程序包 org.apache.flink.kubernetes.operator.service

类 NativeFlinkService

java.lang.Object

org.apache.flink.kubernetes.operator.service.AbstractFlinkService

org.apache.flink.kubernetes.operator.service.NativeFlinkService

所有已实现的接口:

FlinkService

public class NativeFlinkService
extends AbstractFlinkService

Implementation of FlinkService submitting and interacting with Native Kubernetes Flink clusters and jobs.

嵌套类概要

从接口继承的嵌套类/接口 org.apache.flink.kubernetes.operator.service.FlinkService

FlinkService.ScalingResult

字段概要

从类继承的字段 org.apache.flink.kubernetes.operator.service.AbstractFlinkService

artifactManager, executorService, FIELD_NAME_TOTAL_CPU, FIELD_NAME_TOTAL_MEMORY, kubernetesClient, operatorConfig

构造器概要

构造器

构造器	说明
NativeFlinkService(io.fabric8.kubernetes.client.KubernetesClient kubernetesClient, ArtifactManager artifactManager, java.util.concurrent.ExecutorService executorService, FlinkOperatorConfiguration operatorConfig, EventRecorder eventRecorder)

方法概要

修饰符和类型	方法	说明
void	cancelJob(org.apache.flink.kubernetes.operator.api.FlinkDeployment deployment, org.apache.flink.kubernetes.operator.api.spec.UpgradeMode upgradeMode, org.apache.flink.configuration.Configuration configuration)
protected void	deleteClusterInternal(io.fabric8.kubernetes.api.model.ObjectMeta meta, org.apache.flink.configuration.Configuration conf, boolean deleteHaData, io.fabric8.kubernetes.api.model.DeletionPropagation deletionPropagation)	Delete Flink kubernetes cluster by deleting the kubernetes resources directly.
protected void	deployApplicationCluster(org.apache.flink.kubernetes.operator.api.spec.JobSpec jobSpec, org.apache.flink.configuration.Configuration conf)
protected io.fabric8.kubernetes.api.model.PodList	getJmPodList(java.lang.String namespace, java.lang.String clusterId)
protected io.fabric8.kubernetes.api.model.PodList	getTmPodList(java.lang.String namespace, java.lang.String clusterId)
protected java.util.Map<org.apache.flink.runtime.jobgraph.JobVertexID,JobVertexResourceRequirements>	getVertexResources(org.apache.flink.client.program.rest.RestClusterClient<java.lang.String> client, org.apache.flink.kubernetes.operator.api.AbstractFlinkResource<?,?> resource)
FlinkService.ScalingResult	scale(FlinkResourceContext<?> ctx, org.apache.flink.configuration.Configuration deployConfig)
boolean	scalingCompleted(FlinkResourceContext<?> ctx)
protected void	submitClusterInternal(org.apache.flink.configuration.Configuration conf)
void	submitSessionCluster(org.apache.flink.configuration.Configuration conf)
protected void	updateVertexResources(org.apache.flink.client.program.rest.RestClusterClient<java.lang.String> client, org.apache.flink.kubernetes.operator.api.AbstractFlinkResource<?,?> resource, java.util.Map<org.apache.flink.runtime.jobgraph.JobVertexID,JobVertexResourceRequirements> newReqs)

从类继承的方法 org.apache.flink.kubernetes.operator.service.AbstractFlinkService

cancelJob, cancelSessionJob, deleteClusterDeployment, deleteHAData, disposeSavepoint, fetchSavepointInfo, getCheckpointInfo, getClusterClient, getClusterInfo, getEffectiveStatus, getJmPodList, getJobDetailsInfo, getKubernetesClient, getLastCheckpoint, getMetrics, getRestClient, isHaMetadataAvailable, isJobManagerPortReady, listJobs, removeOperatorConfigs, requestJobResult, runJar, submitApplicationCluster, submitJobToSessionCluster, triggerSavepoint, updateStatusAfterClusterDeletion, uploadJar, waitForClusterShutdown

