提示:这里有Exit code: 127 Stack trace: ExitCodeException exitCode=127: 错误的解决的方法,在文章最后面
一、首先要配置好java环境
下载地址:http://www.oracle.com/technetwork/java/javase/downloads/jdk8-downloads-2133151.html
mac电脑直接jdk-8u144-macosx-x64.dmg一键安装jdk
然后配置jdk环境,
JAVA_HOME=/Library/Java/JavaVirtualMachines/jdk1.8.0_144.jdk/Contents/Home
CLASSPAHT=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar
PATH=$JAVA_HOME/bin:$PATH:
export JAVA_HOME
export CLASSPATH
export PATH
二、ssh配置
首先确认能够远程登录
系统偏好设置-共享 -远程登录
ssh-keygen -t rsa
Press enter for each line 提示输入直接按回车就好
cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
chmod og-wx ~/.ssh/authorized_keys
ssh localhost
如果ssh localhost还需要密码 查看一下你.ssh目录的权限
.ssh的父目录的权限问题(我的问题就出现在这里):.ssh的父目录文件权限应该是755,即所属用户的 用户文件 (/home下属的一个用户文件)
执行chmod 755 ~/.ssh
三、安装配置hadoop文件
下载地址:http://hadoop.apache.org/releases.html
http://mirrors.tuna.tsinghua.edu.cn/apache/hadoop/common/hadoop-2.8.1/hadoop-2.8.1.tar.gz
tar -zxvf hadoop-2.8.1.tar.gz
cd hadoop-2.8.1
修改配置文件 配置文件在hadoop目录下的/etc/hadoop
1.修改core-site.xml 文件
<configuration>
<property>
<name>fs.defaultFSname>
<value>hdfs://localhost:9000value>
property>
<property>
<name>hadoop.tmp.dirname>
<value>/Users/chenxun/software/hadoop-2.8.1/tempvalue>
property>
configuration>
2.修改hadfs-site.xml
默认副本数3,修改为1,dfs.namenode.name.dir指明fsimage存放目录,多个目录用逗号隔开。dfs.datanode.data.dir指定块文件存放目录,多个目录逗号隔开
<configuration>
<property>
<name>dfs.replicationname>
<value>1value>
property>
<property>
<name>dfs.namenode.name.dirname>
<value>file:/Users/chenxun/software/hadoop-2.8.1/tmp/hdfs/namevalue>
property>
<property>
<name>dfs.datanode.data.dirname>
<value>file:/Users/chenxun/software/hadoop-2.8.1/tmp/hdfs/datavalue>
property>
<property>
<name>dfs.namenode.secondary.http-addressname>
<value>localhost:9001value>
property>
<property>
<name>dfs.webhdfs.enabledname>
<value>truevalue>
property>
configuration>
3.yarn配置
mapred-site.xml
<configuration>
<property>
<name>mapreduce.framework.namename>
<value>yarnvalue>
property>
<property>
<name>mapreduce.admin.user.envname>
<value>HADOOP_MAPRED_HOME=$HADOOP_COMMON_HOMEvalue>
property>
<property>
<name>yarn.app.mapreduce.am.envname>
<value>HADOOP_MAPRED_HOME=$HADOOP_COMMON_HOMEvalue>
property>
<property>
<name>mapreduce.application.classpathname>
<value>
/Users/chenxun/software/hadoop-2.8.1/etc/hadoop,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/common/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/common/lib/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/hdfs/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/hdfs/lib/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/mapreduce/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/mapreduce/lib/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/yarn/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/yarn/lib/*
value>
property>
configuration>
