Diagnosing problems within FunctionMetadataManager.AddMetadataFromCustomProviders

[TWolversonReply]

I am trying to containerise a Logic Apps Standard workflow with nothing but an HTTP trigger as a minimal example, just to have a worked example of how to do it. What I could find on the topic e.g. Azure Tips and Tricks was out of date and doesn't work anymore but I have used this as a basis. Where I am stuck is that my container reaches

Host.Startup[515]
      A host error has occurred during startup operation 'd74d195b-0032-432e-8e69-0d2308ee1cc9'.
      System.TimeoutException: Timeout occurred while retrieving metadata from provider 'Microsoft.Azure.Workflows.WebJobs.Extensions.Initialization.WorkflowFunctionDefinitionProvider'. The operation exceeded the configured timeout of 30 seconds.
         at Microsoft.Azure.WebJobs.Script.FunctionMetadataManager.<>c__DisplayClass30_0.<AddMetadataFromCustomProviders>b__3(Task`1 t) in /src/azure-functions-host/src/WebJobs.Script/Host/FunctionMetadataManager.cs:line 238
....

and restarts indefinitely. There doesn't seem to be any way to see what caused the operation to time out, and this seems to be due to the following:

https://github.com/Azure/azure-functions-host/blob/dev/src/WebJobs.Script/Host/FunctionMetadataManager.cs#L225

var completedTask = Task.WhenAny(getFunctionMetadataFromProviderTask, delayTask).ContinueWith(t =>
                {
                    if (t.Result == getFunctionMetadataFromProviderTask)
                    {
                        return getFunctionMetadataFromProviderTask.Result;
                    }

                    // Timeout case.
                    throw new TimeoutException($"Timeout occurred while retrieving metadata from provider '{functionProvider.GetType().FullName}'. The operation exceeded the configured timeout of {metadataProviderTimeout.TotalSeconds} seconds.");
                });

This means that getFunctionMetadataFromProviderTask can't provide any information as to why it didn't complete and all that is returned is a timeout exception without a reason. This seems to be the same behaviour as Logic Apps hosted in Azure, where generally any syntax error in any workflow makes the app nonresponsive and causes all workflow definitions to vanish from the Portal, with no information given as to why.

Surely the error handling here can be better? If not, could anyone give me guidance as to how to diagnose what is wrong?