HA&Federation【转】

HA&Federation【转】
转自：http://blog.csdn.net/tutucute0000/article/details/39756123

从nameNode1、namenode2克隆出namenode3、namenode4，分别作为1、2的standby node。

把datanode2、3、4作为namenode1的jounalnodes，datanode4、5、6作为namenode2的journalnodes。

同时将datanode2、3、4组成zookeeper cluster1，datanode5、6、7组成zookeeper cluster2，分别为namenode1、2管理的两个nameservice，提供namonode间的HA自动切换。

配置

1. 在datanode2、3、4、5、6、7上安装zookeeper
[plain] view plain copy
1. wget http://mirrors.cnnic.cn/apache/zookeeper/zookeeper-3.4.6/zookeeper-3.4.6.tar.gz
2. tar zxvf zookeeper-3.4.6.tar.gz
3. mv zookeeper-3.4.6 zookeeper
4. cd zookeeper/conf
5. cp zoo_sample.cfg zoo.cfg
修改datanode2、3、4中的zoo.cfg文件为
[plain] view plain copy
1. tickTime=2000
2. initLimit=10
3. syncLimit=5
4. dataDir=/home/hduser/mydata/zookeeper/zkdata
5. dataLogDir=/home/hduser/mydata/zookeeper/zkdatalog
6. clientPort=2181
7. server.1=datanode2:2888:3888
8. server.2=datanode3:2888:3888
9. server.3=datanode4:2888:3888
datanode5、6、7中的zoo.cfg为
[plain] view plain copy
1. tickTime=2000
2. initLimit=10
3. syncLimit=5
4. dataDir=/home/hduser/mydata/zookeeper/zkdata
5. dataLogDir=/home/hduser/mydata/zookeeper/zkdatalog
6. clientPort=2181
7. server.1=datanode5:2888:3888
8. server.2=datanode6:2888:3888
9. server.3=datanode7:2888:3888
在每个节点上创建配置文件中所设的目录，同时在dataDir下新建myid文件
[plain] view plain copy
1. mkdir -p /home/hduser/mydata/zookeeper/zkdata
2. mkdir -p /home/hduser/mydata/zookeeper/zkdatalog
3. vi /home/hduser/mydata/zookeeper/zkdata/myid
myid文件里就写一个数字，就是zoo.cfg文件里所配置的

server.X = datanodeN:2888:3888

中，节点N所对应的数字X。所以datanode2、5中myid文件内容就是数字1，datanode3、6是数字2，datanode4、7是数字3。

最后还要设置一下环境变量，添加一行到~/.bashrc
[plain] view plain copy
1. export ZOOKEEPER_HOME=/usr/local/zookeeper
2. export PATH=$ZOOKEEPER_HOME/bin:$PATH
2. 修改配置文件core-site.xml

基本的配置前面的文章都写了，这里只贴出需要修改和增加的配置项。

修改所有namenode和datanode的core-site.xml
[plain] view plain copy
1. <property>
2. <name>fs.defaultFS</name>
3. <value>viewfs:///</value>
4. </property>
5. <property>
6. <name>fs.viewfs.mounttable.default.link./ns1</name>
7. <value>hdfs://ns1</value>
8. </property>
9. <property>
10. <name>fs.viewfs.mounttable.default.link./ns2</name>
11. <value>hdfs://ns2</value>
12. </property>
其中ns1、ns2与下面hdfs-site.xml中配置的nameservices对应。

在datanode1、3的core-site.xml中添加
[plain] view plain copy
1. <property>
2. <name>ha.zookeeper.quorum</name>
3. <value>datanode2:2181,datanode3:2181,datanode4:2181</value>
4. </property>
在datanode2、4的core-site.xml中添加
[plain] view plain copy
1. <property>
2. <name>ha.zookeeper.quorum</name>
3. <value>datanode5:2181,datanode6:2181,datanode7:2181</value>
4. </property>
3.修改配置文件hdfs-site.xml

