hadoop 在redhat linux5 上部署成功(三机)

chluamsbze

断断续续，折腾了一个礼拜，终于在出差的时候把这个hadoop部署成功了。
我的环境是 VMware + redhat linux5，一台namenode，一台dataqnode，配置的过程中也遇到不少麻烦，一会儿再一一列举。
后来又在slaves里面加了个datanode节点，也成功了。看来这段时间的不断尝试没有白费，接下来就要开始在window+cygwin+eclipse里面写分布式搜索引擎了。
现在将成功的运行结果列出来。

[iyunv@rh0 bin]# hadoop fs -lsr /tmp
drwxr-xr-x   - root supergroup          0 2010-11-07 09:41 /tmp/hadoop-root
drwxr-xr-x   - root supergroup          0 2010-11-07 09:41 /tmp/hadoop-root/tmp
drwxr-xr-x   - root supergroup          0 2010-11-07 09:41 /tmp/hadoop-root/tmp/mapred
drwx-wx-wx   - root supergroup          0 2010-11-07 09:42 /tmp/hadoop-root/tmp/mapred/system
[iyunv@rh0 bin]# hadoop dfsadmin -report
Configured Capacity: 13391486976 (12.47 GB)
Present Capacity: 8943796224 (8.33 GB)
DFS Remaining: 8943742976 (8.33 GB)
DFS Used: 53248 (52 KB)
DFS Used%: 0%
Under replicated blocks: 0
Blocks with corrupt replicas: 0
Missing blocks: 0
-------------------------------------------------
Datanodes available: 1 (1 total, 0 dead)
Name: 192.168.126.101:50010
Decommission Status : Normal
Configured Capacity: 13391486976 (12.47 GB)
DFS Used: 53248 (52 KB)
Non DFS Used: 4447690752 (4.14 GB)
DFS Remaining: 8943742976(8.33 GB)
DFS Used%: 0%
DFS Remaining%: 66.79%
Last contact: Sun Nov 07 09:42:18 CST 2010

[iyunv@rh0 bin]# ./hadoop fs -put /usr/local/cjd/b.txt /tmp/hadoop-root/tmp/cjd/b.txt
[iyunv@rh0 bin]# ./hadoop jar hadoop-0.20.2-examples.jar wordcount /tmp/hadoop-root/tmp/cjd /tmp/hadoop-root/tmp/output-dir
10/11/07 09:43:03 INFO input.FileInputFormat: Total input paths to process : 1
10/11/07 09:43:03 INFO mapred.JobClient: Running job: job_201011070941_0001
10/11/07 09:43:04 INFO mapred.JobClient:  map 0% reduce 0%
10/11/07 09:43:15 INFO mapred.JobClient:  map 100% reduce 0%
10/11/07 09:43:36 INFO mapred.JobClient:  map 100% reduce 100%
10/11/07 09:43:39 INFO mapred.JobClient: Job complete: job_201011070941_0001
10/11/07 09:43:39 INFO mapred.JobClient: Counters: 17
10/11/07 09:43:39 INFO mapred.JobClient:   Job Counters
10/11/07 09:43:39 INFO mapred.JobClient:     Launched reduce tasks=1
10/11/07 09:43:39 INFO mapred.JobClient:     Launched map tasks=1
10/11/07 09:43:39 INFO mapred.JobClient:     Data-local map tasks=1
10/11/07 09:43:39 INFO mapred.JobClient:   FileSystemCounters
10/11/07 09:43:39 INFO mapred.JobClient:     FILE_BYTES_READ=1836
10/11/07 09:43:39 INFO mapred.JobClient:     HDFS_BYTES_READ=1366
10/11/07 09:43:39 INFO mapred.JobClient:     FILE_BYTES_WRITTEN=3704
10/11/07 09:43:39 INFO mapred.JobClient:     HDFS_BYTES_WRITTEN=1306
10/11/07 09:43:39 INFO mapred.JobClient:   Map-Reduce Framework
10/11/07 09:43:39 INFO mapred.JobClient:     Reduce input groups=131
10/11/07 09:43:39 INFO mapred.JobClient:     Combine output records=131
10/11/07 09:43:39 INFO mapred.JobClient:     Map input records=31
10/11/07 09:43:39 INFO mapred.JobClient:     Reduce shuffle bytes=1836
10/11/07 09:43:39 INFO mapred.JobClient:     Reduce output records=131
10/11/07 09:43:39 INFO mapred.JobClient:     Spilled Records=262
10/11/07 09:43:39 INFO mapred.JobClient:     Map output bytes=2055
10/11/07 09:43:39 INFO mapred.JobClient:     Combine input records=179
10/11/07 09:43:39 INFO mapred.JobClient:     Map output records=179
10/11/07 09:43:39 INFO mapred.JobClient:     Reduce input records=131
[iyunv@rh0 bin]# hadoop dfsadmin -report
Configured Capacity: 13391486976 (12.47 GB)
Present Capacity: 8943679754 (8.33 GB)
DFS Remaining: 8943587328 (8.33 GB)
DFS Used: 92426 (90.26 KB)
DFS Used%: 0%
Under replicated blocks: 0
Blocks with corrupt replicas: 0
Missing blocks: 0
-------------------------------------------------
Datanodes available: 1 (1 total, 0 dead)
Name: 192.168.126.101:50010
Decommission Status : Normal
Configured Capacity: 13391486976 (12.47 GB)
DFS Used: 92426 (90.26 KB)
Non DFS Used: 4447807222 (4.14 GB)
DFS Remaining: 8943587328(8.33 GB)
DFS Used%: 0%
DFS Remaining%: 66.79%
Last contact: Sun Nov 07 09:46:22 CST 2010

