JobManager (flink 1.0-SNAPSHOT API)

java.lang.Object
- org.apache.flink.runtime.jobmanager.JobManager

All Implemented Interfaces:

akka.actor.Actor, FlinkActor, SubmittedJobGraphStore.SubmittedJobGraphListener, LeaderContender, LeaderSessionMessageFilter, LogMessages

Direct Known Subclasses:

YarnJobManager
```
public class JobManager
extends Object
implements FlinkActor, LeaderSessionMessageFilter, LogMessages, LeaderContender, SubmittedJobGraphStore.SubmittedJobGraphListener
```
The job manager is responsible for receiving Flink jobs, scheduling the tasks, gathering the job status and managing the task managers. It is realized as an actor and receives amongst others the following messages:
- RegisterTaskManager is sent by a TaskManager which wants to register at the job manager. A successful registration at the instance manager is acknowledged by AcknowledgeRegistration
- SubmitJob is sent by a client which wants to submit a job to the system. The submit message contains the job description in the form of the JobGraph. The JobGraph is appended to the ExecutionGraph and the corresponding ExecutionJobVertices are scheduled for execution on the TaskManagers.
- CancelJob requests to cancel the job with the specified jobID. A successful cancellation is indicated by CancellationSuccess and a failure by CancellationFailure
- UpdateTaskExecutionState is sent by a TaskManager to update the state of an ExecutionVertex contained in the ExecutionGraph. A successful update is acknowledged by true and otherwise false.
- RequestNextInputSplit requests the next input split for a running task on a TaskManager. The assigned input split or null is sent to the sender in the form of the message NextInputSplit.
- JobStatusChanged indicates that the status of job (RUNNING, CANCELING, FINISHED, etc.) has changed. This message is sent by the ExecutionGraph.

Nested Class Summary
- Nested classes/interfaces inherited from interface akka.actor.Actor
  akka.actor.Actor.emptyBehavior$

Constructor Summary

Constructors
Constructor and Description
JobManager(Configuration flinkConfiguration, ExecutorService executorService, InstanceManager instanceManager, Scheduler scheduler, BlobLibraryCacheManager libraryCacheManager, akka.actor.ActorRef archive, RestartStrategy defaultRestartStrategy, scala.concurrent.duration.FiniteDuration timeout, LeaderElectionService leaderElectionService, SubmittedJobGraphStore submittedJobGraphs, CheckpointRecoveryFactory checkpointRecoveryFactory, SavepointStore savepointStore, scala.concurrent.duration.FiniteDuration jobRecoveryTimeout)

