hive安装配置

最新推荐文章于 2020-10-19 10:46:05 发布

iteye_3156

最新推荐文章于 2020-10-19 10:46:05 发布

阅读量130

点赞数

分类专栏： hadoop 文章标签：大数据数据库 java

本文链接：https://blog.csdn.net/iteye_3156/article/details/82272181

版权

hadoop 专栏收录该内容

34 篇文章 0 订阅

订阅专栏

requirement:
    Java 1.6
    Hadoop 0.20.x.

ref:https://cwiki.apache.org/confluence/display/Hive/GettingStarted#GettingStarted-InstallingHivefromaStableRelease

1)download hive
http://hive.apache.org/releases.html
2)Installing Hive
tar zxvf hive-0.7.0-bin.tar.gz

root@hadoop1:/opt#
root@hadoop1:/opt# ln -sf /opt/hadoop/hive-0.7.0-bin/ /opt/hadoop/hive

export HIVE_HOME=/opt/hadoop/hive
export PATH=/opt/hadoop/hive/bin:$PATH
5)running hive

$ $HADOOP_HOME/bin/hadoop fs -mkdir       /tmp
$ $HADOOP_HOME/bin/hadoop fs -mkdir       /user/hive/warehouse
$ $HADOOP_HOME/bin/hadoop fs -chmod g+w   /tmp
$ $HADOOP_HOME/bin/hadoop fs -chmod g+w   /user/hive/warehouse

root@hadoop1:/opt/hadoop/hive/bin# ./hive
Hive history file=/tmp/root/hive_job_log_root_201107121412_939983567.txt

DDL Operations
hive> CREATE TABLE pokes (foo INT, bar STRING);
OK
Time taken: 0.42 seconds
hive> CREATE TABLE invites (foo INT, bar STRING) PARTITIONED BY (ds STRING);
OK
Time taken: 0.099 seconds
hive> SHOW TABLES;
OK
invites
pokes
Time taken: 0.222 seconds
hive> SHOW TABLES '.*s';
OK
invites
pokes
Time taken: 0.134 seconds
hive> DESCRIBE invites;
OK
foo   int
bar   string
ds   string
Time taken: 0.174 seconds
hive> ALTER TABLE pokes ADD COLUMNS (new_col INT);
OK
Time taken: 0.147 seconds
hive> ALTER TABLE invites ADD COLUMNS (new_col2 INT COMMENT 'a comment');
OK
Time taken: 0.115 seconds
hive> DROP TABLE pokes;
OK
Time taken: 1.054 seconds
hive> show tables;
OK
invites
Time taken: 0.131 seconds

DML Operations
从本地加载文件
hive> LOAD DATA LOCAL INPATH '/opt/hadoop/hive/examples/files/kv1.txt' OVERWRITE INTO TABLE pokes;
Copying data from file:/opt/hadoop/hive/examples/files/kv1.txt
Copying file: file:/opt/hadoop/hive/examples/files/kv1.txt
Loading data to table default.pokes
Deleted hdfs://hadoop1:9000/user/hive/warehouse/pokes
OK
Time taken: 0.318 seconds

hive> select * from pokes limit 10;
OK
238   val_238
86   val_86
311   val_311
27   val_27
165   val_165
409   val_409
255   val_255
278   val_278
98   val_98
484   val_484
Time taken: 0.137 seconds

--分区表加载
hive> LOAD DATA LOCAL INPATH '/opt/hadoop/hive/examples/files/kv2.txt' OVERWRITE INTO TABLE invites PARTITION (ds='2008-08-15');
Copying data from file:/opt/hadoop/hive/examples/files/kv2.txt
Copying file: file:/opt/hadoop/hive/examples/files/kv2.txt
Loading data to table default.invites partition (ds=2008-08-15)
OK
Time taken: 0.394 seconds
hive> select * from invites limit 10;
OK
474   val_475   NULL   2008-08-15
281   val_282   NULL   2008-08-15
179   val_180   NULL   2008-08-15
291   val_292   NULL   2008-08-15
62   val_63   NULL   2008-08-15
271   val_272   NULL   2008-08-15
217   val_218   NULL   2008-08-15
135   val_136   NULL   2008-08-15
167   val_168   NULL   2008-08-15
468   val_469   NULL   2008-08-15
Time taken: 0.217 seconds