yarn-site.xml
<configuration>
<property>
<name>yarn.nodemanager.aux-servicesname>
<value>mapreduce_shufflevalue>
property>
<property>
<name>yarn.application.classpathname>
<value>
/Users/chenxun/software/hadoop-2.8.1/etc/hadoop,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/common/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/common/lib/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/hdfs/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/hdfs/lib/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/mapreduce/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/mapreduce/lib/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/yarn/*,
/Users/chenxun/software/hadoop-2.8.1/share/hadoop/yarn/lib/*
value>
property>
configuration>
四、配置hadoop环境变量,可能需要重新编译native library
vim ~/.bash_profile
export HADOOP_HOME=/Users/chenxun/software/hadoop-2.8.1
export PATH=$PATH:$HADOOP_HOME/sbin:$HADOOP_HOME/bin
export LD_LIBRARY_PATH=$HADOOP_HOME/lib/native/
export HADOOP_COMMON_LIB_NATIVE_DIR=$HADOOP_HOME/lib/native
export HADOOP_OPTS="-Djava.library.path=$HADOOP_HOME/lib/native:$HADOOP_COMMON_LIB_NATIVE_DIR"
**上面配置基本完成,一般没什么问题了。可以开始运行你的hadoop了,直接看第五步好了,有问题继续往下看。
下面步骤在你运行hadoop的过程可能碰到,如果你碰到了再回头看这,:
在mac可能还要重新编译hadoop的运行用到动态库;这个时候肯能要配置maven环境,根据自己的maven安装路径配置maven环境,再重新编译/lib/native下动态要用到
如果你在运行出现下面的情况就需要重新编译动态库:
WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform… using builtin-java classes where applicable
参考链接:http://blog.csdn.net/tterminator/article/details/51779689
参考二:Mac下 hadoop-2.7.0编译过程记录http://blog.csdn.net/beijihukk/article/details/53782508?utm_medium=referral&utm_source=itdadao
先下载好hadoop源码:hadoop-2.8.1-src
重新编译之前需要安装
protobuf
和maven以及cmake,自己百度搜索安装即可
protobuf安装方法如下:
1. protobuf 2.5版本的brew安装方法
$ brew search protobuf
protobuf protobuf-c protobuf-swift
homebrew/php/php53-protobuf homebrew/php/php56-protobuf
homebrew/php/php54-protobuf homebrew/versions/protobuf250
homebrew/php/php55-protobuf homebrew/versions/protobuf260
$ brew install homebrew/versions/protobuf250
配置maven环境
export M2_HOME=/Users/chenxun/software/apache-maven-3.5.0
编译动态库过程出现下面的错误可能是zlib没安装
也可能是openssl没设置好,设置openssl的方法如下(我的mac,你根据你自己的openssl安装路径设置)
export OPENSSL_ROOT_DIR=/usr/local/Cellar/openssl/1.0.2k
export OPENSSL_INCLUDE_DIR=/usr/local/Cellar/openssl/1.0.2k/include
**开始编译native库,命令如下:
mvn package -Pdist,native -DskipTests -Dtar**
将编译出的native library复制到下载的二进制版本的hadoop-2.8.1相应目录中
编译出的native library库的位置为
hadoop-2.8.1-src/hadoop-dist/target/hadoop-2.8.1/lib/native
拷贝到二进制版本的hadoop-2.8.1的目录
hadoop-2.8.1/lib/native
修改/etc/hadoop/hadoop-env.sh配置
export HADOOP_OPTS=”$HADOOP_OPTS -Djava.net.preferIPv4Stack=true -Djava.library.path=/hadoop-2.6.0/lib/native”
重新启动Hadoop
此时就不会出现本文开头处的那个警告了。
五、运行hadoop 以及hadoop命令简单的介绍
格式化HDFS
hdfs namenode -format
一次启动hadoop所有进程:
start-all.sh
打开 http://localhost:50070 进入hdfs管理页面
