org.apache.flink.runtime.rpc.akka.AkkaRpcServiceUtils Java Examples

/**
 * Create a RPC service for the task manager.
 *
 * @param configuration The configuration for the TaskManager.
 * @param haServices to use for the task manager hostname retrieval
 */
@VisibleForTesting
static RpcService createRpcService(
	final Configuration configuration,
	final HighAvailabilityServices haServices) throws Exception {

	checkNotNull(configuration);
	checkNotNull(haServices);

	return AkkaRpcServiceUtils.createRemoteRpcService(
		configuration,
		determineTaskManagerBindAddress(configuration, haServices),
		configuration.getString(TaskManagerOptions.RPC_PORT),
		configuration.getString(TaskManagerOptions.BIND_HOST),
		configuration.getOptional(TaskManagerOptions.RPC_BIND_PORT));
}
 

/**
 * Creates a {@link StandaloneLeaderRetrievalService} form the given configuration and the
 * JobManager name. The host and port for the remote Akka URL are retrieved from the provided
 * configuration. Instead of using the standard JobManager Akka name, the provided one is used
 * for the remote Akka URL.
 *
 * @param configuration Configuration instance containing hte host and port information
 * @param resolveInitialHostName If true, resolves the hostname of the StandaloneLeaderRetrievalService
 * @param jobManagerName Name of the JobManager actor
 * @return StandaloneLeaderRetrievalService
 * @throws ConfigurationException if the job manager address cannot be retrieved from the configuration
 * @throws UnknownHostException if the job manager address cannot be resolved
 */
public static StandaloneLeaderRetrievalService createLeaderRetrievalService(
		Configuration configuration,
		boolean resolveInitialHostName,
		String jobManagerName)
	throws ConfigurationException, UnknownHostException {
	Tuple2<String, Integer> hostnamePort = HighAvailabilityServicesUtils.getJobManagerAddress(configuration);

	String jobManagerAkkaUrl = AkkaRpcServiceUtils.getRpcUrl(
		hostnamePort.f0,
		hostnamePort.f1,
		jobManagerName != null ? jobManagerName : JobMaster.JOB_MANAGER_NAME,
		resolveInitialHostName ? AddressResolution.TRY_ADDRESS_RESOLUTION : AddressResolution.NO_ADDRESS_RESOLUTION,
		configuration);

	return new StandaloneLeaderRetrievalService(jobManagerAkkaUrl);
}
 

/**
 * Creates a {@link StandaloneLeaderRetrievalService} form the given configuration and the
 * JobManager name. The host and port for the remote Akka URL are retrieved from the provided
 * configuration. Instead of using the standard JobManager Akka name, the provided one is used
 * for the remote Akka URL.
 *
 * @param configuration Configuration instance containing hte host and port information
 * @param resolveInitialHostName If true, resolves the hostname of the StandaloneLeaderRetrievalService
 * @param jobManagerName Name of the JobManager actor
 * @return StandaloneLeaderRetrievalService
 * @throws ConfigurationException if the job manager address cannot be retrieved from the configuration
 * @throws UnknownHostException if the job manager address cannot be resolved
 */
public static StandaloneLeaderRetrievalService createLeaderRetrievalService(
		Configuration configuration,
		boolean resolveInitialHostName,
		String jobManagerName)
	throws ConfigurationException, UnknownHostException {
	Tuple2<String, Integer> hostnamePort = HighAvailabilityServicesUtils.getJobManagerAddress(configuration);

	String jobManagerAkkaUrl = AkkaRpcServiceUtils.getRpcUrl(
		hostnamePort.f0,
		hostnamePort.f1,
		jobManagerName != null ? jobManagerName : JobMaster.JOB_MANAGER_NAME,
		resolveInitialHostName ? AddressResolution.TRY_ADDRESS_RESOLUTION : AddressResolution.NO_ADDRESS_RESOLUTION,
		configuration);

	return new StandaloneLeaderRetrievalService(jobManagerAkkaUrl);
}
 

/**
 * Factory method to instantiate the remote RPC service.
 *
 * @param configuration Flink configuration.
 * @param bindAddress The address to bind the RPC service to.
 * @param bindPort The port range to bind the RPC service to.
 * @return The instantiated RPC service
 */
protected RpcService createRemoteRpcService(
		Configuration configuration,
		String bindAddress,
		int bindPort) throws Exception {
	return AkkaRpcServiceUtils.remoteServiceBuilder(configuration, bindAddress, String.valueOf(bindPort))
		.withBindAddress(bindAddress)
		.withBindPort(bindPort)
		.withCustomConfig(AkkaUtils.testDispatcherConfig())
		.createAndStart();
}
 

@Test
public void testTryAddressResolution() {
	try {
		AkkaRpcServiceUtils.getRpcUrl(
			NON_EXISTING_HOSTNAME,
			PORT,
			ENDPOINT_NAME,
			HighAvailabilityServicesUtils.AddressResolution.TRY_ADDRESS_RESOLUTION,
			new Configuration());
		fail("This should fail with an UnknownHostException");
	} catch (UnknownHostException ignore) {
		// expected
	}
}
 

@Test
public void testNoAddressResolution() throws UnknownHostException {
	AkkaRpcServiceUtils.getRpcUrl(
		NON_EXISTING_HOSTNAME,
		PORT,
		ENDPOINT_NAME,
		HighAvailabilityServicesUtils.AddressResolution.NO_ADDRESS_RESOLUTION,
		new Configuration());
}
 

protected void initializeServices(Configuration configuration, PluginManager pluginManager) throws Exception {