Method Summary

All Methods Static Methods Instance Methods Concrete Methods
Modifier and Type	Method and Description
`static String`	`ARCHIVE_NAME()` Name of the archive actor
`protected akka.actor.ActorRef`	`archive()`
`protected CheckpointRecoveryFactory`	`checkpointRecoveryFactory()`
`static scala.Tuple12<ExecutorService,InstanceManager,Scheduler,BlobLibraryCacheManager,RestartStrategy,scala.concurrent.duration.FiniteDuration,Object,LeaderElectionService,SubmittedJobGraphStore,CheckpointRecoveryFactory,SavepointStore,scala.concurrent.duration.FiniteDuration>`	`createJobManagerComponents(Configuration configuration, scala.Option<LeaderElectionService> leaderElectionServiceOption)` Create the job manager components as (instanceManager, scheduler, libraryCacheManager, archiverProps, defaultExecutionRetries, delayBetweenRetries, timeout)
`protected scala.collection.mutable.HashMap<JobID,scala.Tuple2<ExecutionGraph,JobInfo>>`	`currentJobs()` Either running or not yet archived jobs (session hasn't been ended).
`protected RestartStrategy`	`defaultRestartStrategy()`
`protected scala.concurrent.ExecutionContext`	`executionContext()` The extra execution context, for futures, with a custom logging reporter
`protected ExecutorService`	`executorService()`
`protected Configuration`	`flinkConfiguration()`
`scala.Option<scala.collection.Seq<scala.concurrent.Future<scala.runtime.BoxedUnit>>>`	`futuresToComplete()` Futures which have to be completed before terminating the job manager
`String`	`getAddress()` Returns the address of the `LeaderContender` under which other instances can connect to it.
`static akka.actor.ActorRef`	`getJobManagerActorRef(InetSocketAddress address, akka.actor.ActorSystem system, Configuration config)` Resolves the JobManager actor reference in a blocking fashion.
`static akka.actor.ActorRef`	`getJobManagerActorRef(InetSocketAddress address, akka.actor.ActorSystem system, scala.concurrent.duration.FiniteDuration timeout)` Resolves the JobManager actor reference in a blocking fashion.
`static akka.actor.ActorRef`	`getJobManagerActorRef(String jobManagerUrl, akka.actor.ActorSystem system, scala.concurrent.duration.FiniteDuration timeout)` Resolves the JobManager actor reference in a blocking fashion.
`static scala.concurrent.Future<akka.actor.ActorRef>`	`getJobManagerActorRefFuture(InetSocketAddress address, akka.actor.ActorSystem system, scala.concurrent.duration.FiniteDuration timeout)`
`static String`	`getJobManagerAkkaURL(akka.actor.ActorSystem system, scala.Option<String> name)`
`static String`	`getLocalJobManagerAkkaURL(scala.Option<String> name)` Builds the akka actor path for the JobManager actor to address the actor within its own actor system.
`static String`	`getRemoteJobManagerAkkaURL(Configuration config)` Returns the JobManager actor's remote Akka URL, given the configured hostname and port.
`static String`	`getRemoteJobManagerAkkaURL(InetSocketAddress address, scala.Option<String> name)` Builds the akka actor path for the JobManager actor, given the socket address where the JobManager's actor system runs.
`void`	`grantLeadership(UUID newLeaderSessionID)` Callback method which is called by the `LeaderElectionService` upon selecting this instance as the new leader.
`void`	`handleError(Exception exception)` Handles error occurring in the leader election service
`scala.PartialFunction<Object,scala.runtime.BoxedUnit>`	`handleMessage()` Central work method of the JobManager actor.
`protected InstanceManager`	`instanceManager()`
`static String`	`JOB_MANAGER_NAME()` Name of the JobManager actor
`protected scala.concurrent.duration.FiniteDuration`	`jobRecoveryTimeout()`
`protected LeaderElectionService`	`leaderElectionService()`
`scala.Option<UUID>`	`leaderSessionID()`
`protected BlobLibraryCacheManager`	`libraryCacheManager()`
`grizzled.slf4j.Logger`	`log()`
`static grizzled.slf4j.Logger`	`LOG()`
`static void`	`main(String[] args)` Entry point (main method) to run the JobManager in a standalone fashion.
`void`	`onAddedJobGraph(JobID jobId)` Callback for `SubmittedJobGraph` instances added by a different `SubmittedJobGraphStore` instance.
`void`	`onRemovedJobGraph(JobID jobId)` Callback for `SubmittedJobGraph` instances removed by a different `SubmittedJobGraphStore` instance.
`static scala.Tuple4<Configuration,JobManagerMode,String,Iterator<Integer>>`	`parseArgs(String[] args)` Loads the configuration, execution mode and the listening address from the provided command line arguments.
`void`	`postStop()`
`void`	`preStart()` Run when the job manager is started.
`protected RecoveryMode`	`recoveryMode()`
`static <T> scala.util.Try<T>`	`retryOnBindException(scala.Function0<T> fn, scala.Function0<Object> stopCond, long maxSleepBetweenRetries)` Retries a function if it fails because of a `BindException`.
`void`	`revokeLeadership()` Callback method which is called by the `LeaderElectionService` upon revoking the leadership of a former leader.
`static void`	`runJobManager(Configuration configuration, JobManagerMode executionMode, String listeningAddress, int listeningPort)` Starts and runs the JobManager with all its components.
`static void`	`runJobManager(Configuration configuration, JobManagerMode executionMode, String listeningAddress, Iterator<Integer> listeningPortRange)` Starts and runs the JobManager with all its components trying to bind to a port in the specified range.
`static int`	`RUNTIME_FAILURE_RETURN_CODE()`
`protected SavepointStore`	`savepointStore()`
`protected Scheduler`	`scheduler()`
`static akka.actor.ActorRef`	`startActor(akka.actor.Props props, akka.actor.ActorSystem actorSystem)`
`static scala.Tuple4<akka.actor.ActorSystem,akka.actor.ActorRef,akka.actor.ActorRef,scala.Option<WebMonitor>>`	`startActorSystemAndJobManagerActors(Configuration configuration, JobManagerMode executionMode, String listeningAddress, int listeningPort, Class<? extends JobManager> jobManagerClass, Class<? extends MemoryArchivist> archiveClass)` Starts an ActorSystem, the JobManager and all its components including the WebMonitor.
`static scala.Tuple2<akka.actor.ActorRef,akka.actor.ActorRef>`	`startJobManagerActors(Configuration configuration, akka.actor.ActorSystem actorSystem, Class<? extends JobManager> jobManagerClass, Class<? extends MemoryArchivist> archiveClass)` Starts the JobManager and job archiver based on the given configuration, in the given actor system.
`static scala.Tuple2<akka.actor.ActorRef,akka.actor.ActorRef>`	`startJobManagerActors(Configuration configuration, akka.actor.ActorSystem actorSystem, scala.Option<String> jobMangerActorName, scala.Option<String> archiveActorName, Class<? extends JobManager> jobManagerClass, Class<? extends MemoryArchivist> archiveClass)` Starts the JobManager and job archiver based on the given configuration, in the given actor system.
`static int`	`STARTUP_FAILURE_RETURN_CODE()`
`protected SubmittedJobGraphStore`	`submittedJobGraphs()`
`protected scala.concurrent.duration.FiniteDuration`	`timeout()`
`void`	`unhandled(Object message)` Handle unmatched messages with an exception.
`int`	`webMonitorPort()` The port of the web monitor as configured.

