Hive 0.13.1 on Hadoop2.2.0 + Oracle10g部署详解

环境：

Hadoop2.2.0

hive0.13.1

Ubuntu 14.04 LTS

java version "1.7.0_60"

Oracle10g

到以下地址下载安装包

http://mirrors.cnnic.cn/apache/hive/stable/apache-hive-0.13.1-bin.tar.gz

安装包解压到服务器上

/home/fulong/Hive/apache-hive-0.13.1-bin

修改环境变量，添加以下内容

export HIVE_HOME=/home/fulong/Hive/apache-hive-0.13.1-bin

export PATH=$HIVE_HOME/bin:$PATH

进到conf目录下拷贝模板配置文件重命名

fulong@FBI006:~/Hive/apache-hive-0.13.1-bin/conf$ ls

hive-default.xml.template  hive-exec-log4j.properties.template

hive-env.sh.template       hive-log4j.properties.template

fulong@FBI006:~/Hive/apache-hive-0.13.1-bin/conf$ cp hive-env.sh.template hive-env.sh

fulong@FBI006:~/Hive/apache-hive-0.13.1-bin/conf$ cp hive-default.xml.template hive-site.xml

fulong@FBI006:~/Hive/apache-hive-0.13.1-bin/conf$ ls

hive-default.xml.template  hive-env.sh.template                 hive-log4j.properties.template

hive-env.sh               hive-exec-log4j.properties.template  hive-site.xml

修改配置文件hive-env.sh中的以下几处，分别制定Hadoop的根目录，Hive的conf和lib目录

# Set HADOOP_HOME to point to a specific hadoop install directory

HADOOP_HOME=/home/fulong/Hadoop/hadoop-2.2.0

# Hive Configuration Directory can be controlled by:

export HIVE_CONF_DIR=/home/fulong/Hive/apache-hive-0.13.1-bin/conf

# Folder containing extra ibraries required for hive compilation/execution can be controlled by:

export HIVE_AUX_JARS_PATH=/home/fulong/Hive/apache-hive-0.13.1-bin/lib

修改配置文件hive-site.sh中的以下几处连接Oracle相关参数

<property>

  <name>javax.jdo.option.ConnectionURL</name>

  <value>jdbc:oracle:thin:@192.168.0.138:1521:orcl</value>

  <description>JDBC connect string for a JDBC metastore</description>

</property>

<property>

  <name>javax.jdo.option.ConnectionDriverName</name>

  <value>oracle.jdbc.driver.OracleDriver</value>

  <description>Driver class name for a JDBC metastore</description>

</property>

<property>

  <name>javax.jdo.option.ConnectionUserName</name>

  <value>hive</value>

  <description>username to use against metastore database</description>

</property>

<property>

  <name>javax.jdo.option.ConnectionPassword</name>

  <value>hivefbi</value>

  <description>password to use against metastore database</description>

</property>

配置log4j

在$HIVE_HOME下创建log4j目录，用于存储日志文件

拷贝模板重命名

fulong@FBI006:~/Hive/apache-hive-0.13.1-bin/conf$ cp hive-log4j.properties.template hive-log4j.properties

修改存放日志的目录

hive.log.dir=/home/fulong/Hive/apache-hive-0.13.1-bin/log4j

拷贝Oracle JDBC的jar包

将对应Oracle的jdbc包拷贝到$HIVE_HOME/lib下

启动Hive

fulong@FBI006:~/Hive/apache-hive-0.13.1-bin$ hive

14/08/20 17:14:05 INFO Configuration.deprecation: mapred.reduce.tasks is deprecated. Instead, use mapreduce.job.reduces

14/08/20 17:14:05 INFO Configuration.deprecation: mapred.min.split.size is deprecated. Instead, use mapreduce.input.fileinputformat.split.minsize

14/08/20 17:14:05 INFO Configuration.deprecation: mapred.reduce.tasks.speculative.execution is deprecated. Instead, use mapreduce.reduce.speculative

14/08/20 17:14:05 INFO Configuration.deprecation: mapred.min.split.size.per.node is deprecated. Instead, use mapreduce.input.fileinputformat.split.minsize.per.node

14/08/20 17:14:05 INFO Configuration.deprecation: mapred.input.dir.recursive is deprecated. Instead, use mapreduce.input.fileinputformat.input.dir.recursive

14/08/20 17:14:05 INFO Configuration.deprecation: mapred.min.split.size.per.rack is deprecated. Instead, use mapreduce.input.fileinputformat.split.minsize.per.rack

14/08/20 17:14:05 INFO Configuration.deprecation: mapred.max.split.size is deprecated. Instead, use mapreduce.input.fileinputformat.split.maxsize

14/08/20 17:14:05 INFO Configuration.deprecation: mapred.committer.job.setup.cleanup.needed is deprecated. Instead, use mapreduce.job.committer.setup.cleanup.needed

14/08/20 17:14:05 WARN conf.HiveConf: DEPRECATED: hive.metastore.ds.retry.* no longer has any effect.  Use hive.hmshandler.retry.* instead

Logging initialized using configuration in file:/home/fulong/Hive/apache-hive-0.13.1-bin/conf/hive-log4j.properties

Java HotSpot(TM) 64-Bit Server VM warning: You have loaded library /home/fulong/Hadoop/hadoop-2.2.0/lib/native/libhadoop.so which might have disabled stack guard. The VM will try to fix the stack guard now.