		LOG.info("Initializing cluster services.");

		synchronized (lock) {
			commonRpcService = AkkaRpcServiceUtils.createRemoteRpcService(
				configuration,
				configuration.getString(JobManagerOptions.ADDRESS),
				getRPCPortRange(configuration),
				configuration.getString(JobManagerOptions.BIND_HOST),
				configuration.getOptional(JobManagerOptions.RPC_BIND_PORT));

			// update the configuration used to create the high availability services
			configuration.setString(JobManagerOptions.ADDRESS, commonRpcService.getAddress());
			configuration.setInteger(JobManagerOptions.PORT, commonRpcService.getPort());

			ioExecutor = Executors.newFixedThreadPool(
				ClusterEntrypointUtils.getPoolSize(configuration),
				new ExecutorThreadFactory("cluster-io"));
			haServices = createHaServices(configuration, ioExecutor);
			blobServer = new BlobServer(configuration, haServices.createBlobStore());
			blobServer.start();
			heartbeatServices = createHeartbeatServices(configuration);
			metricRegistry = createMetricRegistry(configuration, pluginManager);

			final RpcService metricQueryServiceRpcService = MetricUtils.startRemoteMetricsRpcService(configuration, commonRpcService.getAddress());
			metricRegistry.startQueryService(metricQueryServiceRpcService, null);

			final String hostname = RpcUtils.getHostname(commonRpcService);

			processMetricGroup = MetricUtils.instantiateProcessMetricGroup(
				metricRegistry,
				hostname,
				ConfigurationUtils.getSystemResourceMetricsProbingInterval(configuration));

			archivedExecutionGraphStore = createSerializableExecutionGraphStore(configuration, commonRpcService.getScheduledExecutor());
		}
	}
 

private static RpcService startMetricRpcService(
		Configuration configuration, AkkaRpcServiceUtils.AkkaRpcServiceBuilder rpcServiceBuilder) throws Exception {
	final int threadPriority = configuration.getInteger(MetricOptions.QUERY_SERVICE_THREAD_PRIORITY);

	return rpcServiceBuilder
		.withActorSystemName(METRICS_ACTOR_SYSTEM_NAME)
		.withActorSystemExecutorConfiguration(new BootstrapTools.FixedThreadPoolExecutorConfiguration(1, 1, threadPriority))
		.createAndStart();
}
 

public Dispatcher(
		RpcService rpcService,
		DispatcherId fencingToken,
		DispatcherBootstrap dispatcherBootstrap,
		DispatcherServices dispatcherServices) throws Exception {
	super(rpcService, AkkaRpcServiceUtils.createRandomName(DISPATCHER_NAME), fencingToken);
	checkNotNull(dispatcherServices);

	this.configuration = dispatcherServices.getConfiguration();
	this.highAvailabilityServices = dispatcherServices.getHighAvailabilityServices();
	this.resourceManagerGatewayRetriever = dispatcherServices.getResourceManagerGatewayRetriever();
	this.heartbeatServices = dispatcherServices.getHeartbeatServices();
	this.blobServer = dispatcherServices.getBlobServer();
	this.fatalErrorHandler = dispatcherServices.getFatalErrorHandler();
	this.jobGraphWriter = dispatcherServices.getJobGraphWriter();
	this.jobManagerMetricGroup = dispatcherServices.getJobManagerMetricGroup();
	this.metricServiceQueryAddress = dispatcherServices.getMetricQueryServiceAddress();

	this.jobManagerSharedServices = JobManagerSharedServices.fromConfiguration(
		configuration,
		blobServer,
		fatalErrorHandler);

	this.runningJobsRegistry = highAvailabilityServices.getRunningJobsRegistry();

	jobManagerRunnerFutures = new HashMap<>(16);

	this.historyServerArchivist = dispatcherServices.getHistoryServerArchivist();

	this.archivedExecutionGraphStore = dispatcherServices.getArchivedExecutionGraphStore();

	this.jobManagerRunnerFactory = dispatcherServices.getJobManagerRunnerFactory();

	this.jobManagerTerminationFutures = new HashMap<>(2);

	this.shutDownFuture = new CompletableFuture<>();

	this.dispatcherBootstrap = checkNotNull(dispatcherBootstrap);
}
 

/**
 * Factory method to instantiate the remote RPC service.
 *
 * @param configuration Flink configuration.
 * @param externalAddress The external address to access the RPC service.
 * @param externalPortRange The external port range to access the RPC service.
 * @param bindAddress The address to bind the RPC service to.
 * @return The instantiated RPC service
 */
protected RpcService createRemoteRpcService(
	Configuration configuration,
	String externalAddress,
	String externalPortRange,
	String bindAddress) throws Exception {
	return AkkaRpcServiceUtils.remoteServiceBuilder(configuration, externalAddress, externalPortRange)
		.withBindAddress(bindAddress)
		.withCustomConfig(AkkaUtils.testDispatcherConfig())
		.createAndStart();
}
 

/**
 * Create a RPC service for the task manager.
 *
 * @param configuration The configuration for the TaskManager.
 * @param haServices to use for the task manager hostname retrieval
 */
public static RpcService createRpcService(
		final Configuration configuration,
		final HighAvailabilityServices haServices) throws Exception {

	checkNotNull(configuration);
	checkNotNull(haServices);

	final String taskManagerAddress = determineTaskManagerBindAddress(configuration, haServices);
	final String portRangeDefinition = configuration.getString(TaskManagerOptions.RPC_PORT);

	return AkkaRpcServiceUtils.createRpcService(taskManagerAddress, portRangeDefinition, configuration);
}
 

