1、分发HDFS文件（-cacheFile）

需求：wordcount（只统计指定的单词），但是该文件非常大,可以先将该文件上传到hdfs，通过-cacheFile的方式进行分发；

-cachefile hdfs://host:port/path/to/file#linkname #选项在计算节点上缓存文件，streaming程序通过./linkname的方式访问文件。

思路：mapper和reducer程序都不需要修改，只是在运行streaming的时候需要使用-cacheFile 指定hdfs上的文件；

1.1、streaming命令格式

$HADOOP_HOME/bin/hadoop jar hadoop-streaming.jar \
    -jobconf mapred.job.name="streaming_wordcount" \
    -jobconf mapred.job.priority=3 \
    -input /input/ \
    -output /output/ \
    -mapper "python mapper.py whc" \
    -reducer "python reducer.py" \
    -cacheFile "hdfs://master:9000/cache_file/wordwhite#whc"
    -file ./mapper.py \
    -file ./reducer.py 

注：-cacheFile "hdfs://master:9000/cache_file/wordwhite#whc" whc表示在hdfs上该文件的别名，在-mapper "python mapper.py whc"就如同使用本地文件一样。

1.2、上传wordwhite

$ hadoop fs -mkdir /input/cachefile
$ hadoop fs -put wordwhite  /input/cachefile
$ hadoop fs -ls /input/cachefile
Found 1 items
-rw-r--r--   1 hadoop supergroup         12 2018-01-26 15:02 /input/cachefile/wordwhite
$ hadoop fs -text hdfs://localhost:9000/input/cachefile/wordwhite
the
and
had

1.3 run_streaming程序

mapper和reducer程序参考本地分发实例

$ vim runstreaming_cachefile.sh 

#!/bin/bash

HADOOP_CMD="/home/hadoop/app/hadoop/hadoop-2.6.0-cdh5.13.0/bin/hadoop"
STREAM_JAR_PATH="/home/hadoop/app/hadoop/hadoop-2.6.0-cdh5.13.0/share/hadoop/tools/lib/hadoop-streaming-2.6.0-cdh5.13.0.jar"

INPUT_FILE_PATH="/input/The_Man_of_Property"
OUTPUT_FILE_PATH="/output/wordcount/wordwhitecachefiletest"

$HADOOP_CMD jar $STREAM_JAR_PATH \
                -input $INPUT_FILE_PATH \
                -output $OUTPUT_FILE_PATH \
                -jobconf "mapred.job.name=wordcount_wordwhite_cachefile_demo" \
                -mapper "python mapper.py WHF" \
                -reducer "python reducer.py" \
                -cacheFile "hdfs://localhost:9000/input/cachefile/wordwhite#WHF" \
                -file ./mapper.py \
                -file ./reducer.py

1.4、执行程序

$ ./runstreaming_cachefile.sh 
18/01/26 15:38:27 WARN streaming.StreamJob: -file option is deprecated, please use generic option -files instead.
18/01/26 15:38:28 WARN streaming.StreamJob: -cacheFile option is deprecated, please use -files instead.
18/01/26 15:38:28 WARN streaming.StreamJob: -jobconf option is deprecated, please use -D instead.
18/01/26 15:38:28 INFO Configuration.deprecation: mapred.job.name is deprecated. Instead, use mapreduce.job.name
packageJobJar: [./mapper.py, ./reducer.py, /tmp/hadoop-unjar1709565523181962236/] [] /tmp/streamjob6164905989972408041.jar tmpDir=null
18/01/26 15:38:29 INFO client.RMProxy: Connecting to ResourceManager at /0.0.0.0:8032
18/01/26 15:38:29 INFO client.RMProxy: Connecting to ResourceManager at /0.0.0.0:8032
18/01/26 15:38:31 INFO mapred.FileInputFormat: Total input paths to process : 1
18/01/26 15:38:31 INFO mapreduce.JobSubmitter: number of splits:2
18/01/26 15:38:32 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1516345010544_0012
18/01/26 15:38:32 INFO impl.YarnClientImpl: Submitted application application_1516345010544_0012
18/01/26 15:38:32 INFO mapreduce.Job: The url to track the job: http://localhost:8088/proxy/application_1516345010544_0012/
18/01/26 15:38:32 INFO mapreduce.Job: Running job: job_1516345010544_0012
18/01/26 15:38:40 INFO mapreduce.Job: Job job_1516345010544_0012 running in uber mode : false
18/01/26 15:38:40 INFO mapreduce.Job:  map 0% reduce 0%
18/01/26 15:38:49 INFO mapreduce.Job:  map 50% reduce 0%
18/01/26 15:38:50 INFO mapreduce.Job:  map 100% reduce 0%
18/01/26 15:38:57 INFO mapreduce.Job:  map 100% reduce 100%
18/01/26 15:38:57 INFO mapreduce.Job: Job job_1516345010544_0012 completed successfully
18/01/26 15:38:57 INFO mapreduce.Job: Counters: 49
    File System Counters
        FILE: Number of bytes read=73950
        FILE: Number of bytes written=582590
        FILE: Number of read operations=0
        FILE: Number of large read operations=0
        FILE: Number of write operations=0
        HDFS: Number of bytes read=636501
        HDFS: Number of bytes written=27
        HDFS: Number of read operations=9
        HDFS: Number of large read operations=0
        HDFS: Number of write operations=2
    Job Counters 
        Launched map tasks=2
        Launched reduce tasks=1
        Data-local map tasks=2
        Total time spent by all maps in occupied slots (ms)=12921
        Total time spent by all reduces in occupied slots (ms)=5641
        Total time spent by all map tasks (ms)=12921
        Total time spent by all reduce tasks (ms)=5641
        Total vcore-milliseconds taken by all map tasks=12921
        Total vcore-milliseconds taken by all reduce tasks=5641
        Total megabyte-milliseconds taken by all map tasks=13231104
        Total megabyte-milliseconds taken by all reduce tasks=5776384
    Map-Reduce Framework
        Map input records=2866
        Map output records=9243
        Map output bytes=55458
        Map output materialized bytes=73956
        Input split bytes=198
        Combine input records=0
        Combine output records=0
        Reduce input groups=3
        Reduce shuffle bytes=73956
        Reduce input records=9243
        Reduce output records=3
        Spilled Records=18486
        Shuffled Maps =2
        Failed Shuffles=0
        Merged Map outputs=2
        GC time elapsed (ms)=360
        CPU time spent (ms)=3910
        Physical memory (bytes) snapshot=719896576
        Virtual memory (bytes) snapshot=8331550720
        Total committed heap usage (bytes)=602931200
    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=636303
    File Output Format Counters 
        Bytes Written=27
18/01/26 15:38:57 INFO streaming.StreamJob: Output directory: /output/wordcount/wordwhitecachefiletest

1.5、查看结果

$ hadoop fs -ls /output/wordcount/wordwhitecachefiletest
Found 2 items
-rw-r--r--   1 hadoop supergroup          0 2018-01-26 15:38 /output/wordcount/wordwhitecachefiletest/_SUCCESS
-rw-r--r--   1 hadoop supergroup         27 2018-01-26 15:38 /output/wordcount/wordwhitecachefiletest/part-00000

$ hadoop fs -text /output/wordcount/wordwhitecachefiletest/part-00000
and 2573
had 1526
the 5144

以上就完成了分发HDFS上的文件并指定单词的wordcount.

2、hadoop streaming 语法参考

• http://blog.51cto.com/balich/2065419