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface org.apache.flink.runtime.LeaderSessionMessageFilter
decorateMessage, handleDiscardedMessage, receive

Methods inherited from interface org.apache.flink.runtime.LogMessages
receive

Methods inherited from interface akka.actor.Actor
akka$actor$Actor$_setter_$context_$eq, akka$actor$Actor$_setter_$self_$eq, aroundPostRestart, aroundPostStop, aroundPreRestart, aroundPreStart, aroundReceive, context, postRestart, preRestart, self, sender, supervisorStrategy

Constructor Detail

JobManager

public JobManager(Configuration flinkConfiguration,
                  ExecutorService executorService,
                  InstanceManager instanceManager,
                  Scheduler scheduler,
                  BlobLibraryCacheManager libraryCacheManager,
                  akka.actor.ActorRef archive,
                  RestartStrategy defaultRestartStrategy,
                  scala.concurrent.duration.FiniteDuration timeout,
                  LeaderElectionService leaderElectionService,
                  SubmittedJobGraphStore submittedJobGraphs,
                  CheckpointRecoveryFactory checkpointRecoveryFactory,
                  SavepointStore savepointStore,
                  scala.concurrent.duration.FiniteDuration jobRecoveryTimeout)

Method Detail
- LOG
```
public static grizzled.slf4j.Logger LOG()
```
- STARTUP_FAILURE_RETURN_CODE
```
public static int STARTUP_FAILURE_RETURN_CODE()
```
- RUNTIME_FAILURE_RETURN_CODE
```
public static int RUNTIME_FAILURE_RETURN_CODE()
```
- JOB_MANAGER_NAME
```
public static String JOB_MANAGER_NAME()
```
  Name of the JobManager actor
- ARCHIVE_NAME
```
public static String ARCHIVE_NAME()
```
  Name of the archive actor
- main
```
public static void main(String[] args)
```
  Entry point (main method) to run the JobManager in a standalone fashion.
  
  Parameters:
  
  args - The command line arguments.
- runJobManager
```
public static void runJobManager(Configuration configuration,
                                 JobManagerMode executionMode,
                                 String listeningAddress,
                                 int listeningPort)
```
  Starts and runs the JobManager with all its components. First, this method starts a dedicated actor system for the JobManager. Second, its starts all components of the JobManager (including library cache, instance manager, scheduler). Finally, it starts the JobManager actor itself.
  This method blocks indefinitely (or until the JobManager's actor system is shut down).
  