/**
 * Create a RPC service for the task manager.
 *
 * @param configuration The configuration for the TaskManager.
 * @param haServices to use for the task manager hostname retrieval
 */
public static RpcService createRpcService(
		final Configuration configuration,
		final HighAvailabilityServices haServices) throws Exception {

	checkNotNull(configuration);
	checkNotNull(haServices);

	final String taskManagerAddress = determineTaskManagerBindAddress(configuration, haServices);
	final String portRangeDefinition = configuration.getString(TaskManagerOptions.RPC_PORT);

	return AkkaRpcServiceUtils.createRpcService(taskManagerAddress, portRangeDefinition, configuration);
}
 

public static RpcService startMetricsRpcService(Configuration configuration, String hostname) throws Exception {
	final String portRange = configuration.getString(MetricOptions.QUERY_SERVICE_PORT);
	final int threadPriority = configuration.getInteger(MetricOptions.QUERY_SERVICE_THREAD_PRIORITY);

	return AkkaRpcServiceUtils.createRpcService(
		hostname,
		portRange,
		configuration,
		METRICS_ACTOR_SYSTEM_NAME,
		new BootstrapTools.FixedThreadPoolExecutorConfiguration(1, 1, threadPriority));
}
 

public static RpcService startLocalMetricsRpcService(Configuration configuration) throws Exception {
	return startMetricRpcService(configuration, AkkaRpcServiceUtils.localServiceBuilder(configuration));
}
 

public JobMaster(
		RpcService rpcService,
		JobMasterConfiguration jobMasterConfiguration,
		ResourceID resourceId,
		JobGraph jobGraph,
		HighAvailabilityServices highAvailabilityService,
		SlotPoolFactory slotPoolFactory,
		SchedulerFactory schedulerFactory,
		JobManagerSharedServices jobManagerSharedServices,
		HeartbeatServices heartbeatServices,
		JobManagerJobMetricGroupFactory jobMetricGroupFactory,
		OnCompletionActions jobCompletionActions,
		FatalErrorHandler fatalErrorHandler,
		ClassLoader userCodeLoader) throws Exception {

	super(rpcService, AkkaRpcServiceUtils.createRandomName(JOB_MANAGER_NAME));

	this.jobMasterConfiguration = checkNotNull(jobMasterConfiguration);
	this.resourceId = checkNotNull(resourceId);
	this.jobGraph = checkNotNull(jobGraph);
	this.rpcTimeout = jobMasterConfiguration.getRpcTimeout();
	this.highAvailabilityServices = checkNotNull(highAvailabilityService);
	this.blobWriter = jobManagerSharedServices.getBlobWriter();
	this.scheduledExecutorService = jobManagerSharedServices.getScheduledExecutorService();
	this.jobCompletionActions = checkNotNull(jobCompletionActions);
	this.fatalErrorHandler = checkNotNull(fatalErrorHandler);
	this.userCodeLoader = checkNotNull(userCodeLoader);
	this.heartbeatServices = checkNotNull(heartbeatServices);
	this.jobMetricGroupFactory = checkNotNull(jobMetricGroupFactory);

	final String jobName = jobGraph.getName();
	final JobID jid = jobGraph.getJobID();

	log.info("Initializing job {} ({}).", jobName, jid);

	final RestartStrategies.RestartStrategyConfiguration restartStrategyConfiguration =
			jobGraph.getSerializedExecutionConfig()
					.deserializeValue(userCodeLoader)
					.getRestartStrategy();

	this.restartStrategy = RestartStrategyResolving.resolve(restartStrategyConfiguration,
		jobManagerSharedServices.getRestartStrategyFactory(),
		jobGraph.isCheckpointingEnabled());

	log.info("Using restart strategy {} for {} ({}).", this.restartStrategy, jobName, jid);

	resourceManagerLeaderRetriever = highAvailabilityServices.getResourceManagerLeaderRetriever();

	this.slotPool = checkNotNull(slotPoolFactory).createSlotPool(jobGraph.getJobID());

	this.scheduler = checkNotNull(schedulerFactory).createScheduler(slotPool);

	this.registeredTaskManagers = new HashMap<>(4);

	this.backPressureStatsTracker = checkNotNull(jobManagerSharedServices.getBackPressureStatsTracker());
	this.lastInternalSavepoint = null;

	this.jobManagerJobMetricGroup = jobMetricGroupFactory.create(jobGraph);
	this.executionGraph = createAndRestoreExecutionGraph(jobManagerJobMetricGroup);
	this.jobStatusListener = null;

	this.resourceManagerConnection = null;
	this.establishedResourceManagerConnection = null;

	this.accumulators = new HashMap<>();
}
 

public TaskExecutor(
		RpcService rpcService,
		TaskManagerConfiguration taskManagerConfiguration,
		HighAvailabilityServices haServices,
		TaskManagerServices taskExecutorServices,
		HeartbeatServices heartbeatServices,
		TaskManagerMetricGroup taskManagerMetricGroup,
		@Nullable String metricQueryServicePath,
		BlobCacheService blobCacheService,
		FatalErrorHandler fatalErrorHandler) {

	super(rpcService, AkkaRpcServiceUtils.createRandomName(TASK_MANAGER_NAME));

	checkArgument(taskManagerConfiguration.getNumberSlots() > 0, "The number of slots has to be larger than 0.");

