您的位置：首页 > 运维架构

hadoop安装与配置(完全分布模式)

2014-06-11 12:55 423 查看

-------------------------------------------------

一、前言
二、环境
三、配置
1.配置hosts文件及主机名
2.建立hadoop运行账号
3.配置ssh免密码连入
4.安装JDK
5.安装hadoop
6.配置hadoop
7.配置masters和slaves文件
8.向各节点复制hadoop
9.格式化namenode
10.启动hadoop
11.用jps检验各后台进程是否成功启动
四、测试
五、浏览器查看
-------------------------------------------------

一、前言
Hadoop是Apache软件基金会旗下的一个开源分布式计算平台。以Hadoop分布式文件系统（HDFS，Hadoop Distributed Filesystem）和MapReduce（Google MapReduce的开源实现）为核心的Hadoop为用户提供了系统底层细节透明的分布式基础架构。
对于Hadoop的集群来讲，可以分成两大类角色：Master和Salve。一个HDFS集群是由一个NameNode和若干个DataNode组成的。其中NameNode作为主服务器，管理文件系统的命名空间和客户端对文件系统的访问操作；集群中的DataNode管理存储的数据。MapReduce框架是由一个单独运行在主节点上的JobTracker和运行在每个集群从节点的TaskTracker共同组成的。主节点负责调度构成一个作业的所有任务，这些任务分布在不同的从节点上。主节点监控它们的执行情况，并且重新执行之前的失败任务；从节点仅负责由主节点指派的任务。当一个Job被提交时，JobTracker接收到提交作业和配置信息之后，就会将配置信息等分发给从节点，同时调度任务并监控TaskTracker的执行。
从上面的介绍可以看出，HDFS和MapReduce共同组成了Hadoop分布式系统体系结构的核心。HDFS在集群上实现分布式文件系统，MapReduce在集群上实现了分布式计算和任务处理。HDFS在MapReduce任务处理过程中提供了文件操作和存储等支持，MapReduce在HDFS的基础上实现了任务的分发、跟踪、执行等工作，并收集结果，二者相互作用，完成了Hadoop分布式集群的主要任务。

二、环境
1.系统版本:CentOS6.4 32位
JDK版本:jdk-7u45-linux-i586.rpm
Hadoop版本:hadoop-0.20.2.tar.gz
2.角色分类
192.168.2.101 namenode （充当namenode、secondary namenode和ResourceManager角色）
192.168.2.102 datanode1 （充当datanode、nodemanager角色）
192.168.2.103 datanode2 （充当datanode、nodemanager角色）
3.IP地址规划：

4.hadoop组件依赖关系：

5.名词解释
----HDFS-----
HDFS（Hadoop Distributed File System),hadoop分布式文件系统。
NameNode，HDFS命名服务器，负责与DataNode文件信息保存。
DataNode，HDFS数据节点，负责数据存储并汇报给NameNode。
SecondaryNamenode，NameNode的镜像备份节点。
-----Map Reduce------

TackTracker，启动和管理Map和Reduce子任务的节点。
JobTracker，hadoop的Map/Reduce调度器，负责与TackTracker通信分配计算任务并跟踪任务进度。
三、配置
1.配置hosts文件及主机名(每个节点都需要配置，以master主机为例）

# vim /etc/hosts
192.168.2.101 master
192.168.2.102 slave1
192.168.2.103 slave2
# vim /etc/sysconfig/network   //另外两个节点为slave1和slave2
HOSTNAME=master

2.建立hadoop运行账号(每个节点都需要配置)

# useradd hadoop
# passwd hadoop

3.配置ssh免密码连入
3-1.安装openssh和rsync（每个节点都要安装）

# rpm -qa |grep openssh
openssh-5.3p1-84.1.el6.i686
openssh-server-5.3p1-84.1.el6.i686
openssh-clients-5.3p1-84.1.el6.i686
openssh-askpass-5.3p1-84.1.el6.i686
# rpm -qa |grep rsync
rsync-3.0.6-9.el6.i686

3-2.更改ssh配置文档（每个节点都需要修改）

# vim /etc/ssh/sshd_config              //将47-49行注释去掉即可
47 RSAAuthentication yes
48 PubkeyAuthentication yes
49 AuthorizedKeysFile      .ssh/authorized_keys
# service sshd restart
Stopping sshd:                                [  OK  ]
Starting sshd:                                [  OK  ]