  Parameters:
  
  configuration - The configuration object for the JobManager.
  
  executionMode - The execution mode in which to run. Execution mode LOCAL will spawn an an additional TaskManager in the same process.
  
  listeningAddress - The hostname where the JobManager should listen for messages.
  
  listeningPort - The port where the JobManager should listen for messages.
- runJobManager
```
public static void runJobManager(Configuration configuration,
                                 JobManagerMode executionMode,
                                 String listeningAddress,
                                 Iterator<Integer> listeningPortRange)
```
  Starts and runs the JobManager with all its components trying to bind to a port in the specified range.
  
  Parameters:
  
  configuration - The configuration object for the JobManager.
  
  executionMode - The execution mode in which to run. Execution mode LOCAL will spawn an an additional TaskManager in the same process.
  
  listeningAddress - The hostname where the JobManager should listen for messages.
  
  listeningPortRange - The port range where the JobManager should listen for messages.
- retryOnBindException
```
public static <T> scala.util.Try<T> retryOnBindException(scala.Function0<T> fn,
                                                         scala.Function0<Object> stopCond,
                                                         long maxSleepBetweenRetries)
```
  Retries a function if it fails because of a BindException.
  
  Parameters:
  
  fn - The function to retry
  
  stopCond - Flag to signal termination
  
  maxSleepBetweenRetries - Max random sleep time between retries
  
  Returns:
  
  Return value of the the function to retry
- startActorSystemAndJobManagerActors
```
public static scala.Tuple4<akka.actor.ActorSystem,akka.actor.ActorRef,akka.actor.ActorRef,scala.Option<WebMonitor>> startActorSystemAndJobManagerActors(Configuration configuration,
                                                                                                                                                        JobManagerMode executionMode,
                                                                                                                                                        String listeningAddress,
                                                                                                                                                        int listeningPort,
                                                                                                                                                        Class<? extends JobManager> jobManagerClass,
                                                                                                                                                        Class<? extends MemoryArchivist> archiveClass)
```
  Starts an ActorSystem, the JobManager and all its components including the WebMonitor.
  
  Parameters:
  
  configuration - The configuration object for the JobManager
  
  executionMode - The execution mode in which to run. Execution mode LOCAL with spawn an additional TaskManager in the same process.
  
  listeningAddress - The hostname where the JobManager should listen for messages.
  
  listeningPort - The port where the JobManager should listen for messages
  
  jobManagerClass - The class of the JobManager to be started
  
  archiveClass - The class of the Archivist to be started
  
  Returns:
  
  A tuple containing the started ActorSystem, ActorRefs to the JobManager and the Archivist and an Option containing a possibly started WebMonitor
- parseArgs
```
public static scala.Tuple4<Configuration,JobManagerMode,String,Iterator<Integer>> parseArgs(String[] args)
```
  Loads the configuration, execution mode and the listening address from the provided command line arguments.
  
  Parameters:
  
  args - command line arguments
  
  Returns:
  
  Quadruple of configuration, execution mode and an optional listening address
- createJobManagerComponents
```
public static scala.Tuple12<ExecutorService,InstanceManager,Scheduler,BlobLibraryCacheManager,RestartStrategy,scala.concurrent.duration.FiniteDuration,Object,LeaderElectionService,SubmittedJobGraphStore,CheckpointRecoveryFactory,SavepointStore,scala.concurrent.duration.FiniteDuration> createJobManagerComponents(Configuration configuration,
                                                                                                                                                                                                                                                                                                                         scala.Option<LeaderElectionService> leaderElectionServiceOption)
```
  Create the job manager components as (instanceManager, scheduler, libraryCacheManager, archiverProps, defaultExecutionRetries, delayBetweenRetries, timeout)
  
  Parameters:
  
  configuration - The configuration from which to parse the config values.
  
  leaderElectionServiceOption - LeaderElectionService which shall be returned if the option is defined
  
  Returns:
  
  The members for a default JobManager.
- startJobManagerActors
```
public static scala.Tuple2<akka.actor.ActorRef,akka.actor.ActorRef> startJobManagerActors(Configuration configuration,
                                                                                          akka.actor.ActorSystem actorSystem,
                                                                                          Class<? extends JobManager> jobManagerClass,
                                                                                          Class<? extends MemoryArchivist> archiveClass)
```
  Starts the JobManager and job archiver based on the given configuration, in the given actor system.
  
