org.apache.flink.runtime.checkpoint

Class StandaloneCompletedCheckpointStore

java.lang.Object

org.apache.flink.runtime.checkpoint.AbstractCompleteCheckpointStore

org.apache.flink.runtime.checkpoint.StandaloneCompletedCheckpointStore

All Implemented Interfaces:

CompletedCheckpointStore

public class StandaloneCompletedCheckpointStore
extends AbstractCompleteCheckpointStore

CompletedCheckpointStore for JobManagers running in HighAvailabilityMode.NONE.

Constructor Summary

Constructors

Constructor and Description
StandaloneCompletedCheckpointStore(int maxNumberOfCheckpointsToRetain)
StandaloneCompletedCheckpointStore(int maxNumberOfCheckpointsToRetain, SharedStateRegistryFactory sharedStateRegistryFactory, Executor ioExecutor, RecoveryClaimMode recoveryClaimMode) Creates StandaloneCompletedCheckpointStore.

Method Summary

Modifier and Type	Method and Description
CompletedCheckpoint	addCheckpointAndSubsumeOldestOne(CompletedCheckpoint checkpoint, CheckpointsCleaner checkpointsCleaner, Runnable postCleanup) Adds a CompletedCheckpoint instance to the list of completed checkpoints.
List<CompletedCheckpoint>	getAllCheckpoints() Returns all CompletedCheckpoint instances.
int	getMaxNumberOfRetainedCheckpoints() Returns the max number of retained checkpoints.
int	getNumberOfRetainedCheckpoints() Returns the current number of retained checkpoints.
boolean	requiresExternalizedCheckpoints() This method returns whether the completed checkpoint store requires checkpoints to be externalized.
void	shutdown(JobStatus jobStatus, CheckpointsCleaner checkpointsCleaner) Shuts down the store.

Methods inherited from class org.apache.flink.runtime.checkpoint.AbstractCompleteCheckpointStore

findLowest, getSharedStateRegistry, unregisterUnusedState

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.checkpoint.CompletedCheckpointStore

getLatestCheckpoint, getLatestCheckpointId

Constructor Detail

StandaloneCompletedCheckpointStore

@VisibleForTesting
public StandaloneCompletedCheckpointStore(int maxNumberOfCheckpointsToRetain)

StandaloneCompletedCheckpointStore

public StandaloneCompletedCheckpointStore(int maxNumberOfCheckpointsToRetain,
                                          SharedStateRegistryFactory sharedStateRegistryFactory,
                                          Executor ioExecutor,
                                          RecoveryClaimMode recoveryClaimMode)

Creates StandaloneCompletedCheckpointStore.

Parameters:

recoveryClaimMode -

maxNumberOfCheckpointsToRetain - The maximum number of checkpoints to retain (at least 1). Adding more checkpoints than this results in older checkpoints being discarded.

Method Detail

addCheckpointAndSubsumeOldestOne

@Nullable
public CompletedCheckpoint addCheckpointAndSubsumeOldestOne(CompletedCheckpoint checkpoint,
                                                                      CheckpointsCleaner checkpointsCleaner,
                                                                      Runnable postCleanup)
                                                               throws Exception

Description copied from interface: CompletedCheckpointStore

Adds a CompletedCheckpoint instance to the list of completed checkpoints.

Only a bounded number of checkpoints is kept. When exceeding the maximum number of retained checkpoints, the oldest one will be discarded.

After FLINK-24611, SharedStateRegistry.unregisterUnusedState(long) should be called here to subsume unused state. Note, the CompletedCheckpoint passed to SharedStateRegistry.registerAllAfterRestored(org.apache.flink.runtime.checkpoint.CompletedCheckpoint, org.apache.flink.core.execution.RecoveryClaimMode) or SharedStateRegistryFactory.create(java.util.concurrent.Executor, java.util.Collection<org.apache.flink.runtime.checkpoint.CompletedCheckpoint>, org.apache.flink.core.execution.RecoveryClaimMode) must be the same object as the input parameter, otherwise the state may be deleted by mistake.

After FLINK-25872, CheckpointsCleaner.cleanSubsumedCheckpoints(long, java.util.Set<java.lang.Long>, java.lang.Runnable, java.util.concurrent.Executor) should be called explicitly here.

Returns:

the subsumed oldest completed checkpoint if possible, return null if no checkpoint needs to be discarded on subsume.

Throws:

Exception

getAllCheckpoints

public List<CompletedCheckpoint> getAllCheckpoints()

Description copied from interface: CompletedCheckpointStore

Returns all CompletedCheckpoint instances.

Returns an empty list if no checkpoint has been added yet.

getNumberOfRetainedCheckpoints

public int getNumberOfRetainedCheckpoints()

Description copied from interface: CompletedCheckpointStore

Returns the current number of retained checkpoints.

getMaxNumberOfRetainedCheckpoints

public int getMaxNumberOfRetainedCheckpoints()

Description copied from interface: CompletedCheckpointStore

Returns the max number of retained checkpoints.

shutdown

public void shutdown(JobStatus jobStatus,
                     CheckpointsCleaner checkpointsCleaner)
              throws Exception

Description copied from interface: CompletedCheckpointStore

Shuts down the store.

The job status is forwarded and used to decide whether state should actually be discarded or kept. SharedStateRegistry.unregisterUnusedState(long) and CheckpointsCleaner.cleanSubsumedCheckpoints(long, java.util.Set<java.lang.Long>, java.lang.Runnable, java.util.concurrent.Executor) should be called here to subsume unused state.

Specified by:

shutdown in interface CompletedCheckpointStore

Overrides:

shutdown in class AbstractCompleteCheckpointStore

Parameters:

jobStatus - Job state on shut down

checkpointsCleaner - that will cleanup completed checkpoints if needed

Throws:

Exception

requiresExternalizedCheckpoints

public boolean requiresExternalizedCheckpoints()

Description copied from interface: CompletedCheckpointStore

This method returns whether the completed checkpoint store requires checkpoints to be externalized. Externalized checkpoints have their meta data persisted, which the checkpoint store can exploit (for example by simply pointing the persisted metadata).

Returns:

True, if the store requires that checkpoints are externalized before being added, false if the store stores the metadata itself.