	this.taskManagerConfiguration = checkNotNull(taskManagerConfiguration);
	this.heartbeatServices = checkNotNull(heartbeatServices);
	this.taskExecutorServices = checkNotNull(taskExecutorServices);
	this.haServices = checkNotNull(haServices);
	this.fatalErrorHandler = checkNotNull(fatalErrorHandler);
	this.taskManagerMetricGroup = checkNotNull(taskManagerMetricGroup);
	this.blobCacheService = checkNotNull(blobCacheService);
	this.metricQueryServicePath = metricQueryServicePath;

	this.taskSlotTable = taskExecutorServices.getTaskSlotTable();
	this.jobManagerTable = taskExecutorServices.getJobManagerTable();
	this.jobLeaderService = taskExecutorServices.getJobLeaderService();
	this.taskManagerLocation = taskExecutorServices.getTaskManagerLocation();
	this.localStateStoresManager = taskExecutorServices.getTaskManagerStateStore();
	this.networkEnvironment = taskExecutorServices.getNetworkEnvironment();
	this.resourceManagerLeaderRetriever = haServices.getResourceManagerLeaderRetriever();

	this.jobManagerConnections = new HashMap<>(4);

	this.hardwareDescription = HardwareDescription.extractFromSystem(
		taskExecutorServices.getMemoryManager().getMemorySize());

	this.resourceManagerAddress = null;
	this.resourceManagerConnection = null;
	this.currentRegistrationTimeoutId = null;

	this.stackTraceSampleService = new StackTraceSampleService(rpcService.getScheduledExecutor());
}
 

public static HighAvailabilityServices createHighAvailabilityServices(
	Configuration configuration,
	Executor executor,
	AddressResolution addressResolution) throws Exception {

	HighAvailabilityMode highAvailabilityMode = HighAvailabilityMode.fromConfig(configuration);

	switch (highAvailabilityMode) {
		case NONE:
			final Tuple2<String, Integer> hostnamePort = getJobManagerAddress(configuration);

			final String resourceManagerRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
				hostnamePort.f0,
				hostnamePort.f1,
				AkkaRpcServiceUtils.createWildcardName(ResourceManager.RESOURCE_MANAGER_NAME),
				addressResolution,
				configuration);
			final String dispatcherRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
				hostnamePort.f0,
				hostnamePort.f1,
				AkkaRpcServiceUtils.createWildcardName(Dispatcher.DISPATCHER_NAME),
				addressResolution,
				configuration);
			final String webMonitorAddress = getWebMonitorAddress(
				configuration,
				addressResolution);

			return new StandaloneHaServices(
				resourceManagerRpcUrl,
				dispatcherRpcUrl,
				webMonitorAddress);
		case ZOOKEEPER:
			BlobStoreService blobStoreService = BlobUtils.createBlobStoreFromConfig(configuration);

			return new ZooKeeperHaServices(
				ZooKeeperUtils.startCuratorFramework(configuration),
				executor,
				configuration,
				blobStoreService);

		case FACTORY_CLASS:
			return createCustomHAServices(configuration, executor);

		default:
			throw new Exception("Recovery mode " + highAvailabilityMode + " is not supported.");
	}
}
 

public ResourceManager(
		RpcService rpcService,
		ResourceID resourceId,
		HighAvailabilityServices highAvailabilityServices,
		HeartbeatServices heartbeatServices,
		SlotManager slotManager,
		ResourceManagerPartitionTrackerFactory clusterPartitionTrackerFactory,
		JobLeaderIdService jobLeaderIdService,
		ClusterInformation clusterInformation,
		FatalErrorHandler fatalErrorHandler,
		ResourceManagerMetricGroup resourceManagerMetricGroup,
		Time rpcTimeout) {

	super(rpcService, AkkaRpcServiceUtils.createRandomName(RESOURCE_MANAGER_NAME), null);

	this.resourceId = checkNotNull(resourceId);
	this.highAvailabilityServices = checkNotNull(highAvailabilityServices);
	this.heartbeatServices = checkNotNull(heartbeatServices);
	this.slotManager = checkNotNull(slotManager);
	this.jobLeaderIdService = checkNotNull(jobLeaderIdService);
	this.clusterInformation = checkNotNull(clusterInformation);
	this.fatalErrorHandler = checkNotNull(fatalErrorHandler);
	this.resourceManagerMetricGroup = checkNotNull(resourceManagerMetricGroup);

	this.jobManagerRegistrations = new HashMap<>(4);
	this.jmResourceIdRegistrations = new HashMap<>(4);
	this.taskExecutors = new HashMap<>(8);
	this.taskExecutorGatewayFutures = new HashMap<>(8);

	this.jobManagerHeartbeatManager = NoOpHeartbeatManager.getInstance();
	this.taskManagerHeartbeatManager = NoOpHeartbeatManager.getInstance();

	this.clusterPartitionTracker = checkNotNull(clusterPartitionTrackerFactory).get(
		(taskExecutorResourceId, dataSetIds) -> taskExecutors.get(taskExecutorResourceId).getTaskExecutorGateway()
			.releaseClusterPartitions(dataSetIds, rpcTimeout)
			.exceptionally(throwable -> {
				log.debug("Request for release of cluster partitions belonging to data sets {} was not successful.", dataSetIds, throwable);
				throw new CompletionException(throwable);
			})
	);
}
 

