1.hdfs檔案系統目錄要求（建議）

    /user
        /{username}             --使用者名稱
            /mr                 --MapReduce型別應用
                /wordcount      --應用的名稱
                    /input      --要處理的檔案
                    /output     --輸出結果的資料夾

2.按照目錄格式建立輸入資料夾（不用建立輸出資料夾）

    [root@hadoop-yarn hadoop-2.6
 
.5]# bin/hdfs dfs -mkdir -p /user/root/mr/wordcount/input

3.手動建立測試資料

    cyhp supergroup
    hadoop hello
    hdfs hadoop
    mapreduce hadoop
    world yarn

4.將輸入資料放入HDFS中

[root@hadoop-yarn hadoop-2.6.5]# bin/hdfs dfs -put test-datas/wc.input  /user/root/mr/wordcount/input/

5.使用yarn執行示例

[root@hadoop-yarn hadoop-2.6
 
.5]# bin/yarn jar share/hadoop/mapreduce/hadoop-mapreduce-examples-2.6.5.jar wordcount /user/root/mr/wordcount/input/ /user/root/mr/wordcount/output/

6.檢視輸出

[root@hadoop-yarn hadoop-2.6.5]# bin/hdfs dfs -text /user/root/mr/wordcount/output/par*

[root@hadoop-yarn hadoop-2.6.5]# bin/hdfs dfs -text /user/root/mr/wordcount/output/par*
 

cyhp    1
hadoop  3
hdfs    1
hello   1
mapreduce   1
supergroup  1
world   1
yarn    1