public static class Checkpoint.CheckpointDir extends Object
SparkPipelineOptions checkpointDir is used as a root directory under which one
directory is created for Spark's checkpoint and another for Beam's Spark runner's fault
checkpointing needs.
Spark's checkpoint relies on Hadoop's FileSystem and is used for
Beam as well rather than FileSystem to be consistent with Spark.| Constructor and Description |
|---|
CheckpointDir(String rootCheckpointDir) |
| Modifier and Type | Method and Description |
|---|---|
org.apache.hadoop.fs.Path |
getBeamCheckpointDir() |
org.apache.hadoop.fs.Path |
getRootCheckpointDir() |
org.apache.hadoop.fs.Path |
getSparkCheckpointDir() |
public CheckpointDir(String rootCheckpointDir)
Copyright © 2016–2017 The Apache Software Foundation. All rights reserved.