Spark RDD Block Removed Before Use - scala

Question

I am using a Future to perform a blocking operation on an RDD like this:
dStreams.foreach(_.foreachRDD { rdd =>
Future{ writeRDD(rdd) }
})
Sometimes I get this error:
org.apache.spark.SparkException: Job aborted due to stage failure: Task creation failed: org.apache.spark.SparkException: Attempted to use BlockRDD[820] at actorStream at Tests.scala:149 after its blocks have been removed!
It seems like Spark is having trouble knowing when this RDD should be deleted.
Why is this happening and what is the solution?
Update:
I think RDDs might be GC'd before they are used. The only working solution so far involves setting
conf.set("spark.streaming.unpersist", "false")
And unpersist()-ing manually.
Full stack trace in case this is a bug:
15/10/12 23:57:23 ERROR org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation: Aborting job.
org.apache.spark.SparkException: Job aborted due to stage failure: Task creation failed: org.apache.spark.SparkException: Attempted to use BlockRDD[765] at actorStream at NxCoreSparkTests.scala:168 after its blocks have been removed!
org.apache.spark.rdd.BlockRDD.assertValid(BlockRDD.scala:83)
org.apache.spark.rdd.BlockRDD.getPreferredLocations(BlockRDD.scala:56)
org.apache.spark.rdd.RDD$$anonfun$preferredLocations$2.apply(RDD.scala:251)
org.apache.spark.rdd.RDD$$anonfun$preferredLocations$2.apply(RDD.scala:251)
scala.Option.getOrElse(Option.scala:120)
org.apache.spark.rdd.RDD.preferredLocations(RDD.scala:250)
org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal(DAGScheduler.scala:1394)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply$mcVI$sp(DAGScheduler.scala:1405)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply(DAGScheduler.scala:1404)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply(DAGScheduler.scala:1404)
scala.collection.immutable.List.foreach(List.scala:318)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2.apply(DAGScheduler.scala:1404)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2.apply(DAGScheduler.scala:1402)
scala.collection.immutable.List.foreach(List.scala:318)
org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal(DAGScheduler.scala:1402)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply$mcVI$sp(DAGScheduler.scala:1405)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply(DAGScheduler.scala:1404)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply(DAGScheduler.scala:1404)
scala.collection.immutable.List.foreach(List.scala:318)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2.apply(DAGScheduler.scala:1404)
org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2.apply(DAGScheduler.scala:1402)
scala.collection.immutable.List.foreach(List.scala:318)
org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal(DAGScheduler.scala:1402)
org.apache.spark.scheduler.DAGScheduler.getPreferredLocs(DAGScheduler.scala:1368)
org.apache.spark.scheduler.DAGScheduler$$anonfun$16.apply(DAGScheduler.scala:829)
org.apache.spark.scheduler.DAGScheduler$$anonfun$16.apply(DAGScheduler.scala:827)
scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244)
scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244)
scala.collection.Iterator$class.foreach(Iterator.scala:727)
scala.collection.AbstractIterator.foreach(Iterator.scala:1157)
scala.collection.IterableLike$class.foreach(IterableLike.scala:72)
scala.collection.AbstractIterable.foreach(Iterable.scala:54)
scala.collection.TraversableLike$class.map(TraversableLike.scala:244)
scala.collection.AbstractTraversable.map(Traversable.scala:105)
org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$submitMissingTasks(DAGScheduler.scala:827)
org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$submitStage(DAGScheduler.scala:772)
org.apache.spark.scheduler.DAGScheduler.handleJobSubmitted(DAGScheduler.scala:757)
org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.doOnReceive(DAGScheduler.scala:1463)
org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:1455)
org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:1444)
org.apache.spark.util.EventLoop$$anon$1.run(EventLoop.scala:48)
