createAndRestoreExecutionGraph
ExecutionGraph createAndRestoreExecutionGraph(JobGraph jobGraph,
CompletedCheckpointStore completedCheckpointStore,
CheckpointsCleaner checkpointsCleaner,
CheckpointIDCounter checkpointIdCounter,
CheckpointStatsTracker checkpointStatsTracker,
TaskDeploymentDescriptorFactory.PartitionLocationConstraint partitionLocationConstraint,
long initializationTimestamp,
VertexAttemptNumberStore vertexAttemptNumberStore,
VertexParallelismStore vertexParallelismStore,
ExecutionStateUpdateListener executionStateUpdateListener,
MarkPartitionFinishedStrategy markPartitionFinishedStrategy,
ExecutionPlanSchedulingContext executionPlanSchedulingContext,
org.slf4j.Logger log)
throws Exception
- Parameters:
jobGraph - jobGraph to initialize the ExecutionGraph with
completedCheckpointStore - completedCheckpointStore to pass to the CheckpointCoordinator
checkpointsCleaner - checkpointsCleaner to pass to the CheckpointCoordinator
checkpointIdCounter - checkpointIdCounter to pass to the CheckpointCoordinator
checkpointStatsTracker - The CheckpointStatsTracker that's used for collecting
the checkpoint-related statistics.
partitionLocationConstraint - partitionLocationConstraint for this job
initializationTimestamp - initializationTimestamp when the ExecutionGraph was created
vertexAttemptNumberStore - vertexAttemptNumberStore keeping information about the vertex
attempts of previous runs
vertexParallelismStore - vertexMaxParallelismStore keeping information about the vertex
max parallelism settings
executionStateUpdateListener - listener for state transitions of the individual
executions
log - log to use for logging
executionPlanSchedulingContext - execution plan scheduling context that retrieve
execution context details for adaptive batch jobs
- Returns:
- restored
ExecutionGraph
- Throws:
Exception - if the ExecutionGraph could not be created and restored