@Nonnull
private RpcService createRpcService(Configuration configuration, String bindAddress, String portRange) throws Exception {
	return AkkaRpcServiceUtils.createRpcService(bindAddress, portRange, configuration);
}
 

public TaskExecutor(
		RpcService rpcService,
		TaskManagerConfiguration taskManagerConfiguration,
		HighAvailabilityServices haServices,
		TaskManagerServices taskExecutorServices,
		ExternalResourceInfoProvider externalResourceInfoProvider,
		HeartbeatServices heartbeatServices,
		TaskManagerMetricGroup taskManagerMetricGroup,
		@Nullable String metricQueryServiceAddress,
		BlobCacheService blobCacheService,
		FatalErrorHandler fatalErrorHandler,
		TaskExecutorPartitionTracker partitionTracker,
		BackPressureSampleService backPressureSampleService) {

	super(rpcService, AkkaRpcServiceUtils.createRandomName(TASK_MANAGER_NAME));

	checkArgument(taskManagerConfiguration.getNumberSlots() > 0, "The number of slots has to be larger than 0.");

	this.taskManagerConfiguration = checkNotNull(taskManagerConfiguration);
	this.taskExecutorServices = checkNotNull(taskExecutorServices);
	this.haServices = checkNotNull(haServices);
	this.fatalErrorHandler = checkNotNull(fatalErrorHandler);
	this.partitionTracker = partitionTracker;
	this.taskManagerMetricGroup = checkNotNull(taskManagerMetricGroup);
	this.blobCacheService = checkNotNull(blobCacheService);
	this.metricQueryServiceAddress = metricQueryServiceAddress;
	this.backPressureSampleService = checkNotNull(backPressureSampleService);
	this.externalResourceInfoProvider = checkNotNull(externalResourceInfoProvider);

	this.libraryCacheManager = taskExecutorServices.getLibraryCacheManager();
	this.taskSlotTable = taskExecutorServices.getTaskSlotTable();
	this.jobTable = taskExecutorServices.getJobTable();
	this.jobLeaderService = taskExecutorServices.getJobLeaderService();
	this.unresolvedTaskManagerLocation = taskExecutorServices.getUnresolvedTaskManagerLocation();
	this.localStateStoresManager = taskExecutorServices.getTaskManagerStateStore();
	this.shuffleEnvironment = taskExecutorServices.getShuffleEnvironment();
	this.kvStateService = taskExecutorServices.getKvStateService();
	this.ioExecutor = taskExecutorServices.getIOExecutor();
	this.resourceManagerLeaderRetriever = haServices.getResourceManagerLeaderRetriever();

	this.hardwareDescription = HardwareDescription.extractFromSystem(taskExecutorServices.getManagedMemorySize());

	this.resourceManagerAddress = null;
	this.resourceManagerConnection = null;
	this.currentRegistrationTimeoutId = null;

	final ResourceID resourceId = taskExecutorServices.getUnresolvedTaskManagerLocation().getResourceID();
	this.jobManagerHeartbeatManager = createJobManagerHeartbeatManager(heartbeatServices, resourceId);
	this.resourceManagerHeartbeatManager = createResourceManagerHeartbeatManager(heartbeatServices, resourceId);
}
 

public JobMaster(
		RpcService rpcService,
		JobMasterConfiguration jobMasterConfiguration,
		ResourceID resourceId,
		JobGraph jobGraph,
		HighAvailabilityServices highAvailabilityService,
		SlotPoolFactory slotPoolFactory,
		SchedulerFactory schedulerFactory,
		JobManagerSharedServices jobManagerSharedServices,
		HeartbeatServices heartbeatServices,
		JobManagerJobMetricGroupFactory jobMetricGroupFactory,
		OnCompletionActions jobCompletionActions,
		FatalErrorHandler fatalErrorHandler,
		ClassLoader userCodeLoader,
		SchedulerNGFactory schedulerNGFactory,
		ShuffleMaster<?> shuffleMaster,
		PartitionTrackerFactory partitionTrackerFactory) throws Exception {

	super(rpcService, AkkaRpcServiceUtils.createRandomName(JOB_MANAGER_NAME), null);

	this.jobMasterConfiguration = checkNotNull(jobMasterConfiguration);
	this.resourceId = checkNotNull(resourceId);
	this.jobGraph = checkNotNull(jobGraph);
	this.rpcTimeout = jobMasterConfiguration.getRpcTimeout();
	this.highAvailabilityServices = checkNotNull(highAvailabilityService);
	this.blobWriter = jobManagerSharedServices.getBlobWriter();
	this.scheduledExecutorService = jobManagerSharedServices.getScheduledExecutorService();
	this.jobCompletionActions = checkNotNull(jobCompletionActions);
	this.fatalErrorHandler = checkNotNull(fatalErrorHandler);
	this.userCodeLoader = checkNotNull(userCodeLoader);
	this.schedulerNGFactory = checkNotNull(schedulerNGFactory);
	this.heartbeatServices = checkNotNull(heartbeatServices);
	this.jobMetricGroupFactory = checkNotNull(jobMetricGroupFactory);

	final String jobName = jobGraph.getName();
	final JobID jid = jobGraph.getJobID();

	log.info("Initializing job {} ({}).", jobName, jid);

	resourceManagerLeaderRetriever = highAvailabilityServices.getResourceManagerLeaderRetriever();

	this.slotPool = checkNotNull(slotPoolFactory).createSlotPool(jobGraph.getJobID());

	this.scheduler = checkNotNull(schedulerFactory).createScheduler(slotPool);

