一、获取数据
搜狗实验室为我们提供了用户使用搜狗搜索引擎查询的日志,下载地址为
http://download.labs.sogou.com/dl/q.html
本文选择下载精简版。
数据格式如下:
二、上传数据至HDFS
建立hdfs目录
[email protected]:~# hdfs dfs -mkdir -p /library/sougou
上传文件
[email protected]:~# hdfs dfs -put SogouQ1.txt /library/sougou [email protected]:~# hdfs dfs -put SogouQ2.txt /library/sougou [email protected]:~# hdfs dfs -put SogouQ3.txt /library/sougou [email protected]:~#
三、使用Hive创建表
[email protected]:/usr/local/hive/apache-hive-1.2.1/bin# ./hive SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/local/hadoop/hadoop-2.6.0/share/hadoop/common/lib/slf4j-log4j12-1.7.5.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/local/spark/spark-1.6.0-bin-hadoop2.6/lib/spark-assembly-1.6.0-hadoop2.6.0.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation. SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory] SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/local/hadoop/hadoop-2.6.0/share/hadoop/common/lib/slf4j-log4j12-1.7.5.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/local/spark/spark-1.6.0-bin-hadoop2.6/lib/spark-assembly-1.6.0-hadoop2.6.0.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation. SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory] Logging initialized using configuration in jar:file:/usr/local/hive/apache-hive-1.2.1/lib/hive-common-1.2.1.jar!/hive-log4j.properties hive> CREATE TABLE SOUGOU(ID STRING,WEBSESSION STRING,WORD STRING,S_SEQ INT,C_SEQ INT ,WEBSITE STRING) > ROW FORMAT DELIMITED FIELDS TERMINATED BY ‘\t‘ LINES TERMINATED BY ‘\n‘; OK Time taken: 1.995 seconds hive>
四、加载数据
hive> LOAD DATA INPATH ‘/library/sougou/SogouQ1.txt‘ INTO TABLE sougou; Loading data to table default.sougou Table default.sougou stats: [numFiles=1, totalSize=108750574] OK Time taken: 1.101 seconds
此时,我们再次查看源目录
SogouQ1.txt 已经没有啦,该文件跑哪里去了呢?
可见,导入数据其实就是将HDFS上的文件移动一个位置而已。所以速度是非常的快。
那可不可以直接将SogouQ1.txt放置在HDFS的/user/hive/warehouse/sougou/中,而不使用LOAD语句?
因为元数据要知道该表中包含了哪些数据文件,所以必须使用load语句。
五、操作数据
5.1 计算count
hive> select count(*) from sougou; Query ID = root_20160314192407_792e772a-c225-4b37-b948-7050d6b529b4 Total jobs = 1 Launching Job 1 out of 1 Number of reduce tasks determined at compile time: 1 In order to change the average load for a reducer (in bytes): set hive.exec.reducers.bytes.per.reducer=<number> In order to limit the maximum number of reducers: set hive.exec.reducers.max=<number> In order to set a constant number of reducers: set mapreduce.job.reduces=<number> Starting Job = job_1457942575478_0002, Tracking URL = http://spark-master:8088/proxy/application_1457942575478_0002/ Kill Command = /usr/local/hadoop/hadoop-2.6.0/bin/hadoop job -kill job_1457942575478_0002 Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 1 2016-03-14 19:24:29,014 Stage-1 map = 0%, reduce = 0% 2016-03-14 19:24:47,137 Stage-1 map = 100%, reduce = 0%, Cumulative CPU 6.14 sec 2016-03-14 19:25:04,563 Stage-1 map = 100%, reduce = 100%, Cumulative CPU 9.38 sec MapReduce Total cumulative CPU time: 9 seconds 380 msec Ended Job = job_1457942575478_0002 MapReduce Jobs Launched: Stage-Stage-1: Map: 1 Reduce: 1 Cumulative CPU: 9.38 sec HDFS Read: 108757501 HDFS Write: 8 SUCCESS Total MapReduce CPU Time Spent: 9 seconds 380 msec OK 1000000 Time taken: 58.603 seconds, Fetched: 1 row(s)