3-3.配置Master无密码登录所有Salve

# su - hadoop                                    //切换到hadoop用户（在master主机上配置）
$ ssh-keygen -t rsa                             //生成密码对
Generating public/private rsa key pair.
Enter file in which to save the key (/home/hadoop/.ssh/id_rsa): 回车，默认路径
Created directory '/home/hadoop/.ssh'.
Enter passphrase (empty for no passphrase):                     回车，无密码的密码对
Enter same passphrase again:                                    回车
Your identification has been saved in /home/hadoop/.ssh/id_rsa.
Your public key has been saved in /home/hadoop/.ssh/id_rsa.pub.
The key fingerprint is:
2e:5b:d0:e1:8e:8b:c9:14:81:e4:6b:7b:ef:20:0d:09 hadoop@master
The key's randomart image is:
+--[ RSA 2048]----+
|  .              |
| o .             |
|E o .   .        |
| . o . o .       |
|  = . . S        |
| . + . =         |
|  o = o +        |
|   = = =         |
|    +.=          |
+-----------------+
$ cd .ssh
$ ll
-rw-------. 1 hadoop hadoop 1679 Jun 10 18:39 id_rsa
-rw-r--r--. 1 hadoop hadoop  395 Jun 10 18:39 id_rsa.pub
$ cp id_rsa.pub authorized_keys
$ ll
-rw-r--r--. 1 hadoop hadoop  395 Jun 10 18:40 authorized_keys
-rw-------. 1 hadoop hadoop 1679 Jun 10 18:39 id_rsa
-rw-r--r--. 1 hadoop hadoop  395 Jun 10 18:39 id_rsa.pub
$ chmod 600 authorized_keys           //更改权限
$ ll
-rw-------. 1 hadoop hadoop  395 Jun 10 18:40 authorized_keys
-rw-------. 1 hadoop hadoop 1679 Jun 10 18:39 id_rsa
-rw-r--r--. 1 hadoop hadoop  395 Jun 10 18:39 id_rsa.pub

$ ssh localhost                        //利用master主机自身验证
The authenticity of host 'localhost (::1)' can't be established.
RSA key fingerprint is 85:a0:dd:ce:31:7a:c3:94:85:7c:9e:2d:20:f8:2d:2d.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'localhost' (RSA) to the list of known hosts.

# su - hadoop                    //在slave主机上执行，所有slave主机都要配置，以slave1为例
$ ssh-keygen -t rsa
Generating public/private rsa key pair.
Enter file in which to save the key (/home/hadoop/.ssh/id_rsa): 回车，默认路径
Created directory '/home/hadoop/.ssh'.
Enter passphrase (empty for no passphrase):                     回车，无密码的密码对
Enter same passphrase again:                                    回车
Your identification has been saved in /home/hadoop/.ssh/id_rsa.
Your public key has been saved in /home/hadoop/.ssh/id_rsa.pub.
The key fingerprint is:
b1:80:fb:b7:4a:e6:88:d4:48:02:04:de:c1:b4:2d:03 hadoop@slave2
The key's randomart image is:
+--[ RSA 2048]----+
|oEoo             |
|o o.+.           |
|.. =... .        |
|.   o. . o       |
|. . .   S        |
| o o .           |
|  o . + .        |
| . . = . .       |
|  . . o..        |
+-----------------+

$ scp ~/.ssh/authorized_keys slave1:/home/hadoop/.ssh/   //在master主机上执行，将master主机的公钥分发给所有slave主机
$ cd .ssh/                                               //回到slave1主机上查看
$ ll
-rw-------. 1 hadoop hadoop  395 Jun 10 18:46 authorized_keys
-rw-------. 1 hadoop hadoop 1671 Jun 10 18:45 id_rsa
-rw-r--r--. 1 hadoop hadoop  395 Jun 10 18:45 id_rsa.pub
$ chmod 600 authorized_keys

[hadoop@master ~]$ ssh slave1      //测试master无密码连接所有slave
[hadoop@slave2 ~]$ hostname
slave1
[hadoop@slave2 ~]$ exit
logout
Connection to slave1 closed.
[hadoop@master ~]$ ssh slave2
[hadoop@slave2 ~]$ hostname
slave2
[hadoop@slave2 ~]$ exit
logout
Connection to slave2 closed.