在所有节点(datanode1-3、namenode2-7)中的hdfs-site.xml中都添加以下配置
[plain] view plain copy
1. <property>
2. <name>dfs.nameservices</name>
3. <value>ns1,ns2</value>
4. </property>
5. <property>
6. <name>dfs.ha.namenodes.ns1</name>
7. <value>nn1,nn3</value>
8. </property>
9. <property>
10. <name>dfs.namenode.rpc-address.ns1.nn1</name>
11. <value>namenode1:9000</value>
12. </property>
13. <property>
14. <name>dfs.namenode.rpc-address.ns1.nn3</name>
15. <value>namenode3:9000</value>
16. </property>
17. <property>
18. <name>dfs.client.failover.proxy.provider.ns1</name>
19. <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
20. </property>
21. <property>
22. <name>dfs.ha.namenodes.ns2</name>
23. <value>nn2,nn4</value>
24. </property>
25. <property>
26. <name>dfs.namenode.rpc-address.ns2.nn2</name>
27. <value>namenode2:9000</value>
28. </property>
29. <property>
30. <name>dfs.namenode.rpc-address.ns2.nn4</name>
31. <value>namenode4:9000</value>
32. </property>
33. <property>
34. <name>dfs.client.failover.proxy.provider.ns2</name>
35. <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
36. </property>
37. <property>
38. <name>dfs.ha.fencing.methods</name>
39. <value>sshfence</value>
40. </property>
41. <property>
42. <name>dfs.ha.fencing.ssh.private-key-files</name>
43. <value>/home/hduser/.ssh/id_rsa</value>
44. </property>
45. <property>
46. <name>dfs.ha.fencing.ssh.connect-timeout</name>
47. <value>30000</value>
48. </property>
49. <property>
50. <name>dfs.ha.automatic-failover.enabled</name>
51. <value>true</value>
52. </property>
53. <property>
54. <name>dfs.journalnode.edits.dir</name>
55. <value>/home/hduser/mydata/hdfs/journalnode</value>
56. </property>
然后单独在namenode1、3的hdfs-site.xml中添加
[plain] view plain copy
1. <property>
2. <name>dfs.namenode.shared.edits.dir</name>
3. <value>qjournal://datanode2:8485;datanode3:8485;datanode4:8485/ns1</value>
4. </property>
在namenode2、4的hdfs-site.xml中添加
[plain] view plain copy
1. <property>
2. <name>dfs.namenode.shared.edits.dir</name>
3. <value>qjournal://datanode5:8485;datanode6:8485;datanode7:8485/ns2</value>
4. </property>
每个journalnode的上都要创建配置dfs.journalnode.edits.dir中所设的目录
[plain] view plain copy
1. mkdir -p /home/hduser/mydata/hdfs/journalnode
值得注意的是dfs.ha.fencing.methods配置。

它是用来保证同一时刻只能有一个active namenode在工作。如果配置成sshfence，则代表在发生failover的时候，使用ssh登录到active namenode上，然后使用fuser干掉active namenode上的进程。所以在namenode间还需要有如下配置：

1. 配置dfs.ha.fencing.ssh.private-key-files指定私钥。或者直接就在所有节点之间配置好ssh免密码登录，以后在集群间跳转也方便。

2. 确保所有namenode上都能使用fuser命令。如果没有，使用下面的命令安装
[plain] view plain copy
1. sudo apt-get install psmisc
启动集群

1. 启动zookeeper

在namenode2-7上运行
[plain] view plain copy
1. hduser@datanode5:~$ zkServer.sh start
2. JMX enabled by default
3. Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg
4. Starting zookeeper ... STARTED
检查进程
[plain] view plain copy
1. hduser@datanode5:~$ jps
2. 16448 Jps
3. 16425 QuorumPeerMain
2. 启动journalnodes

在namenode2-7上运行
[plain] view plain copy
1. hduser@datanode5:~$ hadoop-daemon.sh start journalnode
2. starting journalnode, logging to /usr/local/hadoop/logs/hadoop-hduser-journalnode-datanode5.out
检查进程
[plain] view plain copy
1. hduser@datanode5:~$ jps
2. 16472 JournalNode
3. 16518 Jps
4. 16425 QuorumPeerMain
3. 启动namenode

由于我是将之前搭建的非HA集群转换成HA集群，所以不需要重新format namenode，只需要做以下两步

首先使用namenode中的信息初始化journalnodes。

在namenode1、2里运行
[plain] view plain copy
1. hdfs namenode -initializeSharedEdits
然后启动namenode1、2
[plain] view plain copy
1. hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs start namenode
登录到namenode3、4，将namenode1、2中的元数据同步到它们各自对应的standby namenode上。
[plain] view plain copy
1. hdfs namenode -bootstrapStandby
然后再启动namenode3、4。

