spark集群中slave访问master被拒绝

Question

16/04/09 23:18:20 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
16/04/09 23:18:45 ERROR RetryingBlockFetcher: Exception while beginning fetch of 1 outstanding blocks 
java.io.IOException: Failed to connect to master/218.192.172.48:35542
at org.apache.spark.network.client.TransportClientFactory.createClient(TransportClientFactory.java:216)
at org.apache.spark.network.client.TransportClientFactory.createClient(TransportClientFactory.java:167)
at org.apache.spark.network.netty.NettyBlockTransferService$$anon$1.createAndStart(NettyBlockTransferService.scala:91)
at org.apache.spark.network.shuffle.RetryingBlockFetcher.fetchAllOutstanding(RetryingBlockFetcher.java:140)
at org.apache.spark.network.shuffle.RetryingBlockFetcher.start(RetryingBlockFetcher.java:120)
at org.apache.spark.network.netty.NettyBlockTransferService.fetchBlocks(NettyBlockTransferService.scala:100)
at org.apache.spark.storage.ShuffleBlockFetcherIterator.sendRequest(ShuffleBlockFetcherIterator.scala:169)
at org.apache.spark.storage.ShuffleBlockFetcherIterator.fetchUpToMaxBytes(ShuffleBlockFetcherIterator.scala:351)
at org.apache.spark.storage.ShuffleBlockFetcherIterator.initialize(ShuffleBlockFetcherIterator.scala:286)
at org.apache.spark.storage.ShuffleBlockFetcherIterator.<init>(ShuffleBlockFetcherIterator.scala:119)
at org.apache.spark.shuffle.BlockStoreShuffleReader.read(BlockStoreShuffleReader.scala:43)
at org.apache.spark.rdd.ShuffledRDD.compute(ShuffledRDD.scala:98)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:313)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:277)
at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:313)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:277)
at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:313)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:277)
at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:313)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:277)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:69)
at org.apache.spark.scheduler.Task.run(Task.scala:82)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:227)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:745)
Caused by: java.net.ConnectException: 拒绝连接: master/218.192.172.48:35542
at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method)
at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:744)
at io.netty.channel.socket.nio.NioSocketChannel.doFinishConnect(NioSocketChannel.java:224)
at io.netty.channel.nio.AbstractNioChannel$AbstractNioUnsafe.finishConnect(AbstractNioChannel.java:289)
at io.netty.channel.nio.NioEventLoop.processSelectedKey(NioEventLoop.java:528)
at io.netty.channel.nio.NioEventLoop.processSelectedKeysOptimized(NioEventLoop.java:468)
at io.netty.channel.nio.NioEventLoop.processSelectedKeys(NioEventLoop.java:382)
at io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:354)
at io.netty.util.concurrent.SingleThreadEventExecutor$2.run(SingleThreadEventExecutor.java:111)
... 1 more
16/04/09 23:18:50 ERROR RetryingBlockFetcher: Exception while beginning fetch of 1 outstanding blocks (after 1 retries)
java.io.IOException: Failed to connect to master/218.192.172.48:35542
at org.apache.spark.network.client.TransportClientFactory.createClient(TransportClientFactory.java:216)
at org.apache.spark.network.client.TransportClientFactory.createClient(TransportClientFactory.java:167)
at org.apache.spark.network.netty.NettyBlockTransferService$$anon$1.createAndStart(NettyBlockTransferService.scala:91)
at org.apache.spark.network.shuffle.RetryingBlockFetcher.fetchAllOutstanding(RetryingBlockFetcher.java:140)
at org.apache.spark.network.shuffle.RetryingBlockFetcher.access$200(RetryingBlockFetcher.java:43)
at org.apache.spark.network.shuffle.RetryingBlockFetcher$1.run(RetryingBlockFetcher.java:170)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
at java.util.concurrent.FutureTask.run(FutureTask.java:262)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:745)
Caused by: java.net.ConnectException: 拒绝连接: master/218.192.172.48:35542
at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method)
at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:744)
at io.netty.channel.socket.nio.NioSocketChannel.doFinishConnect(NioSocketChannel.java:224)
at io.netty.channel.nio.AbstractNioChannel$AbstractNioUnsafe.finishConnect(AbstractNioChannel.java:289)
at io.netty.channel.nio.NioEventLoop.processSelectedKey(NioEventLoop.java:528)
at io.netty.channel.nio.NioEventLoop.processSelectedKeysOptimized(NioEventLoop.java:468)
at io.netty.channel.nio.NioEventLoop.processSelectedKeys(NioEventLoop.java:382)
at io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:354)
at io.netty.util.concurrent.SingleThreadEventExecutor$2.run(SingleThreadEventExecutor.java:111)
... 1 more
16/04/09 23:18:55 ERROR RetryingBlockFetcher: Exception while beginning fetch of 1 outstanding blocks (after 2 retries)
java.io.IOException: Failed to connect to master/218.192.172.48:35542
at org.apache.spark.network.client.TransportClientFactory.createClient(TransportClientFactory.java:216)
at org.apache.spark.network.client.TransportClientFactory.createClient(TransportClientFactory.java:167)
at org.apache.spark.network.netty.NettyBlockTransferService$$anon$1.createAndStart(NettyBlockTransferService.scala:91)
at org.apache.spark.network.shuffle.RetryingBlockFetcher.fetchAllOutstanding(RetryingBlockFetcher.java:140)
at org.apache.spark.network.shuffle.RetryingBlockFetcher.access$200(RetryingBlockFetcher.java:43)
at org.apache.spark.network.shuffle.RetryingBlockFetcher$1.run(RetryingBlockFetcher.java:170)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
at java.util.concurrent.FutureTask.run(FutureTask.java:262)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:745)配置是按照教程：http://wuchong.me/blog/2015/04/04/spark-on-yarn-cluster-deploy/ 来的，但是始终解决不了这个问题，
主从互拼也能拼通，请问各位前辈有遇到过类似情形么，我应该如何排错？
其中我把master同时也作为worder，但是master的日志里我也发现了同样的错误.
另外，SparkPi的例程却可以运行，而我现在运行的程序在local模式下是没有问题的。

Answer 1

我为master映射了多个ip，然而我发现原来拒绝访问的ip是外网的。
这可能是根据hostname来解析的ip，因为在服务器上我不能随便改它

Answer 2

首先再强调一下环境，我用的是Intellij的IDE，并且已经按照提示在conf/spark_env.sh中设置了SPARK_LOCAL_IP对应着局域网下本机的ip，可是在IDE的环境下仍然提示映射到了外网ip中，从机我设置在了虚拟机上，拒绝访问master也是正常的。
然而我发现，在spark-shell的环境下，并没有提示解析到外网ip上，所以我尝试了下用spark-submit提交，拒绝访问的问题就消失了。
当然，如果一开始master和对应的hostname是一样的，应该就不会存在拒绝访问的问题，看起来默认就是按照hostname来解析，只是intellij没有读取spark-env.sh中的设置，这是个需要研究的问题。