Erin Allison

In theory, the condition that causes this is not possible. This is during re-processing an entity that had its event error or be requeued manually, which means the K8S API...

I'd appreciate if you could put a sample together because it almost sounds like an issue with K8S and I'm curious if I can reproduce the behavior.

That workflow seems perfectly valid to me. Think of a situation where your controller is provisioning an external resource (say a VM on a hypervisor), if it comes time to...

That helps a lot; I'll see why it's doing that

I tried your directions and cannot reproduce. There is no solution in KubeOps/22-06-12 so I tried Kubeops/2022-06-12 and get to where an exception is thrown in StatusModified and then nothing...

I still cannot figure this issue out. The source of the exception occurs at L223 in `EventQueue`, because `ResourceCache` cannot handle an entity passed with a null UID. https://github.com/buehler/dotnet-operator-sdk/blob/704c99c52e6c1d5f0a8bfe08b5f0438e6b92cd00/src/KubeOps/Operator/Controller/EventQueue.cs#L208-L225 I...

I guess to proceed from here, I'd like to see a dump of the value of resourceEvent and newResource in that lambda function at the time the exception is thrown....

I'm not a big fan of the bottom example on the page; it's just a copy-paste of an internal test for the library. For your case, there is no need...

I'm interested in why they didn't put the type argument on the class... if they did that, we could handle this with a factory. On a cursory inspection, I agree...

I would want to possibly benchmark the performance/reliability of the method, but how about storing the output of `GetHashCode` for both the object and the status sub-resource. I know at...