只能复制到 0 个节点而不是 minReplication (=1).有 4 个数据节点正在运行，此操作中未排除任何节点 [英] could only be replicated to 0 nodes instead of minReplication (=1). There are 4 datanode(s) running and no node(s) are excluded in this operation

问题描述

我不知道如何解决这个错误:

I don't know how to fix this error:

<代码>顶点失败，vertexName = initialmap，vertexId = vertex_1449805139484_0001_1_00，诊断= [任务失败，的TaskID = task_1449805139484_0001_1_00_000003，诊断= [AttemptID:attempt_1449805139484_0001_1_00_000003_0信息:错误:org.apache.hadoop.ipc.RemoteException(java.io.IOException): 文件/user/hadoop/gridmix-kon/input/_temporary/1/_temporary/attempt_14498051394840_0001_m_000003_0/part-m-00003/segment-121 只能复制到 0 个节点(而不是 minReplication=1).有 4 个数据节点正在运行，此操作中没有排除任何节点.在 org.apache.hadoop.hdfs.server.blockmanagement.BlockManager.chooseTarget(BlockManager.java:1441)在 org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getAdditionalBlock(FSNamesystem.java:2702)在 org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer.addBlock(NameNodeRpcServer.java:584)在 org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolServerSideTranslatorPB.addBlock(ClientNamenodeProtocolServerSideTranslatorPB.java:440)在 org.apache.hadoop.hdfs.protocol.proto.ClientNamenodeProtocolProtos$ClientNamenodeProtocol$2.callBlockingMethod(ClientNamenodeProtocolProtos.java)在 org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:585)在 org.apache.hadoop.ipc.RPC$Server.call(RPC.java:928)在 org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2014)在 org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2010)在 java.security.AccessController.doPrivileged(Native Method)在 javax.security.auth.Subject.doAs(Subject.java:415)在 org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1561)在 org.apache.hadoop.ipc.Server$Handler.run(Server.java:2008)在 org.apache.hadoop.ipc.Client.call(Client.java:1411)在 org.apache.hadoop.ipc.Client.call(Client.java:1364)在 org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:206)在 com.sun.proxy.$Proxy17.addBlock(来源不明)在 sun.reflect.GeneratedMethodAccessor3.invoke(未知来源)在 sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)在 java.lang.reflect.Method.invoke(Method.java:606)在 org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:190)在 org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:103)在 com.sun.proxy.$Proxy17.addBlock(来源不明)在 org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolTranslatorPB.addBlock(ClientNamenodeProtocolTranslatorPB.java:361)在 org.apache.hadoop.hdfs.DFSOutputStream$DataStreamer.locateFollowingBlock(DFSOutputStream.java:1439)在 org.apache.hadoop.hdfs.DFSOutputStream$DataStreamer.nextBlockOutputStream(DFSOutputStream.java:1261)在 org.apache.hadoop.hdfs.DFSOutputStream$DataStreamer.run(DFSOutputStream.java:525)

知道这是怎么回事吗?

推荐答案

这个错误出现在BlockManager::chooseTarget4NewBlock()(我指的是最新的代码)代码中.导致这种情况的特定代码段是:

This error occurs in BlockManager::chooseTarget4NewBlock() (I am referring to the latest code) code. Specific piece of code, which causes this is:

final DatanodeStorageInfo[] targets = blockplacement.chooseTarget(src,
    numOfReplicas, client, excludedNodes, blocksize, 
    favoredDatanodeDescriptors, storagePolicy);

if (targets.length < minReplication) {
  throw new IOException("File " + src + " could only be replicated to "
      + targets.length + " nodes instead of minReplication (="
      + minReplication + ").  There are "
      + getDatanodeManager().getNetworkTopology().getNumOfLeaves()
      + " datanode(s) running and "
      + (excludedNodes == null? "no": excludedNodes.size())
      + " node(s) are excluded in this operation.");
}

这种情况发生在BlockManager试图选择一个目标主机来存储新的数据块但找不到单个主机时(targets.length .minReplication在hdfs-site.xml文件中设置为1(配置参数:dfs.namenode.replication.min).

This occurs, when the BlockManager tries to choose a target host for storing new block of data and can not find a single host (targets.length < minReplication). minReplication is set to 1 (configuration parameter: dfs.namenode.replication.min) in hdfs-site.xml file.

这可能是由于以下原因之一造成的:

This could occur due to one of the following reasons:

• 数据节点实例未运行
• 数据节点实例无法联系名称节点
• 数据节点空间不足，因此无法为它们分配新的数据块

但是，就您而言，错误消息还包含以下信息:

But, in your case, error message also contains following information:

There are 4 datanode(s) running and no node(s) are excluded in this operation.

这意味着，有 4 个数据节点正在运行，并且在此操作中，所有 4 个数据节点都被考虑用于放置数据.

It means, there are 4 Data Nodes running and all the 4 Data Nodes were considered for placement of data, for this operation.

因此，可能怀疑是数据节点上的磁盘空间.您可以使用以下命令检查数据节点上的磁盘空间:

So, possible suspect is disk space on the Data Nodes. You can check the disk space on your Data Nodes, using the following command:

hdfs dfsadmin -report

它为您的每个实时数据节点提供报告.例如就我而言，我得到以下信息:

It gives report for each of your Live Data Nodes. For e.g. in my case, I got the following:

Live datanodes (1):

Name: 192.168.56.1:50010 (192.168.56.1)
Hostname: 192.168.56.1
Decommission Status : Normal
Configured Capacity: 648690003968 (604.14 GB)
DFS Used: 193849055737 (180.54 GB)
Non DFS Used: 186164975111 (173.38 GB)
DFS Remaining: 268675973120 (250.22 GB)
DFS Used%: 29.88%
DFS Remaining%: 41.42%
Configured Cache Capacity: 0 (0 B)
Cache Used: 0 (0 B)
Cache Remaining: 0 (0 B)
Cache Used%: 100.00%
Cache Remaining%: 0.00%
Xceivers: 1
Last contact: Sun Dec 13 17:17:34 IST 2015

检查DFS-Remaining"和DFS-Remaining%".这应该让您了解数据节点上的剩余空间.

Check the "DFS-Remaining" and "DFS-Remaining%". That should give you an idea about the remaining space on your Data Nodes.

您也可以在此处参考维基:https://wiki.apache.org/hadoop/CouldOnlyBeReplicatedTo，描述了此错误的原因和缓解方法.

You can also refer to the wiki here: https://wiki.apache.org/hadoop/CouldOnlyBeReplicatedTo, which describes the reasons for this error and ways to mitigate it.

这篇关于只能复制到 0 个节点而不是 minReplication (=1).有 4 个数据节点正在运行，此操作中未排除任何节点的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！