at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$failJobAndIndependentStages(DAGScheduler.scala:1280)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$abortStage$1.apply(DAGScheduler.scala:1268)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$abortStage$1.apply(DAGScheduler.scala:1267)
at scala.collection.mutable.ResizableArray$class.foreach(ResizableArray.scala:59)
at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:47)
at org.apache.spark.scheduler.DAGScheduler.abortStage(DAGScheduler.scala:1267)
at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$submitMissingTasks(DAGScheduler.scala:836)
at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$submitStage(DAGScheduler.scala:772)
at org.apache.spark.scheduler.DAGScheduler.handleJobSubmitted(DAGScheduler.scala:757)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.doOnReceive(DAGScheduler.scala:1463)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:1455)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:1444)
at org.apache.spark.util.EventLoop$$anon$1.run(EventLoop.scala:48)
at org.apache.spark.scheduler.DAGScheduler.runJob(DAGScheduler.scala:567)
at org.apache.spark.SparkContext.runJob(SparkContext.scala:1813)
at org.apache.spark.SparkContext.runJob(SparkContext.scala:1826)
at org.apache.spark.SparkContext.runJob(SparkContext.scala:1903)
at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply$mcV$sp(InsertIntoHadoopFsRelation.scala:150)
at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply(InsertIntoHadoopFsRelation.scala:108)
at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation$$anonfun$run$1.apply(InsertIntoHadoopFsRelation.scala:108)
at org.apache.spark.sql.execution.SQLExecution$.withNewExecutionId(SQLExecution.scala:56)
at org.apache.spark.sql.execution.datasources.InsertIntoHadoopFsRelation.run(InsertIntoHadoopFsRelation.scala:108)
at org.apache.spark.sql.execution.ExecutedCommand.sideEffectResult$lzycompute(commands.scala:57)
at org.apache.spark.sql.execution.ExecutedCommand.sideEffectResult(commands.scala:57)
at org.apache.spark.sql.execution.ExecutedCommand.doExecute(commands.scala:69)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$5.apply(SparkPlan.scala:140)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$execute$5.apply(SparkPlan.scala:138)
at org.apache.spark.rdd.RDDOperationScope$.withScope(RDDOperationScope.scala:147)
at org.apache.spark.sql.execution.SparkPlan.execute(SparkPlan.scala:138)
at org.apache.spark.sql.SQLContext$QueryExecution.toRdd$lzycompute(SQLContext.scala:927)
at org.apache.spark.sql.SQLContext$QueryExecution.toRdd(SQLContext.scala:927)
at org.apache.spark.sql.execution.datasources.ResolvedDataSource$.apply(ResolvedDataSource.scala:197)
at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:146)
at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:137)
at org.apache.spark.sql.DataFrameWriter.parquet(DataFrameWriter.scala:304)
at vscan.NxCoreSparkDbUtil$.writeToParquetByDay(NxCoreSparkTapeReader.scala:210)
at vscan.NxCoreSparkGoogleHDFS$$anonfun$6$$anonfun$apply$3$$anonfun$1$$anonfun$apply$mcV$sp$1.apply$mcV$sp(NxCoreSparkTests.scala:190)
at vscan.NxCoreSparkGoogleHDFS$$anonfun$6$$anonfun$apply$3$$anonfun$1$$anonfun$apply$mcV$sp$1.apply(NxCoreSparkTests.scala:188)
at vscan.NxCoreSparkGoogleHDFS$$anonfun$6$$anonfun$apply$3$$anonfun$1$$anonfun$apply$mcV$sp$1.apply(NxCoreSparkTests.scala:188)
at vscan.NxCoreSparkGoogleHDFS$.retry(NxCoreSparkTests.scala:217)
at vscan.NxCoreSparkGoogleHDFS$.retry(NxCoreSparkTests.scala:219)
at vscan.NxCoreSparkGoogleHDFS$.retry(NxCoreSparkTests.scala:219)
at vscan.NxCoreSparkGoogleHDFS$.retry(NxCoreSparkTests.scala:219)
at vscan.NxCoreSparkGoogleHDFS$.retry(NxCoreSparkTests.scala:219)
at vscan.NxCoreSparkGoogleHDFS$$anonfun$6$$anonfun$apply$3$$anonfun$1.apply$mcV$sp(NxCoreSparkTests.scala:188)
at vscan.NxCoreSparkGoogleHDFS$$anonfun$6$$anonfun$apply$3$$anonfun$1.apply(NxCoreSparkTests.scala:185)
at vscan.NxCoreSparkGoogleHDFS$$anonfun$6$$anonfun$apply$3$$anonfun$1.apply(NxCoreSparkTests.scala:185)
at scala.concurrent.impl.Future$PromiseCompletingRunnable.liftedTree1$1(Future.scala:24)
at scala.concurrent.impl.Future$PromiseCompletingRunnable.run(Future.scala:24)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:745)
Caused by: org.apache.spark.SparkException: Attempted to use BlockRDD[765] at actorStream at NxCoreSparkTests.scala:168 after its blocks have been removed!