5.2 查看数据
hive> select * from sougou limit 5; OK 20111230000005 57375476989eea12893c0c3811607bcf 1 1 http://www.qiyi.com/ 20111230000005 66c5bb7774e31d0a22278249b26bc83a 3 1 http://www.booksky.org/BookDetail.aspx?BookID=1050804&Level=1 20111230000007 b97920521c78de70ac38e3713f524b50 1 1 http://www.bblianmeng.com/ 20111230000008 6961d0c97fe93701fc9c0d861d096cd9 1 1 http://lib.scnu.edu.cn/ 20111230000008 f2f5a21c764aebde1e8afcc2871e086f 2 1 http://proxyie.cn/ Time taken: 0.246 seconds, Fetched: 5 row(s)
这里出现了乱码,原因是源文件是gb3212编码,但是Hadoop和Hive都使用UTF8编码。我们将文件转码后再次上传到hdfs中
[email protected]:~# iconv -t utf-8 -f gb2312 -c SogouQ1.txt > SogouQ1.txt.utf8 [email protected]:~# rm SogouQ1.txt ; mv SogouQ1.txt.utf8 SogouQ1.txt [email protected]:~# hdfs dfs -rm /user/hive/warehouse/sougou/SogouQ1.txt 16/03/14 19:44:25 INFO fs.TrashPolicyDefault: Namenode trash configuration: Deletion interval = 0 minutes, Emptier interval = 0 minutes. Deleted /user/hive/warehouse/sougou/SogouQ1.txt [email protected]:~# hdfs dfs -put SogouQ1.txt /user/hive/warehouse/sougou/ [email protected]:~#
再次查看
hive> select * from sougou limit 5; OK 20111230000005 57375476989eea12893c0c3811607bcf 奇艺高清 1 1 http://www.qiyi.com/ 20111230000005 66c5bb7774e31d0a22278249b26bc83a 凡人修仙传 3 1 http://www.booksky.org/BookDetail.aspx?BookID=1050804&Level=1 20111230000007 b97920521c78de70ac38e3713f524b50 本本联盟 1 1 http://www.bblianmeng.com/ 20111230000008 6961d0c97fe93701fc9c0d861d096cd9 华南师范大学图书馆 1 1 http://lib.scnu.edu.cn/ 20111230000008 f2f5a21c764aebde1e8afcc2871e086f 在线代理 2 1 http://proxyie.cn/ Time taken: 0.151 seconds, Fetched: 5 row(s)
这样就正常啦。
5.2 再来一个复杂点的查询
hive> select count(*) from sougou where s_seq=1 and c_seq=1 and website like ‘%baidu%‘; Query ID = root_20160314194855_8c9aa844-e088-4695-942f-3579718962f6 Total jobs = 1 Launching Job 1 out of 1 Number of reduce tasks determined at compile time: 1 In order to change the average load for a reducer (in bytes): set hive.exec.reducers.bytes.per.reducer=<number> In order to limit the maximum number of reducers: set hive.exec.reducers.max=<number> In order to set a constant number of reducers: set mapreduce.job.reduces=<number> Starting Job = job_1457942575478_0003, Tracking URL = http://spark-master:8088/proxy/application_1457942575478_0003/ Kill Command = /usr/local/hadoop/hadoop-2.6.0/bin/hadoop job -kill job_1457942575478_0003 Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 1 2016-03-14 19:49:12,041 Stage-1 map = 0%, reduce = 0% 2016-03-14 19:49:33,174 Stage-1 map = 100%, reduce = 0%, Cumulative CPU 7.94 sec 2016-03-14 19:49:48,672 Stage-1 map = 100%, reduce = 100%, Cumulative CPU 11.55 sec MapReduce Total cumulative CPU time: 11 seconds 550 msec Ended Job = job_1457942575478_0003 MapReduce Jobs Launched: Stage-Stage-1: Map: 1 Reduce: 1 Cumulative CPU: 11.55 sec HDFS Read: 114834192 HDFS Write: 6 SUCCESS Total MapReduce CPU Time Spent: 11 seconds 550 msec OK 59090 Time taken: 55.022 seconds, Fetched: 1 row(s)
