大数据架构师，带你深入理解HadoopStreaming原理及实践，快来学

大数据架构师 06-02 1589

前言：

眼前姐妹们对“ubuntu搭建jsp”大致比较着重，咱们都需要剖析一些“ubuntu搭建jsp”的相关知识。那么小编同时在网上搜集了一些有关“ubuntu搭建jsp””的相关资讯，希望朋友们能喜欢，各位老铁们快快来学习一下吧！

前言

Hadoop Streaming提供了一个便于进行MapReduce编程的工具包，使用它可以基于一些可执行命令、脚本语言或其他编程语言来实现Mapper和 Reducer，从而充分利用Hadoop并行计算框架的优势和能力，来处理大数据。需要注意的是，Streaming方式是基于Unix系统的标准输入输出来进行MapReduce Job的运行，它区别与Pipes的地方主要是通信协议，Pipes使用的是Socket通信，是对使用C++语言来实现MapReduce Job并通过Socket通信来与Hadopp平台通信，完成Job的执行。任何支持标准输入输出特性的编程语言都可以使用Streaming方式来实现MapReduce Job，基本原理就是输入从Unix系统标准输入，输出使用Unix系统的标准输出。Hadoop是使用Java语言编写的，所以最直接的方式的就是使用Java语言来实现Mapper和Reducer，然后配置MapReduce Job，提交到集群计算环境来完成计算。但是很多开发者可能对Java并不熟悉，而是对一些具有脚本特性的语言，如C++、Shell、Python、 Ruby、PHP、Perl有实际开发经验，Hadoop Streaming为这一类开发者提供了使用Hadoop集群来进行处理数据的工具，即工具包hadoop-streaming-.jar。

Hadoop Streaming使用了Unix的标准输入输出作为Hadoop和其他编程语言的开发接口，因此在其他的编程语言所写的程序中，只需要将标准输入作为程序的输入，将标准输出作为程序的输出就可以了。在标准的输入输出中，Key和Value是以Tab作为分隔符，并且在Reducer的标准输入中，Hadoop框架保证了输入的数据是经过了按Key排序的。

如何使用Hadoop Streaming工具呢？我们可以查看该工具的使用方法，通过命令行来获取，如下所示：

xiaoxiang@ubuntu3:~/hadoop$ bin/hadoop jar ./contrib/streaming/hadoop-streaming-1.0.3.jar -infoUsage: $HADOOP_HOME/bin/hadoop jar $HADOOP_HOME/hadoop-streaming.jar [options]Options:  -input    <path>     DFS input file(s) for the Map step  -output   <path>     DFS output directory for the Reduce step  -mapper   <cmd|JavaClassName>      The streaming command to run  -combiner <cmd|JavaClassName> The streaming command to run  -reducer  <cmd|JavaClassName>      The streaming command to run  -file     <file>     File/dir to be shipped in the Job jar file  -inputformat TextInputFormat(default)|SequenceFileAsTextInputFormat|JavaClassName Optional.  -outputformat TextOutputFormat(default)|JavaClassName  Optional.  -partitioner JavaClassName  Optional.  -numReduceTasks <num>  Optional.  -inputreader <spec>  Optional.  -cmdenv   <n>=<v>    Optional. Pass env.var to streaming commands  -mapdebug <path>  Optional. To run this script when a map task fails  -reducedebug <path>  Optional. To run this script when a reduce task fails  -io <identifier>  Optional.  -verbose Generic options supported are-conf <configuration file>     specify an application configuration file-D <property=value>            use value for given property-fs <local|namenode:port>      specify a namenode-jt <local|jobtracker:port>    specify a job tracker-files <comma separated list of files>    specify comma separated files to be copied to the map reduce cluster-libjars <comma separated list of jars>    specify comma separated jar files to include in the classpath.-archives <comma separated list of archives>    specify comma separated archives to be unarchived on the compute machines. The general command line syntax isbin/hadoop command [genericOptions] [commandOptions]  In -input: globbing on <path> is supported and can have multiple -inputDefault Map input format: a line is a record in UTF-8  the key part ends at first TAB, the rest of the line is the valueCustom input format: -inputformat package.MyInputFormatMap output format, reduce input/output format:  Format defined by what the mapper command outputs. Line-oriented The files named in the -file argument[s] end up in the  working directory when the mapper and reducer are run.  The location of this working directory is unspecified. To set the number of reduce tasks (num. of output files):  -D mapred.reduce.tasks=10To skip the sort/combine/shuffle/sort/reduce step:  Use -numReduceTasks 0  A Task's Map output then becomes a 'side-effect output' rather than a reduce input  This speeds up processing, This also feels more like "in-place" processing  because the input filename and the map input order are preserved  This equivalent -reducer NONE To speed up the last maps:  -D mapred.map.tasks.speculative.execution=trueTo speed up the last reduces:  -D mapred.reduce.tasks.speculative.execution=trueTo name the job (appears in the JobTracker Web UI):  -D mapred.job.name='My Job'To change the local temp directory:  -D dfs.data.dir=/tmp/dfs  -D stream.tmpdir=/tmp/streamingAdditional local temp directories with -cluster local:  -D mapred.local.dir=/tmp/local  -D mapred.system.dir=/tmp/system  -D mapred.temp.dir=/tmp/tempTo treat tasks with non-zero exit status as SUCCEDED:  -D stream.non.zero.exit.is.failure=falseUse a custom hadoopStreaming build along a standard hadoop install:  $HADOOP_HOME/bin/hadoop jar /path/my-hadoop-streaming.jar [...]\    [...] -D stream.shipped.hadoopstreaming=/path/my-hadoop-streaming.jarFor more details about jobconf parameters see: ;To set an environement variable in a streaming command:   -cmdenv EXAMPLE_DIR=/home/example/dictionaries/ Shortcut:   setenv HSTREAMING "$HADOOP_HOME/bin/hadoop jar $HADOOP_HOME/hadoop-streaming.jar" Example: $HSTREAMING -mapper "/usr/local/bin/perl5 filter.pl"           -file /local/filter.pl -input "/logs/0604*/*" [...]  Ships a script, invokes the non-shipped perl interpreter  Shipped files go to the working directory so filter.pl is found by perl  Input files are all the daily logs for days in month 2006-04