  Parameters:
  
  configuration - The configuration for the JobManager
  
  actorSystem - The actor system running the JobManager
  
  jobManagerClass - The class of the JobManager to be started
  
  archiveClass - The class of the MemoryArchivist to be started
  
  Returns:
  
  A tuple of references (JobManager Ref, Archiver Ref)
- startJobManagerActors
```
public static scala.Tuple2<akka.actor.ActorRef,akka.actor.ActorRef> startJobManagerActors(Configuration configuration,
                                                                                          akka.actor.ActorSystem actorSystem,
                                                                                          scala.Option<String> jobMangerActorName,
                                                                                          scala.Option<String> archiveActorName,
                                                                                          Class<? extends JobManager> jobManagerClass,
                                                                                          Class<? extends MemoryArchivist> archiveClass)
```
  Starts the JobManager and job archiver based on the given configuration, in the given actor system.
  
  Parameters:
  
  configuration - The configuration for the JobManager
  
  actorSystem - The actor system running the JobManager
  
  jobMangerActorName - Optionally the name of the JobManager actor. If none is given, the actor will have the name generated by the actor system.
  
  archiveActorName - Optionally the name of the archive actor. If none is given, the actor will have the name generated by the actor system.
  
  jobManagerClass - The class of the JobManager to be started
  
  archiveClass - The class of the MemoryArchivist to be started
  
  Returns:
  
  A tuple of references (JobManager Ref, Archiver Ref)
- startActor
```
public static akka.actor.ActorRef startActor(akka.actor.Props props,
                                             akka.actor.ActorSystem actorSystem)
```
- getRemoteJobManagerAkkaURL
```
public static String getRemoteJobManagerAkkaURL(InetSocketAddress address,
                                                scala.Option<String> name)
```
  Builds the akka actor path for the JobManager actor, given the socket address where the JobManager's actor system runs.
  
  Parameters:
  
  address - The address of the JobManager's actor system.
  
  Returns:
  
  The akka URL of the JobManager actor.
- getRemoteJobManagerAkkaURL
```
public static String getRemoteJobManagerAkkaURL(Configuration config)
```
  Returns the JobManager actor's remote Akka URL, given the configured hostname and port.
  
  Parameters:
  
  config - The configuration to parse
  
  Returns:
  
  JobManager actor remote Akka URL
- getLocalJobManagerAkkaURL
```
public static String getLocalJobManagerAkkaURL(scala.Option<String> name)
```
  Builds the akka actor path for the JobManager actor to address the actor within its own actor system.
  
  Returns:
  
  The local akka URL of the JobManager actor.
- getJobManagerAkkaURL
```
public static String getJobManagerAkkaURL(akka.actor.ActorSystem system,
                                          scala.Option<String> name)
```
- getJobManagerActorRefFuture
```
public static scala.concurrent.Future<akka.actor.ActorRef> getJobManagerActorRefFuture(InetSocketAddress address,
                                                                                       akka.actor.ActorSystem system,
                                                                                       scala.concurrent.duration.FiniteDuration timeout)
```
- getJobManagerActorRef
```
public static akka.actor.ActorRef getJobManagerActorRef(String jobManagerUrl,
                                                        akka.actor.ActorSystem system,
                                                        scala.concurrent.duration.FiniteDuration timeout)
```
  Resolves the JobManager actor reference in a blocking fashion.
  
  Parameters:
  
  jobManagerUrl - The akka URL of the JobManager.
  
  system - The local actor system that should perform the lookup.
  
  timeout - The maximum time to wait until the lookup fails.
  
