共计 7118 个字符，预计需要花费 18 分钟才能阅读完成。
丸趣 TV 小编给大家分享一下 Yarn 上的不健康节点 UNHEALTHY nodes 怎么处理，希望大家阅读完这篇文章之后都有所收获，下面让我们一起去探讨吧！
一、错误
自己的三台虚拟机 hadoop001、hadoop002、hadoop003
检查 23188 发现有 Unhealthy Nodes，正常的 active nodes 数目不对
另外查看
$ yarn node -list -all
Total Nodes:4
Node-Id Node-State Node-Http-Address Number-of-Running-Containers
hadoop001:34354 UNHEALTHY hadoop001:23999 0
hadoop002:60027 RUNNING hadoop002:23999 0
hadoop001:50623 UNHEALTHY hadoop001:23999 0
hadoop003:39700 UNHEALTHY hadoop003:23999 0二、日志检查
查看 resourcemanager 的日志可以看到
2016-09-10 12:02:05,953 INFO org.apache.hadoop.yarn.server.resourcemanager.scheduler.fair.FairScheduler: Added node hadoop002:60027 cluster capacity: memory:4096, vCores:4
2016-09-10 12:02:05,990 INFO org.apache.hadoop.yarn.server.resourcemanager.rmnode.RMNodeImpl: Node hadoop001:50623 reported UNHEALTHY with details: 1/1 local-dirs are bad: /data/disk1/data/yarn/local; 1/1 log-dirs are bad: /opt/beh/logs
/yarn/userlog
2016-09-10 12:02:05,991 INFO org.apache.hadoop.yarn.server.resourcemanager.rmnode.RMNodeImpl: hadoop001:50623 Node Transitioned from RUNNING to UNHEALTHY
2016-09-10 12:02:05,993 INFO org.apache.hadoop.yarn.server.resourcemanager.scheduler.fair.FairScheduler: Removed node hadoop001:50623 cluster capacity: memory:2048, vCores:2
2016-09-10 12:02:06,378 INFO org.apache.hadoop.yarn.util.RackResolver: Resolved hadoop003 to /default-rack检查 nodemanager 的日志可以查看到
2016-09-10 12:02:02,869 INFO org.mortbay.log: jetty-6.1.26.cloudera.4
2016-09-10 12:02:02,905 INFO org.mortbay.log: Extract jar:file:/opt/beh/core/hadoop/share/hadoop/yarn/hadoop-yarn-common-2.6.0-cdh6.4.4.jar!/webapps/node to /tmp/Jetty_0_0_0_0_23999_node____tgfx6h/webapp
2016-09-10 12:02:03,242 INFO org.mortbay.log: Started HttpServer2$SelectChannelConnectorWithSafeStartup@0.0.0.0:23999
2016-09-10 12:02:03,242 INFO org.apache.hadoop.yarn.webapp.WebApps: Web app /node started at 23999
2016-09-10 12:02:03,735 INFO org.apache.hadoop.yarn.webapp.WebApps: Registered webapp guice modules
2016-09-10 12:02:03,775 INFO org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl: Sending out 0 NM container statuses: []
2016-09-10 12:02:03,783 INFO org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl: Registering with RM using containers :[]
2016-09-10 12:02:03,822 INFO org.apache.hadoop.yarn.client.ConfiguredRMFailoverProxyProvider: Failing over to rm2
2016-09-10 12:02:03,824 INFO org.apache.hadoop.io.retry.RetryInvocationHandler: Exception while invoking registerNodeManager of class ResourceTrackerPBClientImpl over rm2 after 1 fail over attempts. Trying to fail over after sleeping fo
r 2138ms.
