Improvements to the Caching Abstraction in ASP.NET Core 

Authority is officially improving the caching abstraction in .NET to make it more intuitive and reliable. This blog showcases a sample project with reusable extension methods for Distributed Caching to greatly simplify the repetitive code on object serialization when setting cached values. It also provides guidance on best practices so developers can focus on business logic. We are actively working on bringing this experience into the .NET and our current ETA is post .NET 8, but we wanted to share early progress and get feedback.

Application developers mostly use distributed caches to improve data query performance and to save web application session objects. While the concept of caching sounds straight forward, we heard from .NET developers that it requires a decent amount of experience to use caching properly. The challenges are:

1. The required object serialization code is repetitive and error-prone.
2. It’s still a common error to use sync over async in caching methods with clients such as StackExchange.Redis
3. There is too much overhead from the “boilerplate” type of work in .NET caching methods.

Ideally, the developer should only have to specify what to cache, and the framework should take care of the rest. We are introducing a better caching abstraction in .NET to optimize for common distributed caching and session store caching scenarios. We would like to show you a concept of our solution, hoping to solve some immediate problems and receive feedback. The sample extension methods are written by Marc Gravell, the maintainer of the popular StackExchange.Redis. You can copy the DistributedCacheExtensions.cs into your ASP.NET core web project, configure a distributed cache in Program.cs, and follow the example code to start using the extension methods.

The extension methods sample code is very intuitive if you want to jump directly into it. The rest of the blog dives into the detailed implementation behind the extension methods with an example Web API application for a weather forecast.