面，我们分别选择几个可以使用Hadoop Streaming工具来进行计算的例子，比如对单词词频进行统计计算，即WordCount功能。首先，我们准备测试使用的数据集，如下所示：

xiaoxiang@ubuntu3:~/hadoop$ bin/hadoop fs -lsr /user/xiaoxiang/dataset/join/-rw-r--r--   3 xiaoxiang supergroup  391103029 2013-03-26 12:19 /user/xiaoxiang/dataset/join/irc_basic_info.ds-rw-r--r--   3 xiaoxiang supergroup   11577164 2013-03-26 12:20 /user/xiaoxiang/dataset/join/irc_net_block.ds-rw-r--r--   3 xiaoxiang supergroup    8335235 2013-03-26 12:20 /user/xiaoxiang/dataset/join/irc_org_info.ds

一共有3个数据文件，大概将近400M大小。下面，选择Python语言来实现MapReduce Job的运行。使用Python实现Mapper，代码文件为word_count_mapper.py，代码如下所示：

#!/usr/bin/env python import sys for line in sys.stdin:    line = line.strip()    words = filter(lambda word: word, line.split())    for word in words:        print '%s\t%s' % (word, 1)

使用Python实现Reducer，代码文件为word_count_reducer.py，代码如下所示：

#!/usr/bin/env python import sysfrom operator import itemgetter wc_dict = {} for line in sys.stdin:    line = line.strip()    word, count = line.split()    try:        count = int(count)        wc_dict[word] = wc_dict.get(word, 0) + count    except ValueError:        pass sorted_dict = sorted(wc_dict.items(), key=itemgetter(0))for word, count in sorted_dict:    print '%s\t%s' % (word, count)