3-4.配置所有Salve无密码登录Master

$ scp ~/.ssh/id_rsa.pub  master:/home/hadoop/        //在所有slave主机上配置，以slave1主机为例

$ cat ~/id_rsa.pub >> ~/.ssh/authorized_keys        //在master主机上执行，将所有slave主机上的ssh公钥追加到自身中
$ rm -f ~/id_rsa.pub

[hadoop@slave1 .ssh]$ ssh master                   //测试所有的slave主机登录master主机
Last login: Tue Jun 10 18:42:00 2014 from localhost.localdomain
[hadoop@master ~]$ hostname
master
[hadoop@master ~]$ exit
logout
Connection to master closed.
[hadoop@slave2 .ssh]$ ssh master
Last login: Tue Jun 10 18:43:18 2014 from slave1
[hadoop@master ~]$ hostname
master
[hadoop@master ~]$ exit
logout
Connection to master closed.

4.安装JDK（每个节点都要安装，使用root用户）

# rpm -ivh jdk-7u45-linux-i586.rpm
# rpm -ql jdk |less
/usr/java/jdk1.7.0_45              //JDK安装路径，非常重要
# vim /etc/profile                   //写入搜索路径中
54 JAVA_HOME=/usr/java/jdk1.7.0_45
55 PATH=$PATH:$JAVA_HOME/bin
56 export PATH USER LOGNAME MAIL HISTSIZE HISTCONTROL JAVA_HOME
# . /etc/profile
# java -version
java version "1.7.0_45"
Java(TM) SE Runtime Environment (build 1.7.0_45-b18)
Java HotSpot(TM) Client VM (build 24.45-b08, mixed mode, sharing)

5.安装hadoop（先在master主机中安装并配置，然后再拷贝到所有slave主机中）

# tar -zxvf hadoop-0.20.2.tar.gz -C /usr    //以root用户执行
# cd /usr/
# mv hadoop-0.20.2/ hadoop
# ll
drwxr-xr-x.  12 hadoop hadoop  4096 Feb 19  2010 hadoop
# mkdir /usr/hadoop/tmp
# chown -R hadoop:hadoop /usr/hadoop/tmp
# vim /etc/profile                         //此步骤所有节点都需要配置(仿照JAVA_HOME即可)
54 JAVA_HOME=/usr/java/jdk1.7.0_45
55 HADOOP_HOME=/usr/hadoop
56 PATH=$PATH:$JAVA_HOME/bin
57 PATH=$PATH:$HADOOP_HOME/bin
58 export PATH USER LOGNAME MAIL HOSTNAME HISTSIZE HISTCONTROL JAVA_HOME HADOOP_HOME
# . /etc/profile

6.配置hadoop（先在master主机中配置，然后再拷贝到所有slave主机中）

# cd /usr/hadoop/conf/                      //使用root用户操作
# vim hadoop-env.sh
9 export JAVA_HOME=/usr/java/jdk1.7.0_45   //修改JDK路径

core-site.xml和hdfs-site.xml是站在HDFS角度上配置文件；core-site.xml和mapred-site.xml是站在MapReduce角度上配置文件。

# vim core-site.xml
<configuration>
<property>
<name>hadoop.tmp.dir</name>
<value>/usr/hadoop/tmp</value>               //数据存储路径
</property>
<property>
<name>fs.default.name</name>
<value>hdfs://192.168.2.101:9000</value>     //HDFS地址和端口
</property>
</configuration>

# vim hdfs-site.xml
<configuration>
<property>
<name>dfs.replication</name>
<value>1</value>                             //数据保存的份数,默认3份
</property>
</configuration>

# vim mapred-site.xml
<configuration>
<property>
<name>mapred.job.tracker</name>
<value>http://192.168.2.101:9001</value>    //JobTracker的地址和端口
</property>
</configuration>

7.配置masters和slaves文件（依然在master主机中配置，在生产环境中应架设DNS服务器，使用域名，本实验使用的是hosts文件，但是为了防止hosts文件出现错误，使用IP更加可靠）