	this.registeredTaskManagers = new HashMap<>(4);
	this.partitionTracker = checkNotNull(partitionTrackerFactory)
		.create(resourceID -> {
			Tuple2<TaskManagerLocation, TaskExecutorGateway> taskManagerInfo = registeredTaskManagers.get(resourceID);
			if (taskManagerInfo == null) {
				return Optional.empty();
			}

			return Optional.of(taskManagerInfo.f1);
		});

	this.backPressureStatsTracker = checkNotNull(jobManagerSharedServices.getBackPressureStatsTracker());

	this.shuffleMaster = checkNotNull(shuffleMaster);

	this.jobManagerJobMetricGroup = jobMetricGroupFactory.create(jobGraph);
	this.schedulerNG = createScheduler(jobManagerJobMetricGroup);
	this.jobStatusListener = null;

	this.resourceManagerConnection = null;
	this.establishedResourceManagerConnection = null;

	this.accumulators = new HashMap<>();
	this.taskManagerHeartbeatManager = NoOpHeartbeatManager.getInstance();
	this.resourceManagerHeartbeatManager = NoOpHeartbeatManager.getInstance();
}
 

public static HighAvailabilityServices createHighAvailabilityServices(
	Configuration configuration,
	Executor executor,
	AddressResolution addressResolution) throws Exception {

	HighAvailabilityMode highAvailabilityMode = LeaderRetrievalUtils.getRecoveryMode(configuration);

	switch (highAvailabilityMode) {
		case NONE:
			final Tuple2<String, Integer> hostnamePort = getJobManagerAddress(configuration);

			final String jobManagerRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
				hostnamePort.f0,
				hostnamePort.f1,
				JobMaster.JOB_MANAGER_NAME,
				addressResolution,
				configuration);
			final String resourceManagerRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
				hostnamePort.f0,
				hostnamePort.f1,
				ResourceManager.RESOURCE_MANAGER_NAME,
				addressResolution,
				configuration);
			final String dispatcherRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
				hostnamePort.f0,
				hostnamePort.f1,
				Dispatcher.DISPATCHER_NAME,
				addressResolution,
				configuration);

			final String address = checkNotNull(configuration.getString(RestOptions.ADDRESS),
				"%s must be set",
				RestOptions.ADDRESS.key());
			final int port = configuration.getInteger(RestOptions.PORT);
			final boolean enableSSL = SSLUtils.isRestSSLEnabled(configuration);
			final String protocol = enableSSL ? "https://" : "http://";

			return new StandaloneHaServices(
				resourceManagerRpcUrl,
				dispatcherRpcUrl,
				jobManagerRpcUrl,
				String.format("%s%s:%s", protocol, address, port));
		case ZOOKEEPER:
			BlobStoreService blobStoreService = BlobUtils.createBlobStoreFromConfig(configuration);

			return new ZooKeeperHaServices(
				ZooKeeperUtils.startCuratorFramework(configuration),
				executor,
				configuration,
				blobStoreService);

		case FACTORY_CLASS:
			return createCustomHAServices(configuration, executor);

		default:
			throw new Exception("Recovery mode " + highAvailabilityMode + " is not supported.");
	}
}
 

public TaskExecutor(
		RpcService rpcService,
		TaskManagerConfiguration taskManagerConfiguration,
		HighAvailabilityServices haServices,
		TaskManagerServices taskExecutorServices,
		HeartbeatServices heartbeatServices,
		TaskManagerMetricGroup taskManagerMetricGroup,
		@Nullable String metricQueryServiceAddress,
		BlobCacheService blobCacheService,
		FatalErrorHandler fatalErrorHandler,
		PartitionTable<JobID> partitionTable) {

	super(rpcService, AkkaRpcServiceUtils.createRandomName(TASK_MANAGER_NAME));

	checkArgument(taskManagerConfiguration.getNumberSlots() > 0, "The number of slots has to be larger than 0.");

	this.taskManagerConfiguration = checkNotNull(taskManagerConfiguration);
	this.heartbeatServices = checkNotNull(heartbeatServices);
	this.taskExecutorServices = checkNotNull(taskExecutorServices);
	this.haServices = checkNotNull(haServices);
	this.fatalErrorHandler = checkNotNull(fatalErrorHandler);
	this.partitionTable = partitionTable;
	this.taskManagerMetricGroup = checkNotNull(taskManagerMetricGroup);
	this.blobCacheService = checkNotNull(blobCacheService);
	this.metricQueryServiceAddress = metricQueryServiceAddress;

	this.taskSlotTable = taskExecutorServices.getTaskSlotTable();
	this.jobManagerTable = taskExecutorServices.getJobManagerTable();
	this.jobLeaderService = taskExecutorServices.getJobLeaderService();
	this.taskManagerLocation = taskExecutorServices.getTaskManagerLocation();
	this.localStateStoresManager = taskExecutorServices.getTaskManagerStateStore();
	this.shuffleEnvironment = taskExecutorServices.getShuffleEnvironment();
	this.kvStateService = taskExecutorServices.getKvStateService();
	this.resourceManagerLeaderRetriever = haServices.getResourceManagerLeaderRetriever();

	this.jobManagerConnections = new HashMap<>(4);

	this.hardwareDescription = HardwareDescription.extractFromSystem(
		taskExecutorServices.getMemoryManager().getMemorySize());

	this.resourceManagerAddress = null;
	this.resourceManagerConnection = null;
	this.currentRegistrationTimeoutId = null;

	this.stackTraceSampleService = new StackTraceSampleService(rpcService.getScheduledExecutor());
}
 

public static RpcService startRemoteMetricsRpcService(Configuration configuration, String hostname) throws Exception {
	final String portRange = configuration.getString(MetricOptions.QUERY_SERVICE_PORT);

	return startMetricRpcService(configuration, AkkaRpcServiceUtils.remoteServiceBuilder(configuration, hostname, portRange));
}
 