java.net.ConnectException: Call From hadoop002/192.168.30.22 to hadoop002:23125 failed on connection exception: java.net.ConnectException: 拒绝连接; For more details see: http://wiki.apache.org/hadoop/ConnectionRefused
at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:57)
at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
at java.lang.reflect.Constructor.newInstance(Constructor.java:526)
at org.apache.hadoop.net.NetUtils.wrapWithMessage(NetUtils.java:791)
at org.apache.hadoop.net.NetUtils.wrapException(NetUtils.java:731)
at org.apache.hadoop.ipc.Client.call(Client.java:1472)
at org.apache.hadoop.ipc.Client.call(Client.java:1399)
at org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:232)
at com.sun.proxy.$Proxy27.registerNodeManager(Unknown Source)
at org.apache.hadoop.yarn.server.api.impl.pb.client.ResourceTrackerPBClientImpl.registerNodeManager(ResourceTrackerPBClientImpl.java:68)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:606)
at org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:187)
at org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:102)
at com.sun.proxy.$Proxy28.registerNodeManager(Unknown Source)
at org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl.registerWithRM(NodeStatusUpdaterImpl.java:257)
at org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl.serviceStart(NodeStatusUpdaterImpl.java:191)
at org.apache.hadoop.service.AbstractService.start(AbstractService.java:193)
at org.apache.hadoop.service.CompositeService.serviceStart(CompositeService.java:120)
at org.apache.hadoop.yarn.server.nodemanager.NodeManager.serviceStart(NodeManager.java:264)
at org.apache.hadoop.service.AbstractService.start(AbstractService.java:193)
at org.apache.hadoop.yarn.server.nodemanager.NodeManager.initAndStartNodeManager(NodeManager.java:463)
at org.apache.hadoop.yarn.server.nodemanager.NodeManager.main(NodeManager.java:509)
Caused by: java.net.ConnectException: 拒绝连接
at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method)
at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:739)
at org.apache.hadoop.net.SocketIOWithTimeout.connect(SocketIOWithTimeout.java:206)
at org.apache.hadoop.net.NetUtils.connect(NetUtils.java:530)
at org.apache.hadoop.net.NetUtils.connect(NetUtils.java:494)
at org.apache.hadoop.ipc.Client$Connection.setupConnection(Client.java:607)
at org.apache.hadoop.ipc.Client$Connection.setupIOstreams(Client.java:705)
at org.apache.hadoop.ipc.Client$Connection.access$2800(Client.java:368)
at org.apache.hadoop.ipc.Client.getConnection(Client.java:1521)
at org.apache.hadoop.ipc.Client.call(Client.java:1438)
... 19 more
2016-09-10 12:02:05,965 INFO org.apache.hadoop.yarn.client.ConfiguredRMFailoverProxyProvider: Failing over to rm1
2016-09-10 12:02:05,996 INFO org.apache.hadoop.yarn.server.nodemanager.security.NMContainerTokenSecretManager: Rolling master-key for container-tokens, got key with id -1513537506
2016-09-10 12:02:05,998 INFO org.apache.hadoop.yarn.server.nodemanager.security.NMTokenSecretManagerInNM: Rolling master-key for container-tokens, got key with id 701920721
2016-09-10 12:02:05,999 INFO org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl: Registered with ResourceManager as hadoop002:60027 with total resource of memory:2048, vCores:2
2016-09-10 12:02:05,999 INFO org.apache.hadoop.yarn.server.nodemanager.NodeStatusUpdaterImpl: Notifying ContainerManager to unblock new container-requests三、错误分析
NodeManager 默认会每两分钟检查本地磁盘（local-dirs），找出那些目录可以使用。注意如果判定这个磁盘不可用，则在重启 NodeManager 之前，就算磁盘好了，也不会把它变成可用。当好磁盘数少于一定量时，会把这台机器变成 unhealthy，将不会再给这台机器分配任务。
查看自己的虚拟机磁盘情况，发现 001 和 003 的磁盘都要满了，于是清除不需要的文件，腾出剩余空间，UNHEALTHY nodes 立马恢复正常
$ yarn node -list -all
Total Nodes:4
Node-Id Node-State Node-Http-Address Number-of-Running-Containers
hadoop001:34354 RUNNING hadoop001:23999 0
hadoop002:60027 RUNNING hadoop002:23999 0
hadoop003:39700 RUNNING hadoop003:23999 0
hadoop001:50623 LOST hadoop001:23999 0此处为什么有 2 个 hadoop001，因为修改了配置文件重启过一次，所有出现了 2 个，其中有一个为 LOST 状态，另一个正常 RUNNING，不影响使用，yarn 重启后就可恢复正常。
看完了这篇文章，相信你对“Yarn 上的不健康节点 UNHEALTHY nodes 怎么处理”有了一定的了解，如果想了解更多相关知识，欢迎关注丸趣 TV 行业资讯频道，感谢各位的阅读！
正文完
