Distributed caching and data invalidation service

Due to the increased load on the main database, it is necessary to implement a gRPC microservice that will act as a "smart" caching layer for the data catalog. The service should not only return data from the cache but also synchronize changes, minimizing the load on the DBMS (avoiding the Cache Stampede problem).

Technology Stack

• Platform: .NET 8 / 9

• Protocol: gRPC

• Tools: StackExchange.Redis, EF Core (Read-Only DB replica)

• Patterns: Cache-Aside, Singleton, DI

Acceptance Criteria

• High performance of endpoints, using asynchrony (async/await) at all levels.

• Correct handling of Redis failures (the service should not crash; it should temporarily switch to direct requests to the DB — Circuit Breaker mechanism).

• Presence of unit tests for multithreaded locking logic.