/**
 * Creates new YarnPreConfiguredMasterHaServices for the given Flink and YARN configuration.
 * This constructor parses the ResourceManager address from the Flink configuration and sets
 * up the HDFS access to store recovery data in the YARN application's working directory.
 *
 * @param config     The Flink configuration of this component / process.
 * @param hadoopConf The Hadoop configuration for the YARN cluster.
 *
 * @throws IOException
 *             Thrown, if the initialization of the Hadoop file system used by YARN fails.
 * @throws IllegalConfigurationException
 *             Thrown, if the Flink configuration does not properly describe the ResourceManager address and port.
 */
public YarnPreConfiguredMasterNonHaServices(
		Configuration config,
		org.apache.hadoop.conf.Configuration hadoopConf,
		HighAvailabilityServicesUtils.AddressResolution addressResolution) throws IOException {

	super(config, hadoopConf);

	// track whether we successfully perform the initialization
	boolean successful = false;

	try {
		// extract the hostname and port of the resource manager
		final String rmHost = config.getString(YarnConfigOptions.APP_MASTER_RPC_ADDRESS);
		final int rmPort = config.getInteger(YarnConfigOptions.APP_MASTER_RPC_PORT);

		if (rmHost == null) {
			throw new IllegalConfigurationException("Config parameter '" +
					YarnConfigOptions.APP_MASTER_RPC_ADDRESS.key() + "' is missing.");
		}
		if (rmPort < 0) {
			throw new IllegalConfigurationException("Config parameter '" +
					YarnConfigOptions.APP_MASTER_RPC_PORT.key() + "' is missing.");
		}
		if (rmPort <= 0 || rmPort >= 65536) {
			throw new IllegalConfigurationException("Invalid value for '" +
					YarnConfigOptions.APP_MASTER_RPC_PORT.key() + "' - port must be in [1, 65535]");
		}

		this.resourceManagerRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
			rmHost,
			rmPort,
			ResourceManager.RESOURCE_MANAGER_NAME,
			addressResolution,
			config);

		this.dispatcherRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
			rmHost,
			rmPort,
			Dispatcher.DISPATCHER_NAME,
			addressResolution,
			config);

		// all well!
		successful = true;
	}
	finally {
		if (!successful) {
			// quietly undo what the parent constructor initialized
			try {
				super.close();
			} catch (Throwable ignored) {}
		}
	}
}
 

/**
 * Creates new YarnPreConfiguredMasterHaServices for the given Flink and YARN configuration.
 * This constructor parses the ResourceManager address from the Flink configuration and sets
 * up the HDFS access to store recovery data in the YARN application's working directory.
 *
 * @param config     The Flink configuration of this component / process.
 * @param hadoopConf The Hadoop configuration for the YARN cluster.
 *
 * @throws IOException
 *             Thrown, if the initialization of the Hadoop file system used by YARN fails.
 * @throws IllegalConfigurationException
 *             Thrown, if the Flink configuration does not properly describe the ResourceManager address and port.
 */
public YarnPreConfiguredMasterNonHaServices(
		Configuration config,
		org.apache.hadoop.conf.Configuration hadoopConf,
		HighAvailabilityServicesUtils.AddressResolution addressResolution) throws IOException {

	super(config, hadoopConf);

	// track whether we successfully perform the initialization
	boolean successful = false;

	try {
		// extract the hostname and port of the resource manager
		final String rmHost = config.getString(YarnConfigOptions.APP_MASTER_RPC_ADDRESS);
		final int rmPort = config.getInteger(YarnConfigOptions.APP_MASTER_RPC_PORT);

		if (rmHost == null) {
			throw new IllegalConfigurationException("Config parameter '" +
					YarnConfigOptions.APP_MASTER_RPC_ADDRESS.key() + "' is missing.");
		}
		if (rmPort < 0) {
			throw new IllegalConfigurationException("Config parameter '" +
					YarnConfigOptions.APP_MASTER_RPC_PORT.key() + "' is missing.");
		}
		if (rmPort <= 0 || rmPort >= 65536) {
			throw new IllegalConfigurationException("Invalid value for '" +
					YarnConfigOptions.APP_MASTER_RPC_PORT.key() + "' - port must be in [1, 65535]");
		}

		this.resourceManagerRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
			rmHost,
			rmPort,
			ResourceManager.RESOURCE_MANAGER_NAME,
			addressResolution,
			config);

		this.dispatcherRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
			rmHost,
			rmPort,
			Dispatcher.DISPATCHER_NAME,
			addressResolution,
			config);

		// all well!
		successful = true;
	}
	finally {
		if (!successful) {
			// quietly undo what the parent constructor initialized
			try {
				super.close();
			} catch (Throwable ignored) {}
		}
	}
}
 

public static HighAvailabilityServices createHighAvailabilityServices(
	Configuration configuration,
	Executor executor,
	AddressResolution addressResolution) throws Exception {