# vim masters
192.168.2.101
# vim slaves
192.168.2.102
192.168.2.103

8.master主机向各节点slave复制hadoop

# scp -r /usr/hadoop/ slave1:/usr/
# scp -r /usr/hadoop/ slave2:/usr/

# chown -R hadoop:hadoop /usr/hadoop/       //修改所有slave主机hadoop的所属关系

9.格式化namenode

[root@master ~]# su - hadoop
[hadoop@master ~]$ hadoop namenode -format    //如无法执行，使用source /etc/profile更新
14/06/10 19:53:43 INFO namenode.NameNode: STARTUP_MSG:
/************************************************************
STARTUP_MSG: Starting NameNode
STARTUP_MSG:   host = master/192.168.2.101
STARTUP_MSG:   args = [-format]
STARTUP_MSG:   version = 0.20.2
STARTUP_MSG:   build = https://svn.apache.org/repos/asf/hadoop/common/branches/branch-0.20 -r 911707; compiled by

'chrisdo' on Fri Feb 19 08:07:34 UTC 2010
************************************************************/
14/06/10 19:53:43 INFO namenode.FSNamesystem: fsOwner=hadoop,hadoop
14/06/10 19:53:43 INFO namenode.FSNamesystem: supergroup=supergroup
14/06/10 19:53:43 INFO namenode.FSNamesystem: isPermissionEnabled=true
14/06/10 19:53:44 INFO common.Storage: Image file of size 96 saved in 0 seconds.
14/06/10 19:53:44 INFO common.Storage: Storage directory /usr/hadoop/tmp/dfs/name has been successfully formatted.
14/06/10 19:53:44 INFO namenode.NameNode: SHUTDOWN_MSG:
/************************************************************
SHUTDOWN_MSG: Shutting down NameNode at master/192.168.2.101
************************************************************/

10.启动hadoop

[hadoop@master bin]$ start-all.sh
starting namenode, logging to /usr/hadoop/bin/../logs/hadoop-hadoop-namenode-master.out
192.168.2.102: starting datanode, logging to /usr/hadoop/bin/../logs/hadoop-hadoop-datanode-slave1.out
192.168.2.103: starting datanode, logging to /usr/hadoop/bin/../logs/hadoop-hadoop-datanode-slave2.out
192.168.2.101: starting secondarynamenode, logging to /usr/hadoop/bin/../logs/hadoop-hadoop-secondarynamenode-master.out
starting jobtracker, logging to /usr/hadoop/bin/../logs/hadoop-hadoop-jobtracker-master.out
192.168.2.103: starting tasktracker, logging to /usr/hadoop/bin/../logs/hadoop-hadoop-tasktracker-slave2.out
192.168.2.102: starting tasktracker, logging to /usr/hadoop/bin/../logs/hadoop-hadoop-tasktracker-slave1.out

11.用jps检验各后台进程是否成功启动

[hadoop@master ~]$ jps
30758 Jps
28827 NameNode
29018 JobTracker
28954 SecondaryNameNode

[hadoop@salve1 ~]$ jps
27508 TaskTracker
29409 Jps
27436 DataNode

[hadoop@salve2 ~]$ jps
27508 TaskTracker
29523 Jps
27437 DataNode

四、测试
简单测试(计算π值)

