Track kubernetes-client issue 1099

[ljnelson]

I need to remove the reflection hacks to work around https://github.com/fabric8io/kubernetes-client/issues/1099 when that issue is finally fixed.

[xguerin]

Looks like this is actually the case since releases past Aug. 28th. Do you have any plan to update the dependencies?

[ljnelson]

I've been holding off because judging by the various bug reports the 4.x line is more unstable than the 3.x line. If you want to try yourself, you can use <dependencyManagement> stanzas in your pom.xml appropriately.

[xguerin]

Thanks that’s actually fine. I am content for the moment using the work around.

[ljnelson]

I've published a mostly untested SNAPSHOT (0.2.3-SNAPSHOT) to Sonatype snapshots that upgrades the Kubernetes client to 4.1.1. I'd be curious to see what you think.

[xguerin]

Smashing 👍 i’ll definitely give it a shot soon. I’ll keep you posted.

[xguerin]

So far all my unit and system tests are passing without issue. I'll move to actual applications over the week-end and let you know.

EDIT Does 4.1.1 solves the missing deserializer registration on your end? I am still getting the same error as with the older version:

java.lang.ClassCastException: io.fabric8.kubernetes.api.model.batch.Job cannot be cast to my.controller.crds.jobs.Job
	at my.controller.crds.jobs.JobController.onAddition(JobController.java:357)
	at my.controller.GenericEventQueueConsumer.accept(GenericEventQueueConsumer.java:28)
	at org.microbean.kubernetes.controller.ResourceTrackingEventQueueConsumer.accept(ResourceTrackingEventQueueConsumer.java:254)
	at org.microbean.kubernetes.controller.ResourceTrackingEventQueueConsumer.accept(ResourceTrackingEventQueueConsumer.java:48)
	at org.microbean.kubernetes.controller.EventQueueCollection.lambda$createEventQueueConsumptionTask$1(EventQueueCollection.java:940)
	at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
	at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
	at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
	at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
	at java.lang.Thread.run(Thread.java:748)

EDIT 2 Other instance of the same issue, without the name clashing:

2019-02-15 22:12:33 ERROR WatchHTTPManager:293 - Could not deserialize watch event: {"apiVersion":"v1","kind":"List", "items": [{"kind":"Export","apiVersion":"my.controller.com/v1alpha1","metadata":{"annotations":{"my.controller.com/applicationName":"apps.pubsub::Exporter","my.controller.com/applicationScope":"Default"},"labels":{"app":"myapp","job":"test"},"name":"test-0-2","namespace":"test","ownerReferences":[{"apiVersion":"my.controller.com/v1alpha1","kind":"Job","blockOwnerDeletion":true,"controller":true,"name":"test"}]},"spec":{"peId":0,"portId":2,"stream":{"properties":null,"exportOperName":"ExportOp2","name":"hcSectorBargains","allowFilter":true,"congestionPolicy":"WAIT"}}}], "metadata": {"resourceVersion":"v1alpha1"}}
com.fasterxml.jackson.databind.JsonMappingException: No resource type found for:my.controller.com/v1alpha1#Export
 at [Source: N/A; line: -1, column: -1] (through reference chain: io.fabric8.kubernetes.api.model.KubernetesList["items"]->java.util.ArrayList[0])
	at com.fasterxml.jackson.databind.JsonMappingException.from(JsonMappingException.java:148)
	at com.fasterxml.jackson.databind.DeserializationContext.mappingException(DeserializationContext.java:867)
	at io.fabric8.kubernetes.internal.KubernetesDeserializer.deserialize(KubernetesDeserializer.java:78)
	at io.fabric8.kubernetes.internal.KubernetesDeserializer.deserialize(KubernetesDeserializer.java:32)
	at com.fasterxml.jackson.databind.deser.std.CollectionDeserializer.deserialize(CollectionDeserializer.java:245)
	at com.fasterxml.jackson.databind.deser.std.CollectionDeserializer.deserialize(CollectionDeserializer.java:217)
	at com.fasterxml.jackson.databind.deser.std.CollectionDeserializer.deserialize(CollectionDeserializer.java:25)
	at com.fasterxml.jackson.databind.deser.SettableBeanProperty.deserialize(SettableBeanProperty.java:520)
	at com.fasterxml.jackson.databind.deser.impl.MethodProperty.deserializeAndSet(MethodProperty.java:95)
	at com.fasterxml.jackson.databind.deser.BeanDeserializer.vanillaDeserialize(BeanDeserializer.java:258)
	at com.fasterxml.jackson.databind.deser.BeanDeserializer.deserialize(BeanDeserializer.java:125)
	at com.fasterxml.jackson.databind.ObjectMapper._readValue(ObjectMapper.java:3708)
	at com.fasterxml.jackson.databind.ObjectMapper.readValue(ObjectMapper.java:2005)
	at com.fasterxml.jackson.databind.ObjectMapper.treeToValue(ObjectMapper.java:2476)
	at io.fabric8.kubernetes.internal.KubernetesDeserializer.deserialize(KubernetesDeserializer.java:80)
	at io.fabric8.kubernetes.internal.KubernetesDeserializer.deserialize(KubernetesDeserializer.java:32)
	at com.fasterxml.jackson.databind.ObjectMapper._readMapAndClose(ObjectMapper.java:3736)
	at com.fasterxml.jackson.databind.ObjectMapper.readValue(ObjectMapper.java:2726)
	at io.fabric8.kubernetes.client.dsl.internal.WatchHTTPManager.readWatchEvent(WatchHTTPManager.java:312)
	at io.fabric8.kubernetes.client.dsl.internal.WatchHTTPManager.onMessage(WatchHTTPManager.java:247)
	at io.fabric8.kubernetes.client.dsl.internal.WatchHTTPManager$2.onResponse(WatchHTTPManager.java:176)
	at okhttp3.RealCall$AsyncCall.execute(RealCall.java:206)
	at okhttp3.internal.NamedRunnable.run(NamedRunnable.java:32)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
	at java.lang.Thread.run(Thread.java:748)

[ljnelson]

I seem to recall that it does not.

[xguerin]

See the latest comment on OP, looks like your workaround is actually a required step :)

[ljnelson]

Interesting; particularly given that the class involved is from a package with internal in the name.

[ljnelson]

OK, sounds like using this internal class is considered not to be a hack (!). See https://github.com/fabric8io/kubernetes-client/issues/1285 as well.

Strictly speaking, the microbean-kubernetes-controller library doesn't itself need to use this ~workaround~ sanctioned ~hack~ approach, so I will close this issue (unless you have something else to add here, @xguerin?) and consider releasing.

[xguerin]

@ljnelson nope, I'am all good 👍