由于刚启动时所有的namenode都是standby状态，所有namenode log里会有如下错误。
[plain] view plain copy
1. 2014-10-03 19:39:47,971 INFO org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer: Triggering log roll on remote NameNode namenode4/192.168
2. .1.81:9000
3. 2014-10-03 19:39:47,980 WARN org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer: Unable to trigger a roll of the active NN
4. org.apache.hadoop.ipc.RemoteException(org.apache.hadoop.ipc.StandbyException): Operation category JOURNAL is not supported in state standby
5. at org.apache.hadoop.hdfs.server.namenode.ha.StandbyState.checkOperation(StandbyState.java:87)
6. at org.apache.hadoop.hdfs.server.namenode.NameNode$NameNodeHAContext.checkOperation(NameNode.java:1688)
7. at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.checkOperation(FSNamesystem.java:1258)
8. at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.rollEditLog(FSNamesystem.java:5765)
9. at org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer.rollEditLog(NameNodeRpcServer.java:886)
理论上这时候是可以用HAAdmin的命令手动将某一个namenode指定成active的状态的，可奇怪的是当配置了两个nameservices以后，该命令就识别不了namenode了
[plain] view plain copy
1. hduser@namenode4:~$ hdfs haadmin -transitionToActive nn3
2. Illegal argument: Unable to determine the nameservice id.
如果单独配置HA，即只有一个nameservice，就不会有这个问题，暂时不清楚为什么。

所以目前只能通过automatic failover的方式，让ZKFC自动管理namenode的状态。

在namenode1、2里运行
[plain] view plain copy
1. hdfs zkfc -formatZK
然后在所有的namenode上运行
[plain] view plain copy
1. hadoop-daemon.sh start zkfc
这时候再检查每个namenode上ZKFC的log，

namenode1上的ZKFC log最后一行显示
[plain] view plain copy
1. Successfully transitioned NameNode at namenode1/192.168.1.71:9000 to active state
namenode2显示
[plain] view plain copy
1. Successfully transitioned NameNode at namenode2/192.168.1.72:9000 to active state
namenode3显示
[plain] view plain copy
1. Successfully transitioned NameNode at namenode3/192.168.1.80:9000 to standby state
namenode4显示
[plain] view plain copy
1. Successfully transitioned NameNode at namenode4/192.168.1.81:9000 to standby state
至此所有的namenode启动成功了。

4. 启动datanode

到所有的datanode上运行
[plain] view plain copy
1. hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs start datanode
集群启动完毕。

