Caching Strategies for High-Throughput APIs

hey everyone im wrestling with optimizing a high-throughput read-heavy api and considering different caching approaches we're seeing massive load on certain endpoints my current thought is to explore an in-memory cache like redis for frequently accessed, non-critical data but im wondering about cache invalidation strategies how do you handle keeping your caches fresh without impacting performance too much any preferred patterns for distributed caching when multiple service instances need to access it