[hadoop@master ~]$ cd /usr/hadoop/
[hadoop@master hadoop]$ hadoop jar hadoop-0.20.2-examples.jar pi 10 100
Number of Maps  = 10
Samples per Map = 100
Wrote input for Map #0
Wrote input for Map #1
Wrote input for Map #2
Wrote input for Map #3
Wrote input for Map #4
Wrote input for Map #5
Wrote input for Map #6
Wrote input for Map #7
Wrote input for Map #8
Wrote input for Map #9
Starting Job
14/06/14 05:02:13 INFO mapred.FileInputFormat: Total input paths to process : 10
14/06/14 05:02:13 INFO mapred.JobClient: Running job: job_201406132259_0004
14/06/14 05:02:14 INFO mapred.JobClient:  map 0% reduce 0%
14/06/14 05:02:28 INFO mapred.JobClient:  map 20% reduce 0%
14/06/14 05:02:31 INFO mapred.JobClient:  map 40% reduce 0%
14/06/14 05:02:37 INFO mapred.JobClient:  map 80% reduce 0%
14/06/14 05:02:40 INFO mapred.JobClient:  map 80% reduce 26%
14/06/14 05:02:43 INFO mapred.JobClient:  map 100% reduce 26%
14/06/14 05:02:55 INFO mapred.JobClient:  map 100% reduce 100%
14/06/14 05:02:57 INFO mapred.JobClient: Job complete: job_201406132259_0004
14/06/14 05:02:57 INFO mapred.JobClient: Counters: 19
14/06/14 05:02:57 INFO mapred.JobClient:   Job Counters
14/06/14 05:02:57 INFO mapred.JobClient:     Launched reduce tasks=1
14/06/14 05:02:57 INFO mapred.JobClient:     Rack-local map tasks=1
14/06/14 05:02:57 INFO mapred.JobClient:     Launched map tasks=10
14/06/14 05:02:57 INFO mapred.JobClient:     Data-local map tasks=9
14/06/14 05:02:57 INFO mapred.JobClient:   FileSystemCounters
14/06/14 05:02:57 INFO mapred.JobClient:     FILE_BYTES_READ=226
14/06/14 05:02:57 INFO mapred.JobClient:     HDFS_BYTES_READ=1180
14/06/14 05:02:57 INFO mapred.JobClient:     FILE_BYTES_WRITTEN=826
14/06/14 05:02:57 INFO mapred.JobClient:     HDFS_BYTES_WRITTEN=215
14/06/14 05:02:57 INFO mapred.JobClient:   Map-Reduce Framework
14/06/14 05:02:57 INFO mapred.JobClient:     Reduce input groups=20
14/06/14 05:02:57 INFO mapred.JobClient:     Combine output records=0
14/06/14 05:02:57 INFO mapred.JobClient:     Map input records=10
14/06/14 05:02:57 INFO mapred.JobClient:     Reduce shuffle bytes=280
14/06/14 05:02:57 INFO mapred.JobClient:     Reduce output records=0
14/06/14 05:02:57 INFO mapred.JobClient:     Spilled Records=40
14/06/14 05:02:57 INFO mapred.JobClient:     Map output bytes=180
14/06/14 05:02:57 INFO mapred.JobClient:     Map input bytes=240
14/06/14 05:02:57 INFO mapred.JobClient:     Combine input records=0
14/06/14 05:02:57 INFO mapred.JobClient:     Map output records=20
14/06/14 05:02:57 INFO mapred.JobClient:     Reduce input records=20
Job Finished in 45.455 seconds
Estimated value of Pi is 3.14800000000000000000

2.上传本地数据文件文件测试（单词统计,wordcount）

[hadoop@master ~]$ mkdir input
[hadoop@master ~]$ echo "hello word">input/test1.txt
[hadoop@master ~]$ echo "hello hadoop">input/test2.txt

[hadoop@master ~]$ cd /usr/hadoop/
[hadoop@master hadoop]$ hadoop dfs -put ~/input test
[hadoop@master hadoop]$ hadoop dfs -ls test/*
-rw-r--r--   1 hadoop supergroup         11 2014-06-10 20:37 /user/hadoop/test/test1.txt
-rw-r--r--   1 hadoop supergroup         13 2014-06-10 20:37 /user/hadoop/test/test2.txt

[hadoop@master hadoop]$ hadoop jar hadoop-0.20.2-examples.jar wordcount test out
14/06/10 20:40:25 INFO input.FileInputFormat: Total input paths to process : 2
14/06/10 20:40:26 INFO mapred.JobClient: Running job: job_201406102021_0001
14/06/10 20:40:27 INFO mapred.JobClient:  map 0% reduce 0%
14/06/10 20:40:39 INFO mapred.JobClient:  map 50% reduce 0%
14/06/10 20:40:45 INFO mapred.JobClient:  map 100% reduce 0%
14/06/10 20:40:51 INFO mapred.JobClient:  map 100% reduce 100%
14/06/10 20:40:53 INFO mapred.JobClient: Job complete: job_201406102021_0001
14/06/10 20:40:53 INFO mapred.JobClient: Counters: 18
14/06/10 20:40:53 INFO mapred.JobClient:   Job Counters
14/06/10 20:40:53 INFO mapred.JobClient:     Launched reduce tasks=1
14/06/10 20:40:53 INFO mapred.JobClient:     Rack-local map tasks=1
14/06/10 20:40:53 INFO mapred.JobClient:     Launched map tasks=2
14/06/10 20:40:53 INFO mapred.JobClient:     Data-local map tasks=1
14/06/10 20:40:53 INFO mapred.JobClient:   FileSystemCounters
14/06/10 20:40:53 INFO mapred.JobClient:     FILE_BYTES_READ=54
14/06/10 20:40:53 INFO mapred.JobClient:     HDFS_BYTES_READ=24
14/06/10 20:40:53 INFO mapred.JobClient:     FILE_BYTES_WRITTEN=178
14/06/10 20:40:53 INFO mapred.JobClient:     HDFS_BYTES_WRITTEN=24
14/06/10 20:40:53 INFO mapred.JobClient:   Map-Reduce Framework
14/06/10 20:40:53 INFO mapred.JobClient:     Reduce input groups=3
14/06/10 20:40:53 INFO mapred.JobClient:     Combine output records=4
14/06/10 20:40:53 INFO mapred.JobClient:     Map input records=2
14/06/10 20:40:53 INFO mapred.JobClient:     Reduce shuffle bytes=60
14/06/10 20:40:53 INFO mapred.JobClient:     Reduce output records=3
14/06/10 20:40:53 INFO mapred.JobClient:     Spilled Records=8
14/06/10 20:40:53 INFO mapred.JobClient:     Map output bytes=40
14/06/10 20:40:53 INFO mapred.JobClient:     Combine input records=4
14/06/10 20:40:53 INFO mapred.JobClient:     Map output records=4
14/06/10 20:40:53 INFO mapred.JobClient:     Reduce input records=4