[iyunv@rh0 bin]# hadoop fs -lsr /tmp
drwxr-xr-x   - root supergroup          0 2010-11-07 09:41 /tmp/hadoop-root
drwxr-xr-x   - root supergroup          0 2010-11-07 09:43 /tmp/hadoop-root/tmp
drwxr-xr-x   - root supergroup          0 2010-11-07 09:42 /tmp/hadoop-root/tmp/cjd
-rw-r--r--   1 root supergroup       1366 2010-11-07 09:42 /tmp/hadoop-root/tmp/cjd/b.txt
drwxr-xr-x   - root supergroup          0 2010-11-07 09:41 /tmp/hadoop-root/tmp/mapred
drwx-wx-wx   - root supergroup          0 2010-11-07 09:43 /tmp/hadoop-root/tmp/mapred/system
-rw-------   1 root supergroup          4 2010-11-07 09:42 /tmp/hadoop-root/tmp/mapred/system/jobtracker.info
drwxr-xr-x   - root supergroup          0 2010-11-07 09:43 /tmp/hadoop-root/tmp/output-dir
drwxr-xr-x   - root supergroup          0 2010-11-07 09:43 /tmp/hadoop-root/tmp/output-dir/_logs
drwxr-xr-x   - root supergroup          0 2010-11-07 09:43 /tmp/hadoop-root/tmp/output-dir/_logs/history
-rw-r--r--   1 root supergroup      29085 2010-11-07 09:43 /tmp/hadoop-root/tmp/output-dir/_logs/history/rh0_1289094087484_job_201011070941_0001_conf.xml
-rw-r--r--   1 root supergroup       7070 2010-11-07 09:43 /tmp/hadoop-root/tmp/output-dir/_logs/history/rh0_1289094087484_job_201011070941_0001_root_word+count
-rw-r--r--   1 root supergroup       1306 2010-11-07 09:43 /tmp/hadoop-root/tmp/output-dir/part-r-00000
[iyunv@rh0 bin]# hadoop fs -get /tmp/hadoop-root/tmp/output-dir /usr/local/cjd/b
[iyunv@rh0 bin]#


【加入一个节点后查看节点】已经有2个可用的了。
Datanodes available: 2 (2 total, 0 dead)

[iyunv@rh0 bin]# hadoop dfsadmin -report
Configured Capacity: 26782973952 (24.94 GB)
Present Capacity: 18191843343 (16.94 GB)
DFS Remaining: 18191613952 (16.94 GB)
DFS Used: 229391 (224.01 KB)
DFS Used%: 0%
Under replicated blocks: 0
Blocks with corrupt replicas: 0
Missing blocks: 0
-------------------------------------------------
Datanodes available: 2 (2 total, 0 dead)
Name: 192.168.126.102:50010
Decommission Status : Normal
Configured Capacity: 13391486976 (12.47 GB)
DFS Used: 53263 (52.01 KB)
Non DFS Used: 4143476721 (3.86 GB)
DFS Remaining: 9247956992(8.61 GB)
DFS Used%: 0%
DFS Remaining%: 69.06%
Last contact: Sun Nov 07 10:06:55 CST 2010

Name: 192.168.126.101:50010
Decommission Status : Normal
Configured Capacity: 13391486976 (12.47 GB)
DFS Used: 176128 (172 KB)
Non DFS Used: 4447653888 (4.14 GB)
DFS Remaining: 8943656960(8.33 GB)
DFS Used%: 0%
DFS Remaining%: 66.79%
Last contact: Sun Nov 07 10:06:55 CST 2010


【控制台截图】



【遇到的问题】
0, 要把export HADOOP_HOME=/usr/local/cjd/hadoop/hadoop-0.20.2 加入 /etc/profile,否则datanode不知道这个？
1,"Name node is in safe mode",执行以下命令.
  bin/hadoop dfsadmin -safemode leave
  bin/hadoop dfsadmin -safemode off
2,如果修改了配置文件，然后datanode起不来，发现报如下错误，
Incompatible namespaceIDs in /tmp/hadoop-root/tmp/dfs/data: namenode namespaceID = 1952086391; datanode namespaceID = 1896626371
那么到 datanode的 /tmp/hadoop-root/tmp/dfs/data/current/文件系统下，把VERSION里面的namespaceID=xxxx 改成 前面这个：1952086391. 重启hadoop即可。
3,配置文件xml里面的机器名要用ip地址。要用ip保险一些。包括 master and slaves文件里面也写ip。
4,要在每个datanode上，telnet namenode 9000，如果能telnet通，说明没有问题。
5,wordcount启动问题：
  192.168.126.100: Address 192.168.126.100 maps to rh0, but this does not map back to the address - POSSIBLE BREAK-IN ATTEMPT!  
该问题的处理如下：请确保 /etc/xxx/xx/host是如下：

# that require network functionality will fail.
#127.0.0.1        rh0 localhost [配置成这个会报错]
127.0.0.1        localhost
::1        localhost6.localdomain6 localhost6
192.168.126.100 rh0
192.168.126.101 rh1
192.168.126.102 rh2