A comparative study of applicationprogramming interface performancein .NET Framework and .NET Core

Sammanfattning: Many technologies could be applied during the creation of Web applications;thus, developers are left with multiple choices when picking building blocksfor their desired solutions. Often these decisions come down to theresource (such as libraries and online material) availability, community,previous experience, and performance. Microsoft offers some of the mostcommonly used frameworks, currently supporting two .NET implementationsfor building server-side applications: .NET (Core) and .NET Framework.These products share a lot between them, with both being the mainstay in theindustry and even working similarly enough thus a developer experienced inone could switch with little to no drawbacks. This study investigates whetherthe performance differs, making it a potential tiebreaker. This researchwas conducted using both practical experiments and the analysis of relevantliterature. The testing was done on 2 distinct endpoints, each implementingone of the .NET solutions, with great attention placed on the imitation of realworld usage, such that the results are providing actionable information forthe reader, furthermore, the solution was made available on a public Githubrepository: simonbinyamin/api-performance. The outcome was documentedand illustrated for the convenience of the viewer. The findings confirmed theauthor’s hypothesis that .NET Core has overall better performance. However,unexpectedly it was also noted that both implementations experienced highlatency in terms of response time.