tillrohrmann commented on a change in pull request #17485: URL: https://github.com/apache/flink/pull/17485#discussion_r791723572
########## File path: flink-kubernetes/src/main/java/org/apache/flink/kubernetes/highavailability/KubernetesMultipleComponentLeaderElectionHaServices.java ########## @@ -0,0 +1,253 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you under the Apache License, Version 2.0 (the + * "License"); you may not use this file except in compliance + * with the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package org.apache.flink.kubernetes.highavailability; + +import org.apache.flink.api.common.JobID; +import org.apache.flink.configuration.Configuration; +import org.apache.flink.kubernetes.configuration.KubernetesConfigOptions; +import org.apache.flink.kubernetes.configuration.KubernetesLeaderElectionConfiguration; +import org.apache.flink.kubernetes.kubeclient.FlinkKubeClient; +import org.apache.flink.kubernetes.kubeclient.KubernetesConfigMapSharedWatcher; +import org.apache.flink.kubernetes.utils.KubernetesUtils; +import org.apache.flink.runtime.blob.BlobStoreService; +import org.apache.flink.runtime.checkpoint.CheckpointRecoveryFactory; +import org.apache.flink.runtime.highavailability.AbstractHaServices; +import org.apache.flink.runtime.highavailability.RunningJobsRegistry; +import org.apache.flink.runtime.jobmanager.JobGraphStore; +import org.apache.flink.runtime.leaderelection.DefaultLeaderElectionService; +import org.apache.flink.runtime.leaderelection.DefaultMultipleComponentLeaderElectionService; +import org.apache.flink.runtime.leaderelection.LeaderElectionService; +import org.apache.flink.runtime.leaderelection.MultipleComponentLeaderElectionService; +import org.apache.flink.runtime.leaderretrieval.DefaultLeaderRetrievalService; +import org.apache.flink.runtime.leaderretrieval.LeaderRetrievalService; +import org.apache.flink.runtime.rpc.FatalErrorHandler; +import org.apache.flink.util.ExceptionUtils; +import org.apache.flink.util.ExecutorUtils; +import org.apache.flink.util.FlinkRuntimeException; +import org.apache.flink.util.concurrent.ExecutorThreadFactory; + +import javax.annotation.Nullable; +import javax.annotation.concurrent.GuardedBy; + +import java.util.UUID; +import java.util.concurrent.Executor; +import java.util.concurrent.ExecutorService; +import java.util.concurrent.Executors; +import java.util.concurrent.TimeUnit; + +import static org.apache.flink.kubernetes.utils.Constants.LABEL_CONFIGMAP_TYPE_HIGH_AVAILABILITY; +import static org.apache.flink.kubernetes.utils.Constants.NAME_SEPARATOR; +import static org.apache.flink.util.Preconditions.checkNotNull; + +/** Kubernetes HA services that use a single leader election service per JobManager. */ +public class KubernetesMultipleComponentLeaderElectionHaServices extends AbstractHaServices { + + private final Object lock = new Object(); + + private final String clusterId; + + private final FlinkKubeClient kubeClient; + + private final KubernetesConfigMapSharedWatcher configMapSharedWatcher; + private final ExecutorService watchExecutorService; + + private final String lockIdentity; + + private final FatalErrorHandler fatalErrorHandler; + + @Nullable + @GuardedBy("lock") + private DefaultMultipleComponentLeaderElectionService multipleComponentLeaderElectionService = + null; + + KubernetesMultipleComponentLeaderElectionHaServices( + FlinkKubeClient kubeClient, + Executor executor, + Configuration config, + BlobStoreService blobStoreService, + FatalErrorHandler fatalErrorHandler) { + + super(config, executor, blobStoreService); + this.kubeClient = checkNotNull(kubeClient); + this.clusterId = checkNotNull(config.get(KubernetesConfigOptions.CLUSTER_ID)); + this.fatalErrorHandler = checkNotNull(fatalErrorHandler); + + this.configMapSharedWatcher = + this.kubeClient.createConfigMapSharedWatcher( + KubernetesUtils.getConfigMapLabels( + clusterId, LABEL_CONFIGMAP_TYPE_HIGH_AVAILABILITY)); + this.watchExecutorService = + Executors.newCachedThreadPool( + new ExecutorThreadFactory("config-map-watch-handler")); + + lockIdentity = UUID.randomUUID().toString(); + } + + @Override + protected