at org.apache.spark.rdd.BlockRDD.assertValid(BlockRDD.scala:83)
at org.apache.spark.rdd.BlockRDD.getPreferredLocations(BlockRDD.scala:56)
at org.apache.spark.rdd.RDD$$anonfun$preferredLocations$2.apply(RDD.scala:251)
at org.apache.spark.rdd.RDD$$anonfun$preferredLocations$2.apply(RDD.scala:251)
at scala.Option.getOrElse(Option.scala:120)
at org.apache.spark.rdd.RDD.preferredLocations(RDD.scala:250)
at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal(DAGScheduler.scala:1394)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply$mcVI$sp(DAGScheduler.scala:1405)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply(DAGScheduler.scala:1404)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply(DAGScheduler.scala:1404)
at scala.collection.immutable.List.foreach(List.scala:318)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2.apply(DAGScheduler.scala:1404)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2.apply(DAGScheduler.scala:1402)
at scala.collection.immutable.List.foreach(List.scala:318)
at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal(DAGScheduler.scala:1402)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply$mcVI$sp(DAGScheduler.scala:1405)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply(DAGScheduler.scala:1404)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2$$anonfun$apply$2.apply(DAGScheduler.scala:1404)
at scala.collection.immutable.List.foreach(List.scala:318)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2.apply(DAGScheduler.scala:1404)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal$2.apply(DAGScheduler.scala:1402)
at scala.collection.immutable.List.foreach(List.scala:318)
at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$getPreferredLocsInternal(DAGScheduler.scala:1402)
at org.apache.spark.scheduler.DAGScheduler.getPreferredLocs(DAGScheduler.scala:1368)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$16.apply(DAGScheduler.scala:829)
at org.apache.spark.scheduler.DAGScheduler$$anonfun$16.apply(DAGScheduler.scala:827)
at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244)
at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244)
at scala.collection.Iterator$class.foreach(Iterator.scala:727)
at scala.collection.AbstractIterator.foreach(Iterator.scala:1157)
at scala.collection.IterableLike$class.foreach(IterableLike.scala:72)
at scala.collection.AbstractIterable.foreach(Iterable.scala:54)
at scala.collection.TraversableLike$class.map(TraversableLike.scala:244)
at scala.collection.AbstractTraversable.map(Traversable.scala:105)
at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$submitMissingTasks(DAGScheduler.scala:827)
at org.apache.spark.scheduler.DAGScheduler.org$apache$spark$scheduler$DAGScheduler$$submitStage(DAGScheduler.scala:772)
at org.apache.spark.scheduler.DAGScheduler.handleJobSubmitted(DAGScheduler.scala:757)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.doOnReceive(DAGScheduler.scala:1463)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:1455)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:1444)
at org.apache.spark.util.EventLoop$$anon$1.run(EventLoop.scala:48)
15/10/12 23:57:24 ERROR org.apache.spark.sql.execution.datasources.DynamicPartitionWriterContainer: Job job_201510122357_0000 aborted.

Answer 1

I think the problem is that, before your code inside writeRDD(rdd) executes ( since it is in a Future ), the rdd ( or the micro-batch RDD ) is already reclaimed by Apache Spark memory-management or BlockManager.
Therefore, this error
org.apache.spark.SparkException: Job aborted due to stage failure: Task creation failed: org.apache.spark.SparkException: Attempted to use BlockRDD[820] at actorStream at Tests.scala:149 after its blocks have been removed!
You can fix this by first collecting the micro-batch collection and then passing it to writeRDD() function. Something like this:
dStreams.foreach(_.foreachRDD { rdd =>
val coll = rdd.collect()
Future{ writeCollection(coll) }
})