Redis-specific parameters. This ensures that the server actively disconnects idle clients to avoid an increase in the number of connections. This setting is not available on serverless caches.