打开 http://localhost:8088 进入hadoop进程管理页面
hadoop命令简单介绍:
创建目录
hdfs dfs -mkdir -p /user/chenxun/input
查看:
hadoop fs -ls /user/chenxun/
先自己随便建一个file.txt文件,在文件随便写点东西,如下:
vim file.txt
hadoop
chenxun
hadoop
chen
把这个文件用命令上传你刚才建文件input下面:命令如下:
hdfs dfs -put ./file.txt input
用命令查看一下是否成功:如果成功你在input下面看到file.txt文件
hadoop fs -ls /user/chenxun/input
这个时候你测试hadoop提供的wordcount例子:执行下面的语句(如果出现错误不要着急往文章后面看解决方法)
hadoop jar share/hadoop/mapreduce/hadoop-mapreduce-examples-2.8.1.jar wordcount /user/chenxun/input/file.txt output
运行过程如下
17/10/14 01:55:26 INFO client.RMProxy: Connecting to ResourceManager at /0.0.0.0:8032
17/10/14 01:55:28 INFO input.FileInputFormat: Total input files to process : 1
17/10/14 01:55:28 INFO mapreduce.JobSubmitter: number of splits:1
17/10/14 01:55:28 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1507912953622_0017
17/10/14 01:55:28 INFO impl.YarnClientImpl: Submitted application application_1507912953622_0017
17/10/14 01:55:28 INFO mapreduce.Job: The url to track the job: http://chen.local:8088/proxy/application_1507912953622_0017/
17/10/14 01:55:28 INFO mapreduce.Job: Running job: job_1507912953622_0017
17/10/14 01:55:36 INFO mapreduce.Job: Job job_1507912953622_0017 running in uber mode : false
17/10/14 01:55:36 INFO mapreduce.Job: map 0% reduce 0%
17/10/14 01:55:41 INFO mapreduce.Job: map 100% reduce 0%
17/10/14 01:55:47 INFO mapreduce.Job: map 100% reduce 100%
17/10/14 01:55:47 INFO mapreduce.Job: Job job_1507912953622_0017 completed successfully
17/10/14 01:55:48 INFO mapreduce.Job: Counters: 49
File System Counters
FILE: Number of bytes read=44
FILE: Number of bytes written=276523
FILE: Number of read operations=0
FILE: Number of large read operations=0
FILE: Number of write operations=0
HDFS: Number of bytes read=141
HDFS: Number of bytes written=26
HDFS: Number of read operations=6
HDFS: Number of large read operations=0
HDFS: Number of write operations=2
Job Counters
Launched map tasks=1
Launched reduce tasks=1
Data-local map tasks=1
Total time spent by all maps in occupied slots (ms)=3032
Total time spent by all reduces in occupied slots (ms)=3133
Total time spent by all map tasks (ms)=3032
Total time spent by all reduce tasks (ms)=3133
Total vcore-milliseconds taken by all map tasks=3032
Total vcore-milliseconds taken by all reduce tasks=3133
Total megabyte-milliseconds taken by all map tasks=3104768
Total megabyte-milliseconds taken by all reduce tasks=3208192
Map-Reduce Framework
Map input records=4
Map output records=4
Map output bytes=43
Map output materialized bytes=44
Input split bytes=114
Combine input records=4
Combine output records=3
Reduce input groups=3
Reduce shuffle bytes=44
Reduce input records=3
Reduce output records=3
Spilled Records=6
Shuffled Maps =1
Failed Shuffles=0
Merged Map outputs=1
GC time elapsed (ms)=134