  Returns:
  
  The ActorRef to the JobManager
  
  Throws:
  
  IOException - Thrown, if the lookup fails.
- getJobManagerActorRef
```
public static akka.actor.ActorRef getJobManagerActorRef(InetSocketAddress address,
                                                        akka.actor.ActorSystem system,
                                                        scala.concurrent.duration.FiniteDuration timeout)
```
  Resolves the JobManager actor reference in a blocking fashion.
  
  Parameters:
  
  address - The socket address of the JobManager's actor system.
  
  system - The local actor system that should perform the lookup.
  
  timeout - The maximum time to wait until the lookup fails.
  
  Returns:
  
  The ActorRef to the JobManager
  
  Throws:
  
  IOException - Thrown, if the lookup fails.
- getJobManagerActorRef
```
public static akka.actor.ActorRef getJobManagerActorRef(InetSocketAddress address,
                                                        akka.actor.ActorSystem system,
                                                        Configuration config)
```
  Resolves the JobManager actor reference in a blocking fashion.
  
  Parameters:
  
  address - The socket address of the JobManager's actor system.
  
  system - The local actor system that should perform the lookup.
  
  config - The config describing the maximum time to wait until the lookup fails.
  
  Returns:
  
  The ActorRef to the JobManager
  
  Throws:
  
  IOException - Thrown, if the lookup fails.
- flinkConfiguration
```
protected Configuration flinkConfiguration()
```
- executorService
```
protected ExecutorService executorService()
```
- instanceManager
```
protected InstanceManager instanceManager()
```
- scheduler
```
protected Scheduler scheduler()
```
- libraryCacheManager
```
protected BlobLibraryCacheManager libraryCacheManager()
```
- archive
```
protected akka.actor.ActorRef archive()
```
- defaultRestartStrategy
```
protected RestartStrategy defaultRestartStrategy()
```
- timeout
```
protected scala.concurrent.duration.FiniteDuration timeout()
```
- leaderElectionService
```
protected LeaderElectionService leaderElectionService()
```
- submittedJobGraphs
```
protected SubmittedJobGraphStore submittedJobGraphs()
```
- checkpointRecoveryFactory
```
protected CheckpointRecoveryFactory checkpointRecoveryFactory()
```
- savepointStore
```
protected SavepointStore savepointStore()
```
- jobRecoveryTimeout
```
protected scala.concurrent.duration.FiniteDuration jobRecoveryTimeout()
```
- log
```
public grizzled.slf4j.Logger log()
```
  Specified by:
  
  log in interface FlinkActor
- executionContext
```
protected scala.concurrent.ExecutionContext executionContext()
```
  The extra execution context, for futures, with a custom logging reporter
- currentJobs
```
protected scala.collection.mutable.HashMap<JobID,scala.Tuple2<ExecutionGraph,JobInfo>> currentJobs()
```
  Either running or not yet archived jobs (session hasn't been ended).
- recoveryMode
```
protected RecoveryMode recoveryMode()
```
- leaderSessionID
```
public scala.Option<UUID> leaderSessionID()
```
  Specified by:
  
  leaderSessionID in interface LeaderSessionMessageFilter
- futuresToComplete
```
public scala.Option<scala.collection.Seq<scala.concurrent.Future<scala.runtime.BoxedUnit>>> futuresToComplete()
```
  Futures which have to be completed before terminating the job manager
- webMonitorPort
```
public int webMonitorPort()
```
  The port of the web monitor as configured. Make sure that it is actually configured before starting the JobManager. This tightly couples the web monitor with the job manager. It is a temporary workaround until all execution graph components are properly serializable and all web monitors can transparently interact with each job manager. Currently each web server has to run in the actor system of the associated job manager.
- preStart
```
public void preStart()
```
  Run when the job manager is started. Simply logs an informational message. The method also starts the leader election service.
  
  Specified by:
  
  preStart in interface akka.actor.Actor
- postStop
```
public void postStop()
```
  Specified by:
  
  postStop in interface akka.actor.Actor
- handleMessage
```
public scala.PartialFunction<Object,scala.runtime.BoxedUnit> handleMessage()
```
  Central work method of the JobManager actor. Receives messages and reacts to them.
  