运行Python实现的MapReduce程序，如下所示：

xiaoxiang@ubuntu3:~/hadoop$ bin/hadoop jar ./contrib/streaming/hadoop-streaming-1.0.3.jar -input /user/xiaoxiang/dataset/join/ -output /user/xiaoxiang/output/streaming/python -mapper word_count_mapper.py -reducer word_count_reducer.py -numReduceTasks 2 -file *.pypackageJobJar: [word_count_mapper.py, word_count_reducer.py, /opt/stone/cloud/storage/tmp/hadoop-xiaoxiang/hadoop-unjar4066863202997744310/] [] /tmp/streamjob2336302975421423718.jar tmpDir=null13/04/18 17:50:17 INFO util.NativeCodeLoader: Loaded the native-hadoop library13/04/18 17:50:17 WARN snappy.LoadSnappy: Snappy native library not loaded13/04/18 17:50:17 INFO mapred.FileInputFormat: Total input paths to process : 313/04/18 17:50:17 INFO streaming.StreamJob: getLocalDirs(): [/opt/stone/cloud/storage/mapred/local]13/04/18 17:50:17 INFO streaming.StreamJob: Running job: job_201303302227_004713/04/18 17:50:17 INFO streaming.StreamJob: To kill this job, run:13/04/18 17:50:17 INFO streaming.StreamJob: /opt/stone/cloud/hadoop-1.0.3/libexec/../bin/hadoop job  -Dmapred.job.tracker=hdfs://ubuntu3:9001/ -kill job_201303302227_004713/04/18 17:50:17 INFO streaming.StreamJob: Tracking URL:  17:50:18 INFO streaming.StreamJob:  map 0%  reduce 0%13/04/18 17:50:33 INFO streaming.StreamJob:  map 7%  reduce 0%13/04/18 17:50:36 INFO streaming.StreamJob:  map 11%  reduce 0%13/04/18 17:50:39 INFO streaming.StreamJob:  map 15%  reduce 0%13/04/18 17:50:42 INFO streaming.StreamJob:  map 19%  reduce 0%13/04/18 17:50:45 INFO streaming.StreamJob:  map 23%  reduce 0%13/04/18 17:50:48 INFO streaming.StreamJob:  map 25%  reduce 0%13/04/18 17:51:09 INFO streaming.StreamJob:  map 32%  reduce 2%13/04/18 17:51:12 INFO streaming.StreamJob:  map 36%  reduce 4%13/04/18 17:51:15 INFO streaming.StreamJob:  map 40%  reduce 8%13/04/18 17:51:18 INFO streaming.StreamJob:  map 44%  reduce 8%13/04/18 17:51:21 INFO streaming.StreamJob:  map 47%  reduce 8%13/04/18 17:51:24 INFO streaming.StreamJob:  map 50%  reduce 8%13/04/18 17:51:45 INFO streaming.StreamJob:  map 54%  reduce 10%13/04/18 17:51:48 INFO streaming.StreamJob:  map 60%  reduce 15%13/04/18 17:51:51 INFO streaming.StreamJob:  map 65%  reduce 17%13/04/18 17:51:55 INFO streaming.StreamJob:  map 66%  reduce 17%13/04/18 17:51:58 INFO streaming.StreamJob:  map 68%  reduce 17%13/04/18 17:52:01 INFO streaming.StreamJob:  map 72%  reduce 17%13/04/18 17:52:04 INFO streaming.StreamJob:  map 75%  reduce 17%13/04/18 17:52:19 INFO streaming.StreamJob:  map 75%  reduce 19%13/04/18 17:52:22 INFO streaming.StreamJob:  map 87%  reduce 21%13/04/18 17:52:25 INFO streaming.StreamJob:  map 100%  reduce 23%13/04/18 17:52:28 INFO streaming.StreamJob:  map 100%  reduce 27%13/04/18 17:52:31 INFO streaming.StreamJob:  map 100%  reduce 29%13/04/18 17:52:37 INFO streaming.StreamJob:  map 100%  reduce 49%13/04/18 17:52:40 INFO streaming.StreamJob:  map 100%  reduce 69%13/04/18 17:52:43 INFO streaming.StreamJob:  map 100%  reduce 72%13/04/18 17:52:46 INFO streaming.StreamJob:  map 100%  reduce 74%13/04/18 17:52:49 INFO streaming.StreamJob:  map 100%  reduce 76%13/04/18 17:52:52 INFO streaming.StreamJob:  map 100%  reduce 78%13/04/18 17:52:55 INFO streaming.StreamJob:  map 100%  reduce 79%13/04/18 17:52:58 INFO streaming.StreamJob:  map 100%  reduce 81%13/04/18 17:53:01 INFO streaming.StreamJob:  map 100%  reduce 84%13/04/18 17:53:04 INFO streaming.StreamJob:  map 100%  reduce 87%13/04/18 17:53:07 INFO streaming.StreamJob:  map 100%  reduce 90%13/04/18 17:53:10 INFO streaming.StreamJob:  map 100%  reduce 93%13/04/18 17:53:13 INFO streaming.StreamJob:  map 100%  reduce 96%13/04/18 17:53:16 INFO streaming.StreamJob:  map 100%  reduce 98%13/04/18 17:53:19 INFO streaming.StreamJob:  map 100%  reduce 99%13/04/18 17:53:22 INFO streaming.StreamJob:  map 100%  reduce 100%13/04/18 17:54:10 INFO streaming.StreamJob: Job complete: job_201303302227_004713/04/18 17:54:10 INFO streaming.StreamJob: Output: /user/xiaoxiang/output/streaming/python

验证结果，如下所示：

xiaoxiang@ubuntu3:~/hadoop$ bin/hadoop fs -lsr /user/xiaoxiang/output/streaming/python-rw-r--r--   3 xiaoxiang supergroup          0 2013-04-18 17:54 /user/xiaoxiang/output/streaming/python/_SUCCESSdrwxr-xr-x   - xiaoxiang supergroup          0 2013-04-18 17:50 /user/xiaoxiang/output/streaming/python/_logsdrwxr-xr-x   - xiaoxiang supergroup          0 2013-04-18 17:50 /user/xiaoxiang/output/streaming/python/_logs/history-rw-r--r--   3 xiaoxiang supergroup      37646 2013-04-18 17:50 /user/xiaoxiang/output/streaming/python/_logs/history/job_201303302227_0047_1366278617842_xiaoxiang_streamjob2336302975421423718.jar-rw-r--r--   3 xiaoxiang supergroup      21656 2013-04-18 17:50 /user/xiaoxiang/output/streaming/python/_logs/history/job_201303302227_0047_conf.xml-rw-r--r--   3 xiaoxiang supergroup   91367389 2013-04-18 17:52 /user/xiaoxiang/output/streaming/python/part-00000-rw-r--r--   3 xiaoxiang supergroup   91268074 2013-04-18 17:52 /user/xiaoxiang/output/streaming/python/part-00001xiaoxiang@ubuntu3:~/hadoop$ bin/hadoop fs -cat /user/xiaoxiang/output/streaming/python/part-00000 | head -5!    2#    36#039    1#1059)    1#1098    1