It's highly recommended that you fix the library with 'execstack -c <libfile>', or link it with '-z noexecstack'.

hive>

验证

打算创建一张表存储搜狗实验室下载的用户搜索行为日志。

数据下载地址：

http://www.sogou.com/labs/dl/q.html

首先创建表：

hive> create table searchlog (time string,id string,sword string,rank int,clickrank int,url string) row format delimited fields terminated by '\t' lines terminated by '\n' stored as textfile;

此时会报错：

FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.DDLTask. MetaException(message:javax.jdo.JDODataStoreException: An exception was thrown while adding/validating class(es) : ORA-01754: a table may contain only one column of type LONG

解决办法：

用解压缩工具打开${HIVE_HOME}/lib中的hive-metastore-0.13.0.jar，发现名为package.jdo的文件，打开该文件并找到下面的内容。

<field name="viewOriginalText" default-fetch-group="false">

        <column name="VIEW_ORIGINAL_TEXT" jdbc-type="LONGVARCHAR"/>

</field>

<field name="viewExpandedText" default-fetch-group="false">

        <column name="VIEW_EXPANDED_TEXT" jdbc-type="LONGVARCHAR"/>

</field>

可以发现列VIEW_ORIGINAL_TEXT和VIEW_EXPANDED_TEXT的类型都为LONGVARCHAR，对应于Oracle中的LONG，这样就与Oracle表只能存在一列类型为LONG的列的要求相矛盾，所以就出现错误了。

按照Hive官网的建议将该两列的jdbc-type的值改为CLOB，修改后的内容如下所示。

<field name="viewOriginalText"default-fetch-group="false">

             <column name="VIEW_ORIGINAL_TEXT" jdbc-type="CLOB"/>

</field>

<field name="viewExpandedText"default-fetch-group="false">

             <column name="VIEW_EXPANDED_TEXT" jdbc-type="CLOB"/>

</field>

修改以后，重启hive。

重新执行创建表的命令，创建表成功：

hive> create table searchlog (time string,id string,sword string,rank int,clickrank int,url string) row format delimited fields terminated by '\t' lines terminated by '\n' stored as textfile;

OK

Time taken: 0.986 seconds

将本地数据加载进表中：

hive> load data local inpath '/home/fulong/Downloads/SogouQ.reduced' overwrite into table searchlog;

Copying data from file:/home/fulong/Downloads/SogouQ.reduced

Copying file: file:/home/fulong/Downloads/SogouQ.reduced

Loading data to table default.searchlog

rmr: DEPRECATED: Please use 'rm -r' instead.

Deleted hdfs://fulonghadoop/user/hive/warehouse/searchlog

Table default.searchlog stats: [numFiles=1, numRows=0, totalSize=152006060, rawDataSize=0]

OK

Time taken: 25.705 seconds

查看所有表：

hive> show tables;

OK

searchlog

Time taken: 0.139 seconds, Fetched: 1 row(s)

统计行数：

hive> select count(*) from searchlog;

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_1407233914535_0001, Tracking URL = http://FBI003:8088/proxy/application_1407233914535_0001/

Kill Command = /home/fulong/Hadoop/hadoop-2.2.0/bin/hadoop job  -kill job_1407233914535_0001

Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 1

2014-08-20 18:03:17,667 Stage-1 map = 0%,  reduce = 0%

2014-08-20 18:04:05,426 Stage-1 map = 100%,  reduce = 0%, Cumulative CPU 3.46 sec

2014-08-20 18:04:27,317 Stage-1 map = 100%,  reduce = 100%, Cumulative CPU 4.74 sec

MapReduce Total cumulative CPU time: 4 seconds 740 msec

Ended Job = job_1407233914535_0001

MapReduce Jobs Launched:

Job 0: Map: 1  Reduce: 1   Cumulative CPU: 4.74 sec   HDFS Read: 152010455 HDFS Write: 8 SUCCESS

Total MapReduce CPU Time Spent: 4 seconds 740 msec

OK

1724264

Time taken: 103.154 seconds, Fetched: 1 row(s)