从类继承的方法 java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

构造器详细资料

NativeFlinkService

public NativeFlinkService(io.fabric8.kubernetes.client.KubernetesClient kubernetesClient,
                          ArtifactManager artifactManager,
                          java.util.concurrent.ExecutorService executorService,
                          FlinkOperatorConfiguration operatorConfig,
                          EventRecorder eventRecorder)

方法详细资料

deployApplicationCluster

protected void deployApplicationCluster(org.apache.flink.kubernetes.operator.api.spec.JobSpec jobSpec,
                                        org.apache.flink.configuration.Configuration conf)
                                 throws java.lang.Exception

指定者:

deployApplicationCluster 在类中 AbstractFlinkService

抛出:

java.lang.Exception

submitSessionCluster

public void submitSessionCluster(org.apache.flink.configuration.Configuration conf)
                          throws java.lang.Exception

抛出:

java.lang.Exception

cancelJob

public void cancelJob(org.apache.flink.kubernetes.operator.api.FlinkDeployment deployment,
                      org.apache.flink.kubernetes.operator.api.spec.UpgradeMode upgradeMode,
                      org.apache.flink.configuration.Configuration configuration)
               throws java.lang.Exception

抛出:

java.lang.Exception

getJmPodList

protected io.fabric8.kubernetes.api.model.PodList getJmPodList(java.lang.String namespace,
                                                               java.lang.String clusterId)

指定者:

getJmPodList 在类中 AbstractFlinkService

getTmPodList

protected io.fabric8.kubernetes.api.model.PodList getTmPodList(java.lang.String namespace,
                                                               java.lang.String clusterId)

指定者:

getTmPodList 在类中 AbstractFlinkService

submitClusterInternal

protected void submitClusterInternal(org.apache.flink.configuration.Configuration conf)
                              throws java.lang.Exception

抛出:

java.lang.Exception

deleteClusterInternal

protected void deleteClusterInternal(io.fabric8.kubernetes.api.model.ObjectMeta meta,
                                     org.apache.flink.configuration.Configuration conf,
                                     boolean deleteHaData,
                                     io.fabric8.kubernetes.api.model.DeletionPropagation deletionPropagation)

从类复制的说明: AbstractFlinkService

Delete Flink kubernetes cluster by deleting the kubernetes resources directly. Optionally allows deleting the native kubernetes HA resources as well.

指定者:

deleteClusterInternal 在类中 AbstractFlinkService

参数:

meta - ObjectMeta of the deployment

conf - Configuration of the Flink application

deleteHaData - Flag to indicate whether k8s or Zookeeper HA metadata should be removed as well

deletionPropagation - Resource deletion propagation policy

scale

public FlinkService.ScalingResult scale(FlinkResourceContext<?> ctx,
                                        org.apache.flink.configuration.Configuration deployConfig)
                                 throws java.lang.Exception

抛出:

java.lang.Exception

updateVertexResources

@VisibleForTesting
protected void updateVertexResources(org.apache.flink.client.program.rest.RestClusterClient<java.lang.String> client,
                                     org.apache.flink.kubernetes.operator.api.AbstractFlinkResource<?,?> resource,
                                     java.util.Map<org.apache.flink.runtime.jobgraph.JobVertexID,JobVertexResourceRequirements> newReqs)
                              throws java.lang.Exception

抛出:

java.lang.Exception

getVertexResources

@VisibleForTesting
protected java.util.Map<org.apache.flink.runtime.jobgraph.JobVertexID,JobVertexResourceRequirements> getVertexResources(org.apache.flink.client.program.rest.RestClusterClient<java.lang.String> client,
                                                                                                                              org.apache.flink.kubernetes.operator.api.AbstractFlinkResource<?,?> resource)
                                                                                                                       throws java.lang.Exception

抛出:

java.lang.Exception

scalingCompleted

public boolean scalingCompleted(FlinkResourceContext<?> ctx)