验证

访问http://namenode1:50070/dfsclusterhealth.jsp

看着四个namenode都很正常。

先试着把namenode1干掉
[plain] view plain copy
1. hduser@namenode1:~$ jps
2. 7590 Jps
3. 7061 NameNode
4. 17441 DFSZKFailoverController
5. hduser@namenode1:~$ kill -9 7061
登到namenode3中看ZKFC的log
[plain] view plain copy
1. 2014-10-03 21:00:55,943 INFO org.apache.hadoop.ha.NodeFencer: ====== Beginning Service Fencing Process... ======
2. 2014-10-03 21:00:55,943 INFO org.apache.hadoop.ha.NodeFencer: Trying method 1/1: org.apache.hadoop.ha.SshFenceByTcpPort(null)
3. 2014-10-03 21:00:55,944 INFO org.apache.hadoop.ha.SshFenceByTcpPort: Connecting to namenode1...
4. 2014-10-03 21:00:55,944 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Connecting to namenode1 port 22
5. 2014-10-03 21:00:55,946 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Connection established
6. 2014-10-03 21:00:55,953 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Remote version string: SSH-2.0-OpenSSH_5.9p1 Debian-5ubuntu1.1
7. 2014-10-03 21:00:55,953 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Local version string: SSH-2.0-JSCH-0.1.42
8. 2014-10-03 21:00:55,953 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: CheckCiphers: aes256-ctr,aes192-ctr,aes128-ctr,aes256-cbc,aes192-cbc,
9. aes128-cbc,3des-ctr,arcfour,arcfour128,arcfour256
10. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: aes256-ctr is not available.
11. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: aes192-ctr is not available.
12. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: aes256-cbc is not available.
13. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: aes192-cbc is not available.
14. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: arcfour256 is not available.
15. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: SSH_MSG_KEXINIT sent
16. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: SSH_MSG_KEXINIT received
17. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: kex: server->client aes128-ctr hmac-md5 none
18. 2014-10-03 21:00:55,958 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: kex: client->server aes128-ctr hmac-md5 none
19. 2014-10-03 21:00:55,960 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: SSH_MSG_KEXDH_INIT sent
20. 2014-10-03 21:00:55,960 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: expecting SSH_MSG_KEXDH_REPLY
21. 2014-10-03 21:00:55,964 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: ssh_rsa_verify: signature true
22. 2014-10-03 21:00:55,965 WARN org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Permanently added 'namenode1' (RSA) to the list of known hosts.
23. 2014-10-03 21:00:55,965 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: SSH_MSG_NEWKEYS sent
24. 2014-10-03 21:00:55,965 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: SSH_MSG_NEWKEYS received
25. 2014-10-03 21:00:55,966 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: SSH_MSG_SERVICE_REQUEST sent
26. 2014-10-03 21:00:55,966 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: SSH_MSG_SERVICE_ACCEPT received
27. 2014-10-03 21:00:55,968 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Authentications that can continue: publickey,keyboard-interactive,pas
28. sword
29. 2014-10-03 21:00:55,968 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Next authentication method: publickey
30. 2014-10-03 21:00:55,996 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Authentication succeeded (publickey).
31. 2014-10-03 21:00:55,996 INFO org.apache.hadoop.ha.SshFenceByTcpPort: Connected to namenode1
32. 2014-10-03 21:00:55,996 INFO org.apache.hadoop.ha.SshFenceByTcpPort: Looking for process running on port 9000
33. 2014-10-03 21:00:56,062 INFO org.apache.hadoop.ha.SshFenceByTcpPort: Indeterminate response from trying to kill service. Verifying whether it i
34. s running using nc...
35. 2014-10-03 21:00:56,118 INFO org.apache.hadoop.ha.SshFenceByTcpPort: Verified that the service is down.
36. 2014-10-03 21:00:56,118 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Disconnecting from namenode1 port 22
37. 2014-10-03 21:00:56,118 INFO org.apache.hadoop.ha.NodeFencer: ====== Fencing successful by method org.apache.hadoop.ha.SshFenceByTcpPort(null)
38. ======
39. 2014-10-03 21:00:56,118 INFO org.apache.hadoop.ha.ActiveStandbyElector: Writing znode /hadoop-ha/ns1/ActiveBreadCrumb to indicate that the loca
40. l node is the most recent active...
41. 2014-10-03 21:00:56,119 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Caught an exception, leaving main loop due to Socket closed
42. 2014-10-03 21:00:56,183 INFO org.apache.hadoop.ha.ZKFailoverController: Trying to make NameNode at namenode3/192.168.1.80:9000 active...
43. 2014-10-03 21:00:57,323 INFO org.apache.hadoop.ha.ZKFailoverController: Successfully transitioned NameNode at namenode3/192.168.1.80:9000 to ac
44. tive state
看到namenode3成功变成了active的状态。

最后再试一下在拷贝文件的过程中干掉一个active的namenode会不会有问题。

准备一个2G多的大文件，拷贝到hdfs里
[plain] view plain copy
1. hduser@datanode5:~$ hdfs dfs -cp file:///home/hduser/test.tar.gz /ns2/test2/
2. 14/10/04 00:33:56 INFO retry.RetryInvocationHandler: Exception while invoking renewLease of class ClientNamenodeProtocolTranslatorPB over namenode4/192.168.1.81:9000. Trying to fail over immediately.
3. java.net.ConnectException: Call From datanode5/192.168.1.77 to namenode4:9000 failed on connection exception: java.net.ConnectException: Connection refused; For more details see: http://wiki.apache.org/hadoop/ConnectionRefused
4. at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
5. at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:57)
7. ...
拷贝了一会儿active namenode居然自己挂了。。。还好还有一个