查询点击排名
hive> select word,count(*) cnt from sougou group by word order by cnt desc limit 5; Query ID = root_20160314202108_58aeca03-8ed6-4626-b15e-af6643c94107 Total jobs = 2 Launching Job 1 out of 2 Number of reduce tasks not specified. Estimated from input data size: 1 In order to change the average load for a reducer (in bytes): set hive.exec.reducers.bytes.per.reducer=<number> In order to limit the maximum number of reducers: set hive.exec.reducers.max=<number> In order to set a constant number of reducers: set mapreduce.job.reduces=<number> Starting Job = job_1457942575478_0007, Tracking URL = http://spark-master:8088/proxy/application_1457942575478_0007/ Kill Command = /usr/local/hadoop/hadoop-2.6.0/bin/hadoop job -kill job_1457942575478_0007 Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 1 2016-03-14 20:21:29,040 Stage-1 map = 0%, reduce = 0% 2016-03-14 20:21:57,425 Stage-1 map = 100%, reduce = 0%, Cumulative CPU 14.98 sec 2016-03-14 20:22:16,021 Stage-1 map = 100%, reduce = 68%, Cumulative CPU 20.27 sec 2016-03-14 20:22:19,268 Stage-1 map = 100%, reduce = 77%, Cumulative CPU 23.16 sec 2016-03-14 20:22:22,593 Stage-1 map = 100%, reduce = 93%, Cumulative CPU 25.9 sec 2016-03-14 20:22:23,721 Stage-1 map = 100%, reduce = 100%, Cumulative CPU 26.9 sec MapReduce Total cumulative CPU time: 26 seconds 900 msec Ended Job = job_1457942575478_0007 Launching Job 2 out of 2 Number of reduce tasks determined at compile time: 1 In order to change the average load for a reducer (in bytes): set hive.exec.reducers.bytes.per.reducer=<number> In order to limit the maximum number of reducers: set hive.exec.reducers.max=<number> In order to set a constant number of reducers: set mapreduce.job.reduces=<number> Starting Job = job_1457942575478_0008, Tracking URL = http://spark-master:8088/proxy/application_1457942575478_0008/ Kill Command = /usr/local/hadoop/hadoop-2.6.0/bin/hadoop job -kill job_1457942575478_0008 Hadoop job information for Stage-2: number of mappers: 1; number of reducers: 1 2016-03-14 20:22:44,377 Stage-2 map = 0%, reduce = 0% 2016-03-14 20:23:07,303 Stage-2 map = 100%, reduce = 0%, Cumulative CPU 9.95 sec 2016-03-14 20:23:25,482 Stage-2 map = 100%, reduce = 82%, Cumulative CPU 15.54 sec 2016-03-14 20:23:26,563 Stage-2 map = 100%, reduce = 100%, Cumulative CPU 16.88 sec MapReduce Total cumulative CPU time: 16 seconds 880 msec Ended Job = job_1457942575478_0008 MapReduce Jobs Launched: Stage-Stage-1: Map: 1 Reduce: 1 Cumulative CPU: 26.9 sec HDFS Read: 114832713 HDFS Write: 15044297 SUCCESS Stage-Stage-2: Map: 1 Reduce: 1 Cumulative CPU: 16.88 sec HDFS Read: 15048892 HDFS Write: 153 SUCCESS Total MapReduce CPU Time Spent: 43 seconds 780 msec OK 百度 7564 baidu 3652 人体艺术 2786 馆陶县县长闫宁的父亲 2388 4399小游戏 2119 Time taken: 140.18 seconds, Fetched: 5 row(s)