	HighAvailabilityMode highAvailabilityMode = HighAvailabilityMode.fromConfig(configuration);

	switch (highAvailabilityMode) {
		case NONE:
			final Tuple2<String, Integer> hostnamePort = getJobManagerAddress(configuration);

			final String jobManagerRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
				hostnamePort.f0,
				hostnamePort.f1,
				JobMaster.JOB_MANAGER_NAME,
				addressResolution,
				configuration);
			final String resourceManagerRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
				hostnamePort.f0,
				hostnamePort.f1,
				ResourceManager.RESOURCE_MANAGER_NAME,
				addressResolution,
				configuration);
			final String dispatcherRpcUrl = AkkaRpcServiceUtils.getRpcUrl(
				hostnamePort.f0,
				hostnamePort.f1,
				Dispatcher.DISPATCHER_NAME,
				addressResolution,
				configuration);

			final String address = checkNotNull(configuration.getString(RestOptions.ADDRESS),
				"%s must be set",
				RestOptions.ADDRESS.key());
			final int port = configuration.getInteger(RestOptions.PORT);
			final boolean enableSSL = SSLUtils.isRestSSLEnabled(configuration);
			final String protocol = enableSSL ? "https://" : "http://";

			return new StandaloneHaServices(
				resourceManagerRpcUrl,
				dispatcherRpcUrl,
				jobManagerRpcUrl,
				String.format("%s%s:%s", protocol, address, port));
		case ZOOKEEPER:
			BlobStoreService blobStoreService = BlobUtils.createBlobStoreFromConfig(configuration);

			return new ZooKeeperHaServices(
				ZooKeeperUtils.startCuratorFramework(configuration),
				executor,
				configuration,
				blobStoreService);

		case FACTORY_CLASS:
			return createCustomHAServices(configuration, executor);

		default:
			throw new Exception("Recovery mode " + highAvailabilityMode + " is not supported.");
	}
}
 

public JobMaster(
		RpcService rpcService,
		JobMasterConfiguration jobMasterConfiguration,
		ResourceID resourceId,
		JobGraph jobGraph,
		HighAvailabilityServices highAvailabilityService,
		SlotPoolFactory slotPoolFactory,
		SchedulerFactory schedulerFactory,
		JobManagerSharedServices jobManagerSharedServices,
		HeartbeatServices heartbeatServices,
		JobManagerJobMetricGroupFactory jobMetricGroupFactory,
		OnCompletionActions jobCompletionActions,
		FatalErrorHandler fatalErrorHandler,
		ClassLoader userCodeLoader,
		SchedulerNGFactory schedulerNGFactory,
		ShuffleMaster<?> shuffleMaster,
		PartitionTrackerFactory partitionTrackerFactory) throws Exception {

	super(rpcService, AkkaRpcServiceUtils.createRandomName(JOB_MANAGER_NAME));

	this.jobMasterConfiguration = checkNotNull(jobMasterConfiguration);
	this.resourceId = checkNotNull(resourceId);
	this.jobGraph = checkNotNull(jobGraph);
	this.rpcTimeout = jobMasterConfiguration.getRpcTimeout();
	this.highAvailabilityServices = checkNotNull(highAvailabilityService);
	this.blobWriter = jobManagerSharedServices.getBlobWriter();
	this.scheduledExecutorService = jobManagerSharedServices.getScheduledExecutorService();
	this.jobCompletionActions = checkNotNull(jobCompletionActions);
	this.fatalErrorHandler = checkNotNull(fatalErrorHandler);
	this.userCodeLoader = checkNotNull(userCodeLoader);
	this.schedulerNGFactory = checkNotNull(schedulerNGFactory);
	this.heartbeatServices = checkNotNull(heartbeatServices);
	this.jobMetricGroupFactory = checkNotNull(jobMetricGroupFactory);

	final String jobName = jobGraph.getName();
	final JobID jid = jobGraph.getJobID();

	log.info("Initializing job {} ({}).", jobName, jid);

	resourceManagerLeaderRetriever = highAvailabilityServices.getResourceManagerLeaderRetriever();

	this.slotPool = checkNotNull(slotPoolFactory).createSlotPool(jobGraph.getJobID());

	this.scheduler = checkNotNull(schedulerFactory).createScheduler(slotPool);

	this.registeredTaskManagers = new HashMap<>(4);
	this.partitionTracker = checkNotNull(partitionTrackerFactory)
		.create(resourceID -> {
			Tuple2<TaskManagerLocation, TaskExecutorGateway> taskManagerInfo = registeredTaskManagers.get(resourceID);
			if (taskManagerInfo == null) {
				return Optional.empty();
			}

			return Optional.of(taskManagerInfo.f1);
		});

	this.backPressureStatsTracker = checkNotNull(jobManagerSharedServices.getBackPressureStatsTracker());

	this.shuffleMaster = checkNotNull(shuffleMaster);

	this.jobManagerJobMetricGroup = jobMetricGroupFactory.create(jobGraph);
	this.schedulerNG = createScheduler(jobManagerJobMetricGroup);
	this.jobStatusListener = null;

	this.resourceManagerConnection = null;
	this.establishedResourceManagerConnection = null;

	this.accumulators = new HashMap<>();
}
 

@Nonnull
private RpcService createRpcService(Configuration configuration, String bindAddress, String portRange) throws Exception {
	return AkkaRpcServiceUtils.createRpcService(bindAddress, portRange, configuration);
}
 

/**
 * Factory method to instantiate the local RPC service.
 *
 * @param configuration Flink configuration.
 * @return The instantiated RPC service
 */
protected RpcService createLocalRpcService(Configuration configuration) throws Exception {
	return AkkaRpcServiceUtils.localServiceBuilder(configuration)
		.withCustomConfig(AkkaUtils.testDispatcherConfig())
		.createAndStart();
}