控制台输出显示成功切换了namenode过去，继续上传文件。
[plain] view plain copy
1. 14/10/04 00:42:54 WARN retry.RetryInvocationHandler: A failover has occurred since the start of this method invocation attempt.
2. 14/10/04 00:43:49 WARN hdfs.DFSClient: Slow ReadProcessor read fields took 56210ms (threshold=30000ms); ack: seqno: 25725 status: SUCCESS status: SUCCESS status: SUCCESS downstreamAckTimeNanos: 56224427288, targets: [192.168.1.77:50010, 192.168.1.74:50010, 192.168.1.79:50010]
3. ...
不一会儿功夫这个namenode也挂了。看了log，挂的原因都是一个
[plain] view plain copy
1. 2014-10-04 00:37:12,798 INFO org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 6001 ms (timeout=20000 ms) for a response for
2. sendEdits. Succeeded so far: [192.168.1.79:8485]
3. 2014-10-04 00:37:13,799 INFO org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 7002 ms (timeout=20000 ms) for a response for
4. sendEdits. Succeeded so far: [192.168.1.79:8485]
5. 2014-10-04 00:37:14,799 INFO org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 8003 ms (timeout=20000 ms) for a response for
6. sendEdits. Succeeded so far: [192.168.1.79:8485]
7. 2014-10-04 00:37:15,801 INFO org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 9004 ms (timeout=20000 ms) for a response for
8. sendEdits. Succeeded so far: [192.168.1.79:8485]
9. 2014-10-04 00:37:16,802 INFO org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 10005 ms (timeout=20000 ms) for a response fo
10. r sendEdits. Succeeded so far: [192.168.1.79:8485]
11. 2014-10-04 00:37:17,803 INFO org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 11006 ms (timeout=20000 ms) for a response fo
12. r sendEdits. Succeeded so far: [192.168.1.79:8485]
13. 2014-10-04 00:37:18,804 INFO org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 12007 ms (timeout=20000 ms) for a response fo
14. r sendEdits. Succeeded so far: [192.168.1.79:8485]
15. 2014-10-04 00:37:19,804 INFO org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 13008 ms (timeout=20000 ms) for a response fo
16. r sendEdits. Succeeded so far: [192.168.1.79:8485]
17. 2014-10-04 00:37:20,806 WARN org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 14009 ms (timeout=20000 ms) for a response fo
18. r sendEdits. Succeeded so far: [192.168.1.79:8485]
19. 2014-10-04 00:37:21,806 WARN org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 15010 ms (timeout=20000 ms) for a response fo
20. r sendEdits. Succeeded so far: [192.168.1.79:8485]
21. 2014-10-04 00:37:22,808 WARN org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 16011 ms (timeout=20000 ms) for a response fo
22. r sendEdits. Succeeded so far: [192.168.1.79:8485]
23. 2014-10-04 00:37:23,808 WARN org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 17012 ms (timeout=20000 ms) for a response fo
24. r sendEdits. Succeeded so far: [192.168.1.79:8485]
25. 2014-10-04 00:37:23,864 INFO org.apache.hadoop.hdfs.server.blockmanagement.CacheReplicationMonitor: Rescanning after 30000 milliseconds
26. 2014-10-04 00:37:23,864 INFO org.apache.hadoop.hdfs.server.blockmanagement.CacheReplicationMonitor: Scanned 0 directive(s) and 0 block(s) in 1
27. millisecond(s).
28. 2014-10-04 00:37:24,810 WARN org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 18013 ms (timeout=20000 ms) for a response fo
29. r sendEdits. Succeeded so far: [192.168.1.79:8485]
30. 2014-10-04 00:37:25,811 WARN org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 19014 ms (timeout=20000 ms) for a response fo
31. r sendEdits. Succeeded so far: [192.168.1.79:8485]
32. 2014-10-04 00:37:26,798 FATAL org.apache.hadoop.hdfs.server.namenode.FSEditLog: Error: flush failed for required journal (JournalAndStream(mgr
33. =QJM to [192.168.1.77:8485, 192.168.1.78:8485, 192.168.1.79:8485], stream=QuorumOutputStream starting at txid 506))
34. java.io.IOException: Timed out waiting 20000ms for a quorum of nodes to respond.
35. at org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.waitForWriteQuorum(AsyncLoggerSet.java:137)
36. at org.apache.hadoop.hdfs.qjournal.client.QuorumOutputStream.flushAndSync(QuorumOutputStream.java:107)
37. at org.apache.hadoop.hdfs.server.namenode.EditLogOutputStream.flush(EditLogOutputStream.java:113)
38. at org.apache.hadoop.hdfs.server.namenode.EditLogOutputStream.flush(EditLogOutputStream.java:107)
39. at org.apache.hadoop.hdfs.server.namenode.JournalSet$JournalSetOutputStream$8.apply(JournalSet.java:499)
40. at org.apache.hadoop.hdfs.server.namenode.JournalSet.mapJournalsAndReportErrors(JournalSet.java:359)
41. at org.apache.hadoop.hdfs.server.namenode.JournalSet.access$100(JournalSet.java:57)
42. at org.apache.hadoop.hdfs.server.namenode.JournalSet$JournalSetOutputStream.flush(JournalSet.java:495)
43. at org.apache.hadoop.hdfs.server.namenode.FSEditLog.logSync(FSEditLog.java:625)
44. at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getAdditionalBlock(FSNamesystem.java:2839)
45. at org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer.addBlock(NameNodeRpcServer.java:606)
46. at org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolServerSideTranslatorPB.addBlock(ClientNamenodeProtocolServerSideTranslatorP
47. B.java:455)
48. at org.apache.hadoop.hdfs.protocol.proto.ClientNamenodeProtocolProtos$ClientNamenodeProtocol$2.callBlockingMethod(ClientNamenodeProtoc
49. olProtos.java)
50. at org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:585)
51. at org.apache.hadoop.ipc.RPC$Server.call(RPC.java:928)
52. at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2013)
53. at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2009)
54. at java.security.AccessController.doPrivileged(Native Method)
55. at javax.security.auth.Subject.doAs(Subject.java:415)
56. at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1614)
57. at org.apache.hadoop.ipc.Server$Handler.run(Server.java:2007)
58. 2014-10-04 00:37:26,798 WARN org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Aborting QuorumOutputStream starting at txid 506
59. 2014-10-04 00:37:26,799 INFO org.apache.hadoop.util.ExitUtil: Exiting with status 1
60. 2014-10-04 00:37:26,801 INFO org.apache.hadoop.hdfs.server.namenode.NameNode: SHUTDOWN_MSG:
61. /************************************************************
62. SHUTDOWN_MSG: Shutting down NameNode at namenode2/192.168.1.72
连接192.168.1.79超时了，这边的网络确实是不太好。不过这也说明了采用QJM方式的HA集群对Journal Node依赖太大了，需要注意。

