Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[SPARK-11051][Core] Do not allow local checkpointing after the RDD is materialized and checkpointed #9072

Closed
Closed
Show file tree
Hide file tree
Changes from 4 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
Expand Up @@ -547,6 +547,11 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
*/
def isCheckpointed: Boolean = rdd.isCheckpointed

/**
* Return whether this RDD has been checkpointed and materialized or not
*/
private[spark] def isCheckpointedAndMaterialized: Boolean = rdd.isCheckpointedAndMaterialized
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

why add this? I don't think this is used except in tests


/**
* Gets the name of the file to which this RDD was checkpointed
*/
Expand Down
32 changes: 26 additions & 6 deletions core/src/main/scala/org/apache/spark/rdd/RDD.scala
Original file line number Diff line number Diff line change
Expand Up @@ -294,7 +294,11 @@ abstract class RDD[T: ClassTag](
*/
private[spark] def computeOrReadCheckpoint(split: Partition, context: TaskContext): Iterator[T] =
{
if (isCheckpointed) firstParent[T].iterator(split, context) else compute(split, context)
if (isCheckpointedAndMaterialized) {
firstParent[T].iterator(split, context)
} else {
compute(split, context)
}
}

/**
Expand Down Expand Up @@ -1520,12 +1524,22 @@ abstract class RDD[T: ClassTag](
persist(LocalRDDCheckpointData.transformStorageLevel(storageLevel), allowOverride = true)
}

checkpointData match {
case Some(reliable: ReliableRDDCheckpointData[_]) => logWarning(
"RDD was already marked for reliable checkpointing: overriding with local checkpoint.")
case _ =>
// If this RDD is already checkpointed and materialized, its lineage is already truncated.
// We must not override our `checkpointData` in this case because it is needed to recover
// the checkpointed data. If it is overridden, next time materializing on this RDD will
// cause error.
if (isCheckpointedAndMaterialized) {
logWarning("Not marking RDD for local checkpoint because it was already " +
"checkpointed and materialized")
} else {
// Lineage is not truncated yet, so just override any existing checkpoint data with ours
checkpointData match {
case Some(_: ReliableRDDCheckpointData[_]) => logWarning(
"RDD was already marked for reliable checkpointing: overriding with local checkpoint.")
case _ =>
}
checkpointData = Some(new LocalRDDCheckpointData(this))
}
checkpointData = Some(new LocalRDDCheckpointData(this))
this
}

Expand All @@ -1534,6 +1548,12 @@ abstract class RDD[T: ClassTag](
*/
def isCheckpointed: Boolean = checkpointData.exists(_.isCheckpointed)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can you also update the java doc for this:

Whether this RDD is checkpointed and materialized, either reliably or locally.


/**
* Return whether this RDD is marked for checkpointing and materialized,
* either reliably or locally.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I would just say

Whether this RDD is checkpointed and materialized, either reliably or locally.
This is introduced as an alias for `isCheckpointed` to clarify the semantics of
the return value. Exposed for testing.

*/
private[spark] def isCheckpointedAndMaterialized: Boolean = isCheckpointed

/**
* Return whether this RDD is marked for local checkpointing.
* Exposed for testing.
Expand Down
8 changes: 8 additions & 0 deletions core/src/test/java/org/apache/spark/JavaAPISuite.java
Original file line number Diff line number Diff line change
Expand Up @@ -1420,9 +1420,13 @@ public void checkpointAndComputation() {
JavaRDD<Integer> rdd = sc.parallelize(Arrays.asList(1, 2, 3, 4, 5));
sc.setCheckpointDir(tempDir.getAbsolutePath());
Assert.assertFalse(rdd.isCheckpointed());
Assert.assertFalse(rdd.isCheckpointedAndMaterialized());
rdd.checkpoint();
Assert.assertFalse(rdd.isCheckpointed());
Assert.assertFalse(rdd.isCheckpointedAndMaterialized());
rdd.count(); // Forces the DAG to cause a checkpoint
Assert.assertTrue(rdd.isCheckpointed());
Assert.assertTrue(rdd.isCheckpointedAndMaterialized());
Assert.assertEquals(Arrays.asList(1, 2, 3, 4, 5), rdd.collect());
}

Expand All @@ -1431,9 +1435,13 @@ public void checkpointAndRestore() {
JavaRDD<Integer> rdd = sc.parallelize(Arrays.asList(1, 2, 3, 4, 5));
sc.setCheckpointDir(tempDir.getAbsolutePath());
Assert.assertFalse(rdd.isCheckpointed());
Assert.assertFalse(rdd.isCheckpointedAndMaterialized());
rdd.checkpoint();
Assert.assertFalse(rdd.isCheckpointed());
Assert.assertFalse(rdd.isCheckpointedAndMaterialized());
rdd.count(); // Forces the DAG to cause a checkpoint
Assert.assertTrue(rdd.isCheckpointed());
Assert.assertTrue(rdd.isCheckpointedAndMaterialized());

Assert.assertTrue(rdd.getCheckpointFile().isPresent());
JavaRDD<Integer> recovered = sc.checkpointFile(rdd.getCheckpointFile().get());
Expand Down
4 changes: 4 additions & 0 deletions core/src/test/scala/org/apache/spark/CheckpointSuite.scala
Original file line number Diff line number Diff line change
Expand Up @@ -241,9 +241,13 @@ class CheckpointSuite extends SparkFunSuite with LocalSparkContext with Logging
val rdd = new BlockRDD[Int](sc, Array[BlockId]())
assert(rdd.partitions.size === 0)
assert(rdd.isCheckpointed === false)
assert(rdd.isCheckpointedAndMaterialized === false)
checkpoint(rdd, reliableCheckpoint)
assert(rdd.isCheckpointed === false)
assert(rdd.isCheckpointedAndMaterialized === false)
assert(rdd.count() === 0)
assert(rdd.isCheckpointed === true)
assert(rdd.isCheckpointedAndMaterialized === true)
assert(rdd.partitions.size === 0)
}

Expand Down