Question: Why await inside the repository?

Why do you have to await inside the repository? Couldn’t you just return the task? Example code

[ardalis]

Based on the docs (and summed up here: https://stackoverflow.com/a/43497310/13729 ) I believe having the await within each repository method is appropriate, as it ensures no client of the repository will fire off multiple tasks in parallel.

[eluchsinger]

@valdirviana how does removing the await keyboard lose any async feature?

@KhalilMohammad I am not sure what anti-pattern you’re talking about. I see at least two issues here.

The point of removing the await keyboard was to avoid an unnecessary context switch inside the method* (1st issue). But as @ardalis said, EF does not support two parallel operations so it is appropriate to utilize an await inside the method (2nd issue).

* More explanation

What happens when you await? The execution of the asynchronous code happens in a new thread (let’s say Thread 2). When the asynchronous work is done, the further execution of the code will happen again in the originally executing thread (let’s say Thread 1), if it was originally executed by a UI Thread. This causes a possibly unnecessary context switch between these threads. Usually, you don’t need to proceed executing the code after the await in Thread 1 (unless you need to execute it on the UI-Thread). Also, it provides the possibility to avoid the state-machine created by the async-await keywords by using .ConfigureAwait(false).

Edit

Actually, if you are using .NET Core, your Thread 2 will not necessarily resume on Thread 1. This is only true for applications using “UI-Threads”.