The thrown exception's message is completely useless...

Is there an existing issue for this?

• I have searched the existing issues

Describe the bug

When an ASP.NET/gRPC application is first starting, it configures its http or https endpoints. If something is not configured correctly the application throws an error which is completely useless! The offending error is:

System.ObjectDisposedException: Cannot access a disposed object.
Object name: 'IServiceProvider'.
   at Microsoft.Extensions.DependencyInjection.ServiceLookup.ThrowHelper.ThrowObjectDisposedException()
   at Microsoft.Extensions.DependencyInjection.ServiceProvider.GetService(Type serviceType, ServiceProviderEngineScope serviceProviderEngineScope)
   at Microsoft.Extensions.DependencyInjection.ServiceProvider.GetService(Type serviceType)
   at Microsoft.Extensions.DependencyInjection.ServiceProviderServiceExtensions.GetService[T](IServiceProvider provider)
   at Microsoft.Extensions.Hosting.HostingAbstractionsHostExtensions.WaitForShutdownAsync(IHost host, CancellationToken token)
   at Microsoft.Extensions.Hosting.HostingAbstractionsHostExtensions.WaitForShutdown(IHost host)
   at Program.<Main>$(String[] args) in D:\Development\Moby\Dev\MobyFileTransferServices\MobyFileTransferServices\Program.cs:line 206.

I have found that if any of the following occurs, this exception is thrown:

1. If the port is incorrect
2. If the TLS certificate is not found
3. If the TLS certificate's password is not correct
4. and there are others that I will add to this list when I can remember what they are

The internal code (either ASP.NET or gRPC - I am posting this in both forums) knows why it is throwing the exception so why not just say why the exception is being thrown! The Object name: 'IServiceProvider' message is a true red herring and completely useless. As I have come to have more experience on getting this to work, I grow more agitated in the lack of quality of that message. Look, getting SSL to work is hard enough to begin with. Help your developer community by throwing a better, more specific error!!!

For completeness' sage, I am including my Kestrel configuration.
appsettings:

	"Kestrel": {
		"EndpointDefaults": {
			"Protocols": "Http2",
			"SslProtocols": [ "Tls12" ]
		},
		"Endpoints": {
			"Https1": {
				"Certificate": {
					"Password": "123456",
					"Path": "localhost.pfx"
				},
				"Url": "https://*:30054"
			}
		}
	}

program.cs:

	// Add services to the container.
	_ = builder.Services.AddGrpc (grpcServiceOptions => {
		logger?.Debug ("Configuring presence intercepter...");
		grpcServiceOptions.Interceptors.Add<PresenceInterceptor> ();
		_ = grpcServiceOptions.EnableDetailedErrors = true;
	});

There is nothing special about this configuration. It's very simple. Please provide better messages as to why this initialization fails!!!

Regards,
Steve Miller

Expected Behavior

Provide an explanatory message that describes the ACTUAL reason the exception is being thrown!!!

Steps To Reproduce

No response

Exceptions (if any)

No response

.NET Version

7.0.100-rc.2.22477.23 but I'm actually using .NET 6

Anything else?

NOTE: I have downloaded both the gRPC and ASP.NET repositories and can't even find the string "Cannot access a disposed object" in the code... I have no idea where this message is being thrown from...

[davidfowl]

The Object name: 'IServiceProvider' message is a true red herring and completely useless.

I'm not sure that's true. At least, based on the information provided I don't see another explanation for the error as yet. Can you share the entire application or share more of the logic in Program.cs? It's possible something is being disposed prematurely during shutdown causing this problem.

[davidfowl]

PS: Turning on verbose logging is another way to get more information about what might be happening.

Hi David,

Thanks for your input!! Looking at your credentials, you are exactly the right person I need to talk to!!! :-)

1. I have logging set to Trace so I am getting the most verbose output.
2. As I said, this one error is triggered by multiple causes that I listed (missing cert, invalid cert, missing password, password mismatch, incorrect port, and incorrect protocol i.e. https vs http).
3. All of these things trigger this one error, so the message text returned by this error is bogus. As for sharing code, this happens on app startup so the Kestrel config and the tiny delegate for AddGrpc call. With that said, I am trying to setup some other callbacks that I think are triggered when certificate validation is occurring, but so far I'm not getting those breakpoints to hit.

