golang-samber-hot

In-memory caching in Go reduces backend load by storing frequently accessed, low-cardinality resources. This Skill uses the samber/hot library to cache items, automatically deduplicating cache misses and managing TTLs to lower latency.

Beyond standard LRU, Go caching eviction algorithms can include LFU, TinyLFU, W-TinyLFU, ARC, S3FIFO, TwoQueue, SIEVE, and FIFO. This Skill supports selecting among these algorithms to match specific access patterns.

Yes, Prometheus metrics integration is supported for cache observability. This Skill provides fast in-memory caching with built-in Prometheus metrics, allowing you to monitor hit rates, miss rates, and overall cache performance effectively.

To prevent cache stampedes when loading missing keys in Go, this Skill provides loader deduplication. It ensures that concurrent requests for the same missing key result in a single backend fetch, protecting the database from sudden load spikes.

Yes, you can cache missing keys to avoid repeatedly querying the database for non-existent data. This Skill supports missing-key caching, ensuring that queries for absent resources are remembered without hitting the backend repeatedly.

The best way to cache user profiles and session data for a high-frequency Go API is using an in-memory cache with W-TinyLFU eviction. This Skill handles high-frequency fetches of medium-to-low cardinality resources, accelerating API responses.