AspNetCore.Middleware | Utilitary middleware for ASP.NET Core applications | Runtime Evironment library

I believe that registering LocalizationContext with transient scope will not help as EFStringLocalizerFactory is registered as singleton anyway.

Correct.

Is there any better/proper way of handling concurrency within IStringLocalizerFactory aside from introducing global locks or other inefficient techniques?

Not as far as I know.

EF Core DbContexts only support one operation at a time, which I think the error message is clear about. The other factor is that the in-memory cache implementation doesn't do any sort of locking, so the lambda expression used to create the cache entry can be executed concurrently by several consumers wanting to read from the cache.

Explicitly locking is the way to go IMO, with two options:

• Around the GetOrCreate method, meaning you can guarantee the EF Core query will only be run once, but no 2 consumers will be able to read from the cache concurrently; or
• Around the EF Core query, meaning you can potentially override an existing cache entry, but consumers can then read from the cache concurrently.

I'd personally go with option 2 and use a SemaphoreSlim instance.

The problem seems to be in your following package references:

You need to mention 'list' in Route not in HttpPost, because that's in your route

In your ExternalLoginCallback method, are you not supposed to signin the external user with IdentityServer, so it can issue its own access token to the client(Angular?)

including code like

After reading the error raised by my .NET application when visiting localhost:5001/swagger it became immediately obvious that there was something wrong in either my models or controllers that prevented the Swagger specification from being generated. Turns out I was attempting to use a float on a model attribute validation when the model attribute was a double.

UserStrengthMovement.cs

OK, @Crypt32 got me on the right path here, but it's not obvious.

1) The problem had nothing to do with the password. It had to do with the fact that I made the self-signed certificate using openssl with algorithms that Windows 10 apparently supports... but Azure doesn't. I created a new certificate on Azure and that worked fine.

2) I also had a problem with the keystore (this was the error about not finding the file), for reasons I do not understand. It seems Azure can't figure out how to create a service account keystore for an app service and that blows my mind. I switched to a blob-based keystore (see here) and all of a sudden things seem to be running again... or at least, not failing to start.

Moral of the story: X509 error messages are worse than worthless. Would it kill MS to make an error message that actually described the problem? Hopefully this answer will help someone else out.

Did you try switching to the VS 2019 agents ? Although i dont see any reason why this wouldn't work based on the dependancies installed on the VS2017 agents (1). Microsoft does suggest using the VS2019 agents for DotNet Core development (2).

Something else to look into, are there other projects in your solution that are referencing an older version of core which might cause this conflict in versions?

1: https://github.com/Microsoft/azure-pipelines-image-generation/blob/master/images/win/Vs2017-Server2016-Readme.md

2: https://docs.microsoft.com/en-us/azure/devops/pipelines/agents/hosted?view=azure-devops#use-a-microsoft-hosted-agent