When would ShuffleBlockFetcherIterator throw "Failed to get block(s)" exceptions? - apache-spark

Question

In my spark application which is run in a cluster mode, I get below exception. I know somehow this coud be due to emery issue. But as the error says, it can not connect to a node. But I ma sure the node is available and it can be connected. Can anyone know what is the main cause of this error and how to resolve it?
17/10/31 17:10:54 ERROR ShuffleBlockFetcherIterator: Failed to get block(s) from AUPER01-02-10-12-0.prod.vroc.com.au:36787
java.io.IOException: Failed to connect to AUPER01-02-10-12-0.prod.vroc.com.au/192.168.11.22:36787
at org.apache.spark.network.client.TransportClientFactory.createClient(TransportClientFactory.java:232)
at org.apache.spark.network.client.TransportClientFactory.createClient(TransportClientFactory.java:182)
at org.apache.spark.network.netty.NettyBlockTransferService$$anon$1.createAndStart(NettyBlockTransferService.scala:97)
at org.apache.spark.network.shuffle.RetryingBlockFetcher.fetchAllOutstanding(RetryingBlockFetcher.java:141)
at org.apache.spark.network.shuffle.RetryingBlockFetcher.access$200(RetryingBlockFetcher.java:43)
at org.apache.spark.network.shuffle.RetryingBlockFetcher$1.run(RetryingBlockFetcher.java:171)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at io.netty.util.concurrent.DefaultThreadFactory$DefaultRunnableDecorator.run(DefaultThreadFactory.java:144)
at java.lang.Thread.run(Thread.java:745)
Caused by: io.netty.channel.AbstractChannel$AnnotatedConnectException: Connection refused: AUPER01-02-10-12-0.prod.vroc.com.au/192.168.11.22:36787
at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method)
at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:717)
at io.netty.channel.socket.nio.NioSocketChannel.doFinishConnect(NioSocketChannel.java:257)
at io.netty.channel.nio.AbstractNioChannel$AbstractNioUnsafe.finishConnect(AbstractNioChannel.java:291)
at io.netty.channel.nio.NioEventLoop.processSelectedKey(NioEventLoop.java:631)
at io.netty.channel.nio.NioEventLoop.processSelectedKeysOptimized(NioEventLoop.java:566)
at io.netty.channel.nio.NioEventLoop.processSelectedKeys(NioEventLoop.java:480)
at io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:442)
at io.netty.util.concurrent.SingleThreadEventExecutor$2.run(SingleThreadEventExecutor.java:131)
... 2 more

Answer 1

It appears that one of the executors died while the other executors tried to pull blocks from earlier shuffle stages to complete a Spark job.
Right after you've spark-submited a Spark application to a cluster, the application gets a set of machines for executors. They are responsible for executing tasks and caching their results (in memory and/or disk).
Every executor has its own BlockManager that is responsible for managing datasets (as blocks).
The BlockManagers in a Spark application have all to be available or the Spark application will re-trigger task execution.
ShuffleBlockFetcherIterator is a Scala Iterator that fetches multiple shuffle blocks (aka shuffle map outputs) from local and remote BlockManagers.