CPU time spent (ms)=0
Physical memory (bytes) snapshot=0
Virtual memory (bytes) snapshot=0
Total committed heap usage (bytes)=306708480
Shuffle Errors
BAD_ID=0
CONNECTION=0
IO_ERROR=0
WRONG_LENGTH=0
WRONG_MAP=0
WRONG_REDUCE=0
File Input Format Counters
Bytes Read=27
File Output Format Counters
Bytes Written=26
再执行看看执行结果:
hadoop fs -ls /user/chenxun/output
这时候你看到:
Found 2 items
-rw-r--r-- 1 chenxun supergroup 0 2017-10-14 01:55 /user/chenxun/output/_SUCCESS
-rw-r--r-- 1 chenxun supergroup 26 2017-10-14 01:55 /user/chenxun/output/part-r-00000
接下来真正看一下你的单词统计结果:
hadoop fs -cat /user/chenxun/output/part-r-00000
因为我的file.txt文件内容是
[chenxun@chen.local 11:13 ~/software/hadoop-2.8.1]$cat file.txt
hadoop
chenxun
hadoop
chen
所有统计结果如下:
chen 1
chenxun 1
hadoop 2
Exit code: 127 Stack trace: ExitCodeException exitCode=127: 的错误解决方法如下:
六、如果你在第五步执行过程出现错误:Exit code: 127 Stack trace: ExitCodeException exitCode=127: 不要着急这里了解决方法:
今天在Mac配置伪分布式环境。部署完毕后,运行MapReduce程序,程序运行刚开始没有问题, 但是到启动Map Task时,就报错误,报exitCode: 127错误。
错误日志如下:
15/04/06 00:08:01 INFO mapreduce.Job: Job job_1428250045856_0002 failed with state FAILED due to: Application application_1428250045856_0002 failed 2 times due to AM Container for appattempt_1428250045856_0002_000002 exited with exitCode: 127 due to: Exception from container-launch: org.apache.hadoop.util.Shell$ExitCodeException:
org.apache.hadoop.util.Shell$ExitCodeException:
at org.apache.hadoop.util.Shell.runCommand(Shell.java:505)
at org.apache.hadoop.util.Shell.run(Shell.java:418)
at org.apache.hadoop.util.Shell$ShellCommandExecutor.execute(Shell.java:650)
解决办法:
1 从错误日志上看应该是某个Shell命令在Mac上没有找到。
2 查看每个程序的日志:hadoop-2.3.0/logs/userlogs/application_1428247759749_0002/container_1428247759749_0002_02_000001中的错误日志:
/bin/bash: /bin/java: No such file or directory
可以看到:/bin/java 的命令没有找到。
建立一个软链接,链接到java程序即可。
上文中的日志文件和异常中的文件名对应不上,这个是我编写时的问题, 实际可以根据日志中的文件名找到具体的错误文件。但是整个流程,是这个。
是因为Hadoop默认检查/bin/java路径下的java,可是Mac的Java不是装这里的,它的路径是/usr/bin/java。你不信啊,那你输入命令$ /usr/bin/java -version, 看看是不是会出现类似以下的信息:
Java version "1.8.0_45"
Java(TM) SE Runtime Environment (build 1.8.0_45-b14)
Java Hotspot(TM) 64-Bit Server VM (build 25.45-b02, mixed mode)
如果对了的话,那么下一步就是在/bin/java创建一个快捷方式,让Hadoop读到/usr/bin/java里的信息:
$ sudo ln -s /usr/bin/java /bin/java
再输入命令$ /bin/java -version验证一遍,显示:
引用
Java version "1.8.0_45"
Java(TM) SE Runtime Environment (build 1.8.0_45-b14)
Java Hotspot(TM) 64-Bit Server VM (build 25.45-b02, mixed mode)
这个时候你在执行sudo ln -s /usr/bin/java /bin/java出现问题:ln: /bin/java: Operation not permitted
这是mac系统的原因不是你的错。
这是因为苹果在OS X 10.11中引入的SIP特性使得即使加了sudo(也就是具有root权限)也无法修改系统级的目录,其中就包括了/usr/bin。要解决这个问题有两种做法:一种是比较不安全的就是关闭SIP,也就是rootless特性;另一种是将本要链接到/usr/bin下的改链接到/usr/local/bin下就好了。
废话不多说上解决方法:
重启按住 Command+R,进入恢复模式,打开Terminal
csrutil disable
重启即可。如果要恢复默认,最好这么做,那么
csrutil enable
参考链接:http://www.jianshu.com/p/22b89f19afd6