试图重新启动两个namenode，发现ZKFC也挂了，日志报错如下
[plain] view plain copy
1. 2014-10-04 00:38:15,708 INFO org.apache.zookeeper.ClientCnxn: Opening socket connection to server datanode5/192.168.1.77:2181. Will not attemp
2. t to authenticate using SASL (unknown error)
3. 2014-10-04 00:38:15,717 INFO org.apache.zookeeper.ClientCnxn: Socket connection established to datanode5/192.168.1.77:2181, initiating session
4. 2014-10-04 00:38:17,386 INFO org.apache.zookeeper.ClientCnxn: Client session timed out, have not heard from server in 1669ms for sessionid 0x0
5. , closing socket connection and attempting reconnect
6. 2014-10-04 00:38:18,288 INFO org.apache.zookeeper.ClientCnxn: Opening socket connection to server datanode6/192.168.1.78:2181. Will not attemp
7. t to authenticate using SASL (unknown error)
8. 2014-10-04 00:38:18,289 INFO org.apache.zookeeper.ClientCnxn: Socket connection established to datanode6/192.168.1.78:2181, initiating session
9. 2014-10-04 00:38:19,957 INFO org.apache.zookeeper.ClientCnxn: Client session timed out, have not heard from server in 1668ms for sessionid 0x0
10. , closing socket connection and attempting reconnect
11. 2014-10-04 00:38:20,585 INFO org.apache.zookeeper.ClientCnxn: Opening socket connection to server datanode7/192.168.1.79:2181. Will not attemp
12. t to authenticate using SASL (unknown error)
13. 2014-10-04 00:38:20,586 INFO org.apache.zookeeper.ClientCnxn: Socket connection established to datanode7/192.168.1.79:2181, initiating session
14. 2014-10-04 00:38:20,587 INFO org.apache.zookeeper.ClientCnxn: Unable to read additional data from server sessionid 0x0, likely server has clos
15. ed socket, closing socket connection and attempting reconnect
16. 2014-10-04 00:38:20,710 ERROR org.apache.hadoop.ha.ActiveStandbyElector: Connection timed out: couldn't connect to ZooKeeper in 5000 milliseco
17. nds
18. 2014-10-04 00:38:21,750 INFO org.apache.zookeeper.ZooKeeper: Session: 0x0 closed
19. 2014-10-04 00:38:21,753 INFO org.apache.zookeeper.ClientCnxn: EventThread shut down
20. 2014-10-04 00:38:21,756 FATAL org.apache.hadoop.ha.ZKFailoverController: Unable to start failover controller. Unable to connect to ZooKeeper q
21. uorum at datanode5:2181,datanode6:2181,datanode7:2181. Please check the configured value for ha.zookeeper.quorum and ensure that ZooKeeper is
22. running.
试着telnet一下，端口没挂，但是连接速度慢也有卡死现象，看来还是网络问题。这又说明了采用ZKFC方式做automatic failover对zookeeper的依赖。