六、外部表
我们在第三步创建的表是内部表,内部表创建成功后会在/user/hive/warehouse下创建和表同名的目录。并且当导入数据时,源文件会被放置在表对应的目录下。当进行表删除时,目录和文件一同被删除
hive> drop table sougou; OK Time taken: 0.983 seconds
查看hdfs
[email protected]:~# hdfs dfs -ls /user/hive/warehouse/ Found 1 items drwxr-xr-x - root supergroup 0 2016-03-14 17:10 /user/hive/warehouse/t1
Hive还提供了另一种表,称之为外部表。
表创建方式如下:
hive> CREATE EXTERNAL TABLE SOUGOU(ID STRING,WEBSESSION STRING,WORD STRING,S_SEQ INT,C_SEQ INT ,WEBSITE STRING) > ROW FORMAT DELIMITED FIELDS TERMINATED BY ‘\t‘ LINES TERMINATED BY ‘\n‘ > STORED AS TEXTFILE LOCATION ‘/library/sougou/sougouExternal‘; OK Time taken: 0.123 seconds
[email protected]:~# hdfs dfs -ls /user/hive/warehouse/ 16/03/14 20:02:14 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable Found 1 items drwxr-xr-x - root supergroup 0 2016-03-14 17:10 /user/hive/warehouse/t1 [email protected]:~# hdfs dfs -ls /library/sougou/ 16/03/14 20:02:29 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable Found 3 items -rw-r--r-- 3 root supergroup 217441417 2016-03-14 18:55 /library/sougou/SogouQ2.txt -rw-r--r-- 3 root supergroup 1086552775 2016-03-14 18:56 /library/sougou/SogouQ3.txt drwxr-xr-x - root supergroup 0 2016-03-14 20:01 /library/sougou/sougouExternal
目录直接创建在指定的位置。
上传文件
[email protected]:~# hdfs dfs -put SogouQ1.txt /library/sougou/sougouExternal
在Hive中查询数据
hive> select count(*) from sougou; Query ID = root_20160314200414_b514251b-58d3-40aa-a9ee-4a9cf5eef8f2 Total jobs = 1 Launching Job 1 out of 1 Number of reduce tasks determined at compile time: 1 In order to change the average load for a reducer (in bytes): set hive.exec.reducers.bytes.per.reducer=<number> In order to limit the maximum number of reducers: set hive.exec.reducers.max=<number> In order to set a constant number of reducers: set mapreduce.job.reduces=<number> Starting Job = job_1457942575478_0004, Tracking URL = http://spark-master:8088/proxy/application_1457942575478_0004/ Kill Command = /usr/local/hadoop/hadoop-2.6.0/bin/hadoop job -kill job_1457942575478_0004 Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 1 2016-03-14 20:04:27,514 Stage-1 map = 0%, reduce = 0% 2016-03-14 20:04:41,458 Stage-1 map = 100%, reduce = 0%, Cumulative CPU 4.66 sec 2016-03-14 20:04:52,341 Stage-1 map = 100%, reduce = 100%, Cumulative CPU 7.26 sec MapReduce Total cumulative CPU time: 7 seconds 260 msec Ended Job = job_1457942575478_0004 MapReduce Jobs Launched: Stage-Stage-1: Map: 1 Reduce: 1 Cumulative CPU: 7.26 sec HDFS Read: 114832746 HDFS Write: 8 SUCCESS Total MapReduce CPU Time Spent: 7 seconds 260 msec OK 1000000 Time taken: 39.823 seconds, Fetched: 1 row(s)
外部表被删除后,hdfs上的文件并不会被删除
hive> drop table sougou; OK Time taken: 0.363 seconds
[email protected]:~# hdfs dfs -ls /library/sougou/sougouExternal/ 16/03/14 20:16:28 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable Found 1 items -rw-r--r-- 3 root supergroup 114825752 2016-03-14 20:03 /library/sougou/sougouExternal/SogouQ1.txt
时间: 2024-10-04 23:33:31