Leaking connection and API POST apps.openshift.io/v1 bug

[nbn]

Hi,

Was using version 4.10.2 without any problems, but failed some consequences migrating to 4.11.1. My usual context is openshift, so I utilize deploymentconfigs and routes quite often, but needed the monitoring stuff included with 4.11

Here's my example code snippet, which creates a deploymentconfig, then retrives it and updates it.

public static void main(String[] args) {
    String namespace = args[0];
    System.out.println ("Hello namespace (" + namespace + ")");
    DefaultOpenShiftClient client = new DefaultOpenShiftClient();
    try {
        createDefaultDeploymentConfigAsInTheExamples(namespace, client);

        DeploymentConfig nginx = client.deploymentConfigs().inNamespace(namespace).withName("nginx").get();
        System.out.println(" I found " + nginx.getMetadata().getName());

        DeploymentConfig replace = client.deploymentConfigs().inNamespace(namespace).createOrReplace(nginx);
        System.out.println("Updated " + replace.getMetadata().getName());
    } finally {
        client.close();
    }
}

Running with 4.10.2, this is not a problem. With 4.11.0 and onward I get the exception below and the console hangs indefinitely..
Does anyone know if I should do things differently (with respect to the leaking and the server config) after migrating, or is this an error.

I run with these versions of openshift.

$ oc version
oc v3.10.0+dd10d17
kubernetes v1.10.0+b81c8f8
features: Basic-Auth

Server https://my-secret-cluster.com
openshift v3.11.170
kubernetes v1.11.0+d4cacc0


 io.fabric8.kubernetes.client.KubernetesClientException: Failure executing: POST at: https://my-secret-cluster.com/apis/apps.openshift.io/v1/namespaces/secret-namespace/deploymentconfigs. Message: the API version in the data (apps.openshift.io/v1) does not match the expected API version (v1). Received status: Status(apiVersion=v1, code=400, details=null, kind=Status, message=the API version in the data (apps.openshift.io/v1) does not match the expected API version (v1), metadata=ListMeta(_continue=null, remainingItemCount=null, resourceVersion=null, selfLink=null, additionalProperties={}), reason=BadRequest, status=Failure, additionalProperties={}).
    at io.fabric8.kubernetes.client.dsl.base.OperationSupport.requestFailure(OperationSupport.java:589)
    at io.fabric8.kubernetes.client.dsl.base.OperationSupport.assertResponseCode(OperationSupport.java:528)
    at io.fabric8.kubernetes.client.dsl.base.OperationSupport.handleResponse(OperationSupport.java:492)
    at io.fabric8.kubernetes.client.dsl.base.OperationSupport.handleResponse(OperationSupport.java:451)
    at io.fabric8.kubernetes.client.dsl.base.OperationSupport.handleCreate(OperationSupport.java:252)
    at io.fabric8.kubernetes.client.dsl.base.BaseOperation.handleCreate(BaseOperation.java:867)
    at io.fabric8.kubernetes.client.dsl.base.BaseOperation.create(BaseOperation.java:356)
    at io.fabric8.kubernetes.client.dsl.base.BaseOperation.createOrReplace(BaseOperation.java:421)
    at io.fabric8.kubernetes.client.dsl.base.BaseOperation.createOrReplace(BaseOperation.java:83)
    at io.fabric8.kubernetes.client.dsl.base.BaseOperation.createOrReplace(BaseOperation.java:415)
    at io.fabric8.kubernetes.client.dsl.base.BaseOperation.lambda$createOrReplaceWithNew$1(BaseOperation.java:387)
    at io.fabric8.openshift.api.model.DoneableDeploymentConfig.done(DoneableDeploymentConfig.java:27)

[rohanKanojia]

Umm, Would it be possible for you to check if this issue is reproducible on master? I think we have provided a fix for this in #2475 . I think we now have a test to check scenario you described:

kubernetes-client/kubernetes-itests/src/test/java/io/fabric8/openshift/DeploymentConfigIT.java

Lines 92 to 102 in fd12aff

	@Test
	public void createOrReplace() {
	// Given
	DeploymentConfig deploymentConfig = client.deploymentConfigs().inNamespace(session.getNamespace()).withName("dc-createorreplace").get();
	// When
	deploymentConfig.getSpec().getTemplate().getSpec().getContainers().get(0).setImage("openshift/hello-openshift:v3.8");
	deploymentConfig = client.deploymentConfigs().inNamespace(session.getNamespace()).createOrReplace(deploymentConfig);
	// Then
	assertNotNull(deploymentConfig);

[nbn]

I have tried to check out and build master, but cannot get past the kubernetes-client project before it has a compilation failure of

[ERROR] Failed to execute goal org.apache.maven.plugins:maven-compiler-plugin:3.8.1:testCompile (default-testCompile) on project kubernetes-client: Compilation failure: Compilation failure:
[ERROR] /C:/kubernetes-client/kubernetes-client/src/test/java/io/fabric8/kubernetes/client/HandlersTest.java:[18,53] package io.fabric8.kubernetes.client.handlers.apps.v1 does not exist
[ERROR] /C:/kubernetes-client/kubernetes-client/src/test/java/io/fabric8/kubernetes/client/ResourceHandlerTest.java:[25,53] package io.fabric8.kubernetes.client.handlers.core.v1 does not exist
[ERROR] -> [Help 1]

I can see the folder gets created, but it is empty. Is this some of the velocity template generators that fails?

Do you publish snapshots anywhere (bintray or something), so I won't need to compile?

[rohanKanojia]

Could you please try https://jitpack.io/

[nbn]

Interesting tool, thanks.

Got it, and it seems to work on master. When is next release scheduled?

[nbn]

Have verified with 4.12.0 that everything works again, thanks.