Here is the additional code I added to program.cs:

	_ = builder.Services.AddAuthentication (CertificateAuthenticationDefaults.AuthenticationScheme)
		.AddCertificate (certificateAuthenticationOptions => {
			certificateAuthenticationOptions.AllowedCertificateTypes = CertificateTypes.All;
			certificateAuthenticationOptions.Events = new CertificateAuthenticationEvents {
				OnAuthenticationFailed = certificateAuthenticationFailedContext => {

					certificateAuthenticationFailedContext.NoResult ();

					certificateAuthenticationFailedContext.Response.StatusCode = 403;
					certificateAuthenticationFailedContext.Response.ContentType = "text/plain";
					certificateAuthenticationFailedContext.Response.WriteAsync (certificateAuthenticationFailedContext.Exception.ToString ()).Wait ();

					return Task.CompletedTask;
				},
				OnCertificateValidated = certificateValidatedContext => {

					var claims = new[] {
						new Claim (ClaimTypes.Name, certificateValidatedContext.ClientCertificate.Subject, ClaimValueTypes.String, certificateValidatedContext.Options.ClaimsIssuer)
					};

					certificateValidatedContext.Principal = new ClaimsPrincipal (new ClaimsIdentity (claims, certificateValidatedContext.Scheme.Name));

					Console.WriteLine ($"Client certificate thumbprint: {certificateValidatedContext.ClientCertificate.Thumbprint}");
					Console.WriteLine ($"Client certificate subject: {certificateValidatedContext.ClientCertificate.Subject}");

					certificateValidatedContext.Success ();

					return Task.CompletedTask;
				},
				OnChallenge = certificateChallengeContext => {

					return Task.CompletedTask;
				}
			};
		});

	ServicePointManager.ServerCertificateValidationCallback =
		delegate (object sender, X509Certificate? certificate, X509Chain? chain, SslPolicyErrors sslPolicyErrors) {
			return true;
		};

This is some code I copied from a text, "Microservices Communication in .NET Using gRPC".

I think that somewhere in the ASP.NET code the error manifests itself and it is being caught in another try block and it loses the context of the original message... That is my observation anyway...

Thoughts?
Steve

[davidfowl]

Can you show the try catch? It looks like the exception is coming from WaitForShutdown but I'd like to better understand the code flow that causes this to happen.

The try/catch is not in my code… it’s in your code somewhere. I downloaded the aspnet and gRPC repos and did a string search for exception names and message text and couldn’t find it so I am only speculating, but it’s not in my code. I have overloaded every delegate and callback in my program.cs file as well as setting a specific break in the exceptions window for the exceptions and the code does not stop anywhere during startup.

…

On Sat, Nov 5, 2022 at 11:20 AM David Fowler ***@***.***> wrote: Can you show the try catch? It looks like the exception is coming from WaitForShutdown but I'd like to better understand the code flow that causes this to happen. — Reply to this email directly, view it on GitHub <#44895 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AOPRYYO63JN4RSLTCF7RJNTWG2QNTANCNFSM6AAAAAARX5U7CA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[davidfowl]

Can you reproduce this at will? You should be able to debug the original exception (turning on first chance exceptions in the debugger) and figure out where things are falling apart. All I'm looking for are repro steps so I can debug this myself 😅

Yes, I can. (I do it more than I wish!! ;-) ) I will try to gin up a repo and upload it to github. I will get to that later today! Thanks so much for your help...

Hi @mobycorp. We have added the "Needs: Author Feedback" label to this issue, which indicates that we have an open question for you before we can take further action. This issue will be closed automatically in 7 days if we do not hear back from you by then - please feel free to re-open it if you come back to this issue after that time.

This issue has been automatically marked as stale because it has been marked as requiring author feedback but has not had any activity for 4 days. It will be closed if no further activity occurs within 3 days of this comment. If it is closed, feel free to comment when you are able to provide the additional information and we will re-investigate.

See our Issue Management Policies for more information.

@davidfowl - I was going to come back and update this after my list of issues regarding this error after some of them manifest themselves... I am once again listing this as a gRPC issue (but it may be a purely Kestrel issue).

I was installing an update of one of my services on an Ubuntu box and found this one again. When I use systemctl to stop a service, it does not always close the port that the app was using. If you try to start an app on a port that is locked, you will get this error. Again, the error displayed has NOTHING to do with the actual error!! I have run across at least three instances of this error being propagated and only found them by painstaking debugging. Seeing the ACTUAL cause would have saved me at least three days of development!!!