Version 20.0.0 is not registering model in processAot task
Describe the bug
Project build with gradle, using org.springframework.boot and org.graalvm.buildtools.native.
As dependency we have, among others:
- io.kubernetes:client-java
- io.kubernetes:client-java-extended
- io.kubernetes:client-java-spring-aot-integration
There are no problems with java build and java runtime, problem is native build with graalvm plugin.
With client versions 19.0.0 and 20.0.0-legacy, during processAot task, all model from io.kubernetes.client.openapi.models package is registered and processed to reflect-config.json.
But not with version 20.0.0, which in native build runtime ends up with root cause (for example):
java.lang.IllegalArgumentException: Type io.kubernetes.client.openapi.models.VersionInfo is instantiated reflectively but was never registered. Register the type by adding "unsafeAllocated" for the type in reflect-config.json.
I assume, that something was forgotten when implementing breaking changes..
I have already rewritten all the calls to the changes brought by version 20.0.0, I would not like to revert everything to the legacy version,
Client Version 20.0.0
Kubernetes Version 1.28.7
Java Version With client 19.0.0 was used Java 17, spring boot 3.1.5 and graalvm build tools 0.9.27. With client 20.0.0 and 20.0.0-legacy was used Java 21, spring boot 3.2.2 and graalvm build tools 0.10.0
Expected behavior Model should be registered with version 20.0.0 in same way as with 19.0.0 or 20.0.0-legacy
same problem
The issue is linked to either a missing annotation on generated model class or the way class are added in the registration class.
Today it is: https://github.com/kubernetes-client/java/blob/f13d4f4064a836fa59cf93e39532d12c2c378ce1/spring-aot/src/main/java/io/kubernetes/client/spring/aot/KubernetesBeanFactoryInitializationAotProcessor.java#L80
so scan of ApiModel but it is not existing anymore in 20 (non legacy). ex: deployment look like: https://github.com/kubernetes-client/java/blob/f13d4f4064a836fa59cf93e39532d12c2c378ce1/kubernetes/src/main/java/io/kubernetes/client/openapi/models/V1Deployment.java#L54-L56
The Kubernetes project currently lacks enough contributors to adequately respond to all issues.
This bot triages un-triaged issues according to the following rules:
- After 90d of inactivity,
lifecycle/staleis applied - After 30d of inactivity since
lifecycle/stalewas applied,lifecycle/rottenis applied - After 30d of inactivity since
lifecycle/rottenwas applied, the issue is closed
You can:
- Mark this issue as fresh with
/remove-lifecycle stale - Close this issue with
/close - Offer to help out with Issue Triage
Please send feedback to sig-contributor-experience at kubernetes/community.
/lifecycle stale
/remove-lifecycle stale
The Kubernetes project currently lacks enough contributors to adequately respond to all issues.
This bot triages un-triaged issues according to the following rules:
- After 90d of inactivity,
lifecycle/staleis applied - After 30d of inactivity since
lifecycle/stalewas applied,lifecycle/rottenis applied - After 30d of inactivity since
lifecycle/rottenwas applied, the issue is closed
You can:
- Mark this issue as fresh with
/remove-lifecycle stale - Close this issue with
/close - Offer to help out with Issue Triage
Please send feedback to sig-contributor-experience at kubernetes/community.
/lifecycle stale
/remove-lifecycle stale
The Kubernetes project currently lacks enough contributors to adequately respond to all issues.
This bot triages un-triaged issues according to the following rules:
- After 90d of inactivity,
lifecycle/staleis applied - After 30d of inactivity since
lifecycle/stalewas applied,lifecycle/rottenis applied - After 30d of inactivity since
lifecycle/rottenwas applied, the issue is closed
You can:
- Mark this issue as fresh with
/remove-lifecycle stale - Close this issue with
/close - Offer to help out with Issue Triage
Please send feedback to sig-contributor-experience at kubernetes/community.
/lifecycle stale
/remove-lifecycle stale
The Kubernetes project currently lacks enough contributors to adequately respond to all issues.
This bot triages un-triaged issues according to the following rules:
- After 90d of inactivity,
lifecycle/staleis applied - After 30d of inactivity since
lifecycle/stalewas applied,lifecycle/rottenis applied - After 30d of inactivity since
lifecycle/rottenwas applied, the issue is closed
You can:
- Mark this issue as fresh with
/remove-lifecycle stale - Close this issue with
/close - Offer to help out with Issue Triage
Please send feedback to sig-contributor-experience at kubernetes/community.
/lifecycle stale
/remove-lifecycle stale
This is finally solved, thanks to new regenerated model since version 24.0.0.