  Specified by:
  
  handleMessage in interface FlinkActor
  
  Returns:
- unhandled
```
public void unhandled(Object message)
```
  Handle unmatched messages with an exception.
  
  Specified by:
  
  unhandled in interface akka.actor.Actor
- grantLeadership
```
public void grantLeadership(UUID newLeaderSessionID)
```
  Description copied from interface: LeaderContender
  
  Callback method which is called by the LeaderElectionService upon selecting this instance as the new leader. The method is called with the new leader session ID.
  
  Specified by:
  
  grantLeadership in interface LeaderContender
  
  Parameters:
  
  newLeaderSessionID - New leader session ID
- revokeLeadership
```
public void revokeLeadership()
```
  Description copied from interface: LeaderContender
  
  Callback method which is called by the LeaderElectionService upon revoking the leadership of a former leader. This might happen in case that multiple contenders have been granted leadership.
  
  Specified by:
  
  revokeLeadership in interface LeaderContender
- onAddedJobGraph
```
public void onAddedJobGraph(JobID jobId)
```
  Description copied from interface: SubmittedJobGraphStore.SubmittedJobGraphListener
  
  Callback for SubmittedJobGraph instances added by a different SubmittedJobGraphStore instance.
  Important: It is possible to get false positives and be notified about a job graph, which was added by this instance.
  
  Specified by:
  
  onAddedJobGraph in interface SubmittedJobGraphStore.SubmittedJobGraphListener
  
  Parameters:
  
  jobId - The JobID of the added job graph
- onRemovedJobGraph
```
public void onRemovedJobGraph(JobID jobId)
```
  Description copied from interface: SubmittedJobGraphStore.SubmittedJobGraphListener
  
  Callback for SubmittedJobGraph instances removed by a different SubmittedJobGraphStore instance.
  
  Specified by:
  
  onRemovedJobGraph in interface SubmittedJobGraphStore.SubmittedJobGraphListener
  
  Parameters:
  
  jobId - The JobID of the removed job graph
- getAddress
```
public String getAddress()
```
  Description copied from interface: LeaderContender
  
  Returns the address of the LeaderContender under which other instances can connect to it.
  
  Specified by:
  
  getAddress in interface LeaderContender
  
  Returns:
  
  Address of this contender.
- handleError
```
public void handleError(Exception exception)
```
  Handles error occurring in the leader election service
  
  Specified by:
  
  handleError in interface LeaderContender
  
  Parameters:
  
  exception - Exception being thrown in the leader election service

Back to Flink Website

Class JobManager

Nested Class Summary

Nested classes/interfaces inherited from interface akka.actor.Actor

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Methods inherited from interface org.apache.flink.runtime.LeaderSessionMessageFilter

Methods inherited from interface org.apache.flink.runtime.LogMessages

Methods inherited from interface akka.actor.Actor

Constructor Detail

JobManager

Method Detail

LOG

STARTUP_FAILURE_RETURN_CODE

RUNTIME_FAILURE_RETURN_CODE

JOB_MANAGER_NAME

ARCHIVE_NAME

main

runJobManager

runJobManager

retryOnBindException

startActorSystemAndJobManagerActors

parseArgs

createJobManagerComponents

startJobManagerActors

startJobManagerActors

startActor

getRemoteJobManagerAkkaURL

getRemoteJobManagerAkkaURL

getLocalJobManagerAkkaURL

getJobManagerAkkaURL

getJobManagerActorRefFuture

getJobManagerActorRef

getJobManagerActorRef

getJobManagerActorRef

flinkConfiguration

executorService

instanceManager

scheduler

libraryCacheManager

archive

defaultRestartStrategy

timeout

leaderElectionService

submittedJobGraphs

checkpointRecoveryFactory

savepointStore

jobRecoveryTimeout

log

executionContext

currentJobs

recoveryMode

leaderSessionID

futuresToComplete

webMonitorPort

preStart

postStop

handleMessage

unhandled

grantLeadership

revokeLeadership

onAddedJobGraph

onRemovedJobGraph

getAddress

handleError

Back to Flink Website