search

binakot / Abstract-Rest-Service-Benchmark

A set of template services in different languages and frameworks. An abstract performance comparison based on the endpoint throughput. There are no objective benefits, just an elementary comparison of primitive RESTful API services. :shipit:
MIT License
16 stars 14 forks source link

Fixes and speed up for .NET Core sample #14

Closed ForNeVeR closed 6 years ago

ForNeVeR commented 6 years ago

On my Windows machine, this improved the performance of .NET Core sample more that twice (from 11k RPS to 24k RPS). Although my current computational environment is very unstable. I'd be happy to hear your results.

Thanks.

ju-bear commented 6 years ago

ASP.NET Core is capable to handle requests by either using Controllers or using request handlers as in your example. Spring Boot service in this benchmark uses RestController to handle requests.

Maybe 2 separate versions of .NET Core service should be made? One that uses Controllers without any MVC-specific features and another that handles the request directly?

Ubloobok commented 6 years ago

@dredd92 it's a great idea about 2 versions for .NET Core, it'll be more correct for demonstration.

ForNeVeR commented 6 years ago

Superseded by #16.