[hadoop@master hadoop]$ hadoop dfs -ls
drwxr-xr-x   - hadoop supergroup          0 2014-06-10 20:40 /user/hadoop/out
drwxr-xr-x   - hadoop supergroup          0 2014-06-10 20:37 /user/hadoop/test
[hadoop@master hadoop]$ hadoop dfs -ls ./out
drwxr-xr-x   - hadoop supergroup          0 2014-06-10 20:40 /user/hadoop/out/_logs
-rw-r--r--   1 hadoop supergroup         24 2014-06-10 20:40 /user/hadoop/out/part-r-00000
[hadoop@master hadoop]$ hadoop dfs -cat ./out/*
hadoop	1
hello	2
word	1
cat: Source must be a file.

总结：经测试结果显示hello出现2次，hadoop出现1次，word出现1次。
五、浏览器查看
1.通过用浏览器访问jobtracker所在节点的50030端口监控jobtracker

2.通过用浏览器访问namenode所在节点的50070端口监控集群

PS：错误解决方法：
错误1：
14/06/10 19:51:20 ERROR namenode.NameNode: java.io.IOException: Cannot create directory /usr/hadoop/tmp/dfs/name/current
at org.apache.hadoop.hdfs.server.common.Storage$StorageDirectory.clearDirectory(Storage.java:295)
at org.apache.hadoop.hdfs.server.namenode.FSImage.format(FSImage.java:1086)
at org.apache.hadoop.hdfs.server.namenode.FSImage.format(FSImage.java:1110)
at org.apache.hadoop.hdfs.server.namenode.NameNode.format(NameNode.java:856)
at org.apache.hadoop.hdfs.server.namenode.NameNode.createNameNode(NameNode.java:948)
at org.apache.hadoop.hdfs.server.namenode.NameNode.main(NameNode.java:965)
解决方案：
# chown -R hadoop:hadoop /usr/hadoop/tmp/ //修改数据存储路径的所属关系
错误2：利用jps检测进程时没有NameNode以及SecondaryNameNode
$ jps
29018 JobTracker
29091 Jps
解决方案：
# netstat -tupln //查看9000以及9001端口是否启动
错误3：
当再次执行bin/hadoop namenode -format时，master主机NameNode进程没有启动或者slave主机的DataNode进程无法启动，因为在每次执行bin/hadoop namenode -format时,会为namenode生成namespaceID, 但是在tmp文件夹下的datanode还是保留上次的namespaceID,在启动时,由于namespaceID不一致,导致datanode无法启动。所以只要在每次bin/hadoop namenode -format之前先删除"临时文件夹"就可以启动成功。
解决方案：（所有节点都要删除）
“临时文件夹”是你的conf/core-site.xml文件里的hadoop.tmp.dir这个变量定义的。如果你没定义这个变量，那么就是默认值。即：haddoop/tmp文件夹。无论如何删掉这些主机上的tmp文件夹即可。（不会真正删除，你下次运行时又会生成，而这时候生成的namespaceID就是对的了）。

内容来自用户分享和网络整理，不保证内容的准确性，如有侵权内容，可联系管理员处理

标签： 配置 hadoop namenode

相关文章推荐

新的分享

章节导航