重新启动zkfc和namenode，磕磕跘跘的终于拷完了，临了还出了个小问题
[plain] view plain copy
1. 14/10/04 00:46:30 WARN hdfs.DFSClient: DFSOutputStream ResponseProcessor exception for block BP-1850652326-192.168.1.72-1411567722328:blk_1073741916_1098
2. java.net.SocketTimeoutException: 75000 millis timeout while waiting for channel to be ready for read. ch : java.nio.channels.SocketChannel[connected local=/192.168.1.77:38739 remote=/192.168.1.77:50010]
3. at org.apache.hadoop.net.SocketIOWithTimeout.doIO(SocketIOWithTimeout.java:164)
4. at org.apache.hadoop.net.SocketInputStream.read(SocketInputStream.java:161)
5. at org.apache.hadoop.net.SocketInputStream.read(SocketInputStream.java:131)
6. at org.apache.hadoop.net.SocketInputStream.read(SocketInputStream.java:118)
7. at java.io.FilterInputStream.read(FilterInputStream.java:83)
8. at java.io.FilterInputStream.read(FilterInputStream.java:83)
9. at org.apache.hadoop.hdfs.protocolPB.PBHelper.vintPrefixed(PBHelper.java:1998)
10. at org.apache.hadoop.hdfs.protocol.datatransfer.PipelineAck.readFields(PipelineAck.java:176)
11. at org.apache.hadoop.hdfs.DFSOutputStream$DataStreamer$ResponseProcessor.run(DFSOutputStream.java:798)
12. 14/10/04 00:46:30 WARN hdfs.DFSClient: Error Recovery for block BP-1850652326-192.168.1.72-1411567722328:blk_1073741916_1098 in pipeline 192.168.1.77:50010, 192.168.1.74:50010, 192.168.1.76:50010: bad datanode 192.168.1.77:50010
13. 14/10/04 00:46:33 WARN hdfs.DFSClient: Slow waitForAckedSeqno took 78393ms (threshold=30000ms)
14. hduser@datanode5:~$
15. hduser@datanode5:~$ hdfs dfs -ls /ns2/test2/
16. Found 1 items
17. -rw-r--r-- 3 hduser supergroup 2001616896 2014-10-04 00:43 /ns2/test2/test.tar.gz
把文件再从hdfs上拉下来，使用md5校验是否与原文件一致
[plain] view plain copy
1. // 目标路径
2. hdfs dfs -get /ns2/test2/test.tar.gz
3. // 原始文件路径
4. md5sum test.tar.gz > test.tar.gz.md5
5. // 拷贝test.tar.gz.md5到目标路径
6. md5sum -c test.tar.gz.md5
7. test.tar.gz: OK
看来虽然上传的不太顺利，但是结果还是好的。

一些关于HA+Federation的意义和相关组件的介绍，可以参见以下文章

http://www.infoq.com/cn/articles/Hadoop-2-0-namenode-ha-federation-practice-zh/
相关阅读:
[bzoj3295][Cqoi2011][动态逆序对] (树套树)
[bzoj3209][花神的数论题] (数位dp+费马小定理)
[bzoj1026][SCOI2009][windy数] (数位dp)
[bzoj4521][Cqoi2016][手机号码] (数位dp+记忆化搜索)
[bzoj1833][ZJOI2010][count] (数位dp)
[spoj1182][Sorted Bit Sequence] (数位dp)
[ural1057][Amount of Degrees] (数位dp+进制模型)
[hdu3652][B-number] (数位dp)
【bzoj2523】【CTSC2001】聪明的学生
 友情链接（有的是单向的）
原文地址：https://www.cnblogs.com/seaspring/p/7088791.html