LeaderElectionService createLeaderElectionService(String leaderName) { + final MultipleComponentLeaderElectionService multipleComponentLeaderElectionService = + getOrInitializeSingleLeaderElectionService(); + + return new DefaultLeaderElectionService( + multipleComponentLeaderElectionService.createDriverFactory(leaderName)); + } + + private DefaultMultipleComponentLeaderElectionService + getOrInitializeSingleLeaderElectionService() { + synchronized (lock) { + if (multipleComponentLeaderElectionService == null) { + try { + + final KubernetesLeaderElectionConfiguration leaderElectionConfiguration = + new KubernetesLeaderElectionConfiguration( + getClusterConfigMap(), lockIdentity, configuration); + multipleComponentLeaderElectionService = + new DefaultMultipleComponentLeaderElectionService( + fatalErrorHandler, + "Single leader election service", + new KubernetesMultipleComponentLeaderElectionDriverFactory( + kubeClient, + leaderElectionConfiguration, + configMapSharedWatcher, + watchExecutorService, + fatalErrorHandler)); + } catch (Exception e) { + throw new FlinkRuntimeException( + "Could not initialize the default single leader election service.", e); + } + } + + return multipleComponentLeaderElectionService; + } + } + + @Override + protected LeaderRetrievalService createLeaderRetrievalService(String componentId) { + return new DefaultLeaderRetrievalService( + new KubernetesMultipleComponentLeaderRetrievalDriverFactory( + kubeClient, + configMapSharedWatcher, + watchExecutorService, + getClusterConfigMap(), + componentId)); + } + + @Override + protected CheckpointRecoveryFactory createCheckpointRecoveryFactory() { + return new KubernetesCheckpointRecoveryFactory( + kubeClient, configuration, ioExecutor, this::getJobSpecificConfigMap, lockIdentity); + } + + private String getJobSpecificConfigMap(JobID jobID) { + return clusterId + NAME_SEPARATOR + jobID.toString() + NAME_SEPARATOR + "config-map"; + } + + @Override + protected JobGraphStore createJobGraphStore() throws Exception { + return KubernetesUtils.createJobGraphStore( + configuration, kubeClient, getClusterConfigMap(), lockIdentity); + } + + private String getClusterConfigMap() { + return clusterId + NAME_SEPARATOR + "cluster-config-map"; + } + + @Override + protected RunningJobsRegistry createRunningJobsRegistry() { + return new KubernetesRunningJobsRegistry(kubeClient, getClusterConfigMap(), lockIdentity); + } + + @Override + public void internalClose() throws Exception { + Exception exception = null; + try { + closeK8sServices(); + } catch (Exception e) { + exception = e; + } + + kubeClient.close(); + ExecutorUtils.gracefulShutdown(5, TimeUnit.SECONDS, this.watchExecutorService); + + ExceptionUtils.tryRethrowException(exception); + } + + private void closeK8sServices() throws Exception { + Exception exception = null; + synchronized (lock) { + if (multipleComponentLeaderElectionService != null) { + try { + multipleComponentLeaderElectionService.close(); + } catch (Exception e) { + exception = e; + } + multipleComponentLeaderElectionService = null; + } + } + + configMapSharedWatcher.close(); + + ExceptionUtils.tryRethrowException(exception); + } + + @Override + public void internalCleanup() throws Exception { + Exception exception = null; + // in order to clean up, we first need to stop the services that rely on the config maps + try { + closeK8sServices(); + } catch (Exception e) { + exception = e; + } + + kubeClient + .deleteConfigMapsByLabels( + KubernetesUtils.getConfigMapLabels( + clusterId, LABEL_CONFIGMAP_TYPE_HIGH_AVAILABILITY)) + .get(); + + ExceptionUtils.tryRethrowException(exception); + } + + @Override + public void internalCleanupJobData(JobID jobID) throws Exception { + kubeClient.deleteConfigMap(getJobSpecificConfigMap(jobID)).get(); + // need to delete job specific leader address from leader config map + } + + @Override + protected String getLeaderPathForResourceManager() { + return "resourcemanager"; Review comment: `KubernetesHaServices` will hopefully be removed in the next version. Hence, this will probably be work that can be spared. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
