Timeout Exception in Apache-Spark during program E

问题:

I am running a Bash Script in MAC. This script calls a spark method written in Scala language for a large number of times. I am currently trying to call this spark method for 100,000 times using a for loop.

The code exits with the following exception after running a small number of iterations, around 3000 iterations.

org.apache.spark.rpc.RpcTimeoutException: Futures timed out after [10 seconds]. This timeout is controlled by spark.executor.heartbeatInterval
    at org.apache.spark.rpc.RpcTimeout.org$apache$spark$rpc$RpcTimeout$$createRpcTimeoutException(RpcTimeout.scala:48)
    at org.apache.spark.rpc.RpcTimeout$$anonfun$addMessageIfTimeout$1.applyOrElse(RpcTimeout.scala:63)
    at org.apache.spark.rpc.RpcTimeout$$anonfun$addMessageIfTimeout$1.applyOrElse(RpcTimeout.scala:59)
    at scala.PartialFunction$OrElse.apply(PartialFunction.scala:167)
    at org.apache.spark.rpc.RpcTimeout.awaitResult(RpcTimeout.scala:83)
    at org.apache.spark.rpc.RpcEndpointRef.askWithRetry(RpcEndpointRef.scala:102)
    at org.apache.spark.executor.Executor.org$apache$spark$executor$Executor$$reportHeartBeat(Executor.scala:518)
    at org.apache.spark.executor.Executor$$anon$1$$anonfun$run$1.apply$mcV$sp(Executor.scala:547)
    at org.apache.spark.executor.Executor$$anon$1$$anonfun$run$1.apply(Executor.scala:547)
    at org.apache.spark.executor.Executor$$anon$1$$anonfun$run$1.apply(Executor.scala:547)
    at org.apache.spark.util.Utils$.logUncaughtExceptions(Utils.scala:1877)
    at org.apache.spark.executor.Executor$$anon$1.run(Executor.scala:547)
    at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
    at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:304)

Exception in thread "dag-scheduler-event-loop" 16/11/22 13:37:32 WARN NioEventLoop: Unexpected exception in the selector loop.
java.lang.OutOfMemoryError: Java heap space
    at io.netty.util.internal.MpscLinkedQueue.offer(MpscLinkedQueue.java:126)
    at io.netty.util.internal.MpscLinkedQueue.add(MpscLinkedQueue.java:221)
    at io.netty.util.concurrent.SingleThreadEventExecutor.fetchFromScheduledTaskQueue(SingleThreadEventExecutor.java:259)
    at io.netty.util.concurrent.SingleThreadEventExecutor.runAllTasks(SingleThreadEventExecutor.java:346)
    at io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:357)
    at io.netty.util.concurrent.SingleThreadEventExecutor$2.run(SingleThreadEventExecutor.java:111)
    at java.lang.Thread.run(Thread.java:745)
java.lang.OutOfMemoryError: Java heap space
    at java.util.regex.Pattern.compile(Pattern.java:1047)
    at java.lang.String.replace(String.java:2180)
    at org.apache.spark.util.Utils$.getFormattedClassName(Utils.scala:1728)
    at org.apache.spark.storage.RDDInfo$$anonfun$1.apply(RDDInfo.scala:57)
    at org.apache.spark.storage.RDDInfo$$anonfun$1.apply(RDDInfo.scala:57)
    at scala.Option.getOrElse(Option.scala:121)
    at org.apache.spark.storage.RDDInfo$.fromRdd(RDDInfo.scala:57)
    at org.apache.spark.scheduler.StageInfo$$anonfun$1.apply(StageInfo.scala:87)

Can someone help please, is this error being caused because of a large number of calls to spark method?

回答1:

Its RpcTimeoutException .. so spark.network.timeout (spark.rpc.askTimeout) could be tuned with larger-than-default values in order to handle complex workload. You can start with these values and adjust accordingly to your workloads. Please see latest

spark.network.timeout 120s Default timeout for all network interactions. This config will be used in place of spark.core.connection.ack.wait.timeout, spark.storage.blockManagerSlaveTimeoutMs, spark.shuffle.io.connectionTimeout, spark.rpc.askTimeout or spark.rpc.lookupTimeout if they are not configured.

Also consider increasing executor memory i.e spark.executor.memory and most imp thing is review your code, to check whether that is candidate for further optimization.

Solution : value 600 is based on requirement

set by SparkConf: conf.set("spark.network.timeout", "600s")
set by spark-defaults.conf: spark.network.timeout 600s
set when calling spark-submit: --conf spark.network.timeout=600s

回答2:

The above stack trace is also shown java heap space its OOM error so once try to increase the memory and run it and regarding timeout its rpc timeout so you can set spark.network.timeout with timeout value according to your need...

回答3:

pls increase the executer memory so that OOM will go away else make chnage in code so that your RDD wont have big memory foot print.

--executer-memory = 3G

回答4:

Just increase the spark.executor.heartbeatInterval to 20s, the error says that.

回答5:

You are seeing this issue due to the executor memory. Try increasing the memory to (x 2) so the containers don't time out while waiting on the remaining containers.