hive> LOAD DATA LOCAL INPATH './examples/files/kv3.txt' OVERWRITE INTO TABLE invites PARTITION (ds='2008-08-08');root@hadoop1:/opt/hadoop/hive/bin#
root@hadoop1:/opt/hadoop/hive/bin# ./hive
Hive history file=/tmp/root/hive_job_log_root_201107121431_842989549.txt
hive> LOAD DATA LOCAL INPATH '/opt/hadoop/hive/examples/files/kv3.txt' OVERWRITE INTO TABLE invites PARTITION (ds='2008-08-08');
Copying data from file:/opt/hadoop/hive/examples/files/kv3.txt
Copying file: file:/opt/hadoop/hive/examples/files/kv3.txt
Loading data to table default.invites partition (ds=2008-08-08)
OK
Time taken: 6.787 seconds
hive> select * from invites limit 10;
OK
238   val_238   NULL   2008-08-08
NULL       NULL   2008-08-08
311   val_311   NULL   2008-08-08
NULL   val_27   NULL   2008-08-08
NULL   val_165   NULL   2008-08-08
NULL   val_409   NULL   2008-08-08
255   val_255   NULL   2008-08-08
278   val_278   NULL   2008-08-08
98   val_98   NULL   2008-08-08
NULL   val_484   NULL   2008-08-08
Time taken: 0.589 seconds

SQL Operations
hive>
    >
    >
    > SELECT a.foo FROM invites a WHERE a.ds='2008-08-15' limit 10;
Total MapReduce jobs = 1
Launching Job 1 out of 1
Number of reduce tasks is set to 0 since there's no reduce operator
Execution log at: /tmp/root/root_20110712144040_e058253d-bb7f-45b9-97b8-f6c78c5483b1.log
Job running in-process (local Hadoop)
2011-07-12 14:40:52,786 null map = 100%, reduce = 0%
Ended Job = job_local_0001
OK
474
281
179
291
62
271
217
135
167
468
Time taken: 3.62 seconds

hive> INSERT OVERWRITE DIRECTORY '/tmp/hdfs_out' SELECT a.* FROM invites a WHERE a.ds='2008-08-15';
Total MapReduce jobs = 2
Launching Job 1 out of 2
Number of reduce tasks is set to 0 since there's no reduce operator
Execution log at: /tmp/root/root_20110712144141_7b6e4021-a419-42b5-a6eb-c45010872c0a.log
Job running in-process (local Hadoop)
2011-07-12 14:41:39,056 null map = 100%, reduce = 0%
Ended Job = job_local_0001
Ended Job = -1864542964, job is filtered out (removed at runtime).
Moving data to: hdfs://hadoop1:9000/tmp/hive-root/hive_2011-07-12_14-41-36_001_2590472032748705056/-ext-10000
Moving data to: /tmp/hdfs_out
OK
Time taken: 3.247 seconds

hive> INSERT OVERWRITE LOCAL DIRECTORY '/tmp/local_out' SELECT a.* FROM pokes a;
Total MapReduce jobs = 1
Launching Job 1 out of 1
Number of reduce tasks is set to 0 since there's no reduce operator
Execution log at: /tmp/root/root_20110712144242_920398b9-3c37-431b-b088-dcffe1c54aa2.log
Job running in-process (local Hadoop)
2011-07-12 14:42:19,666 null map = 100%, reduce = 0%
Ended Job = job_local_0001
Copying data to local directory /tmp/local_out
Copying data to local directory /tmp/local_out
OK
Time taken: 3.189 seconds

lpxuan@hadoop1:/tmp/local_out$ more 000000_0
238val_238
86val_86
311val_311
27val_27
165val_165

--group by operation
hive>
    >
    > SELECT a.bar, count(*) FROM invites a WHERE a.foo > 0 GROUP BY a.bar;
Total MapReduce jobs = 1
Launching Job 1 out of 1
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 mapred.reduce.tasks=<number>
Execution log at: /tmp/root/root_20110712144545_01fc3105-f98f-4d77-841f-61c5d65f80fc.log
Job running in-process (local Hadoop)
2011-07-12 14:45:45,313 null map = 0%, reduce = 0%
2011-07-12 14:45:53,745 null map = 100%, reduce = 0%
2011-07-12 14:45:55,748 null map = 100%, reduce = 100%
Ended Job = job_local_0001
OK
   3
val_100   1
val_101   2
val_79   1
val_81   1
val_83   2
val_86   1
val_87   1
val_88   2
val_9   1
val_90   3
val_92   1
val_94   3
val_95   1
val_98   3
..
Time taken: 18.354 seconds

--join
hive> SELECT t1.bar, t1.foo, t2.foo FROM pokes t1 JOIN invites t2 ON (t1.bar = t2.bar) limit 10;
Total MapReduce jobs = 1
Launching Job 1 out of 1
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 mapred.reduce.tasks=<number>
Execution log at: /tmp/root/root_20110712144848_0aa68f57-4d70-4cdc-abb7-fa48a5e379dd.log
Job running in-process (local Hadoop)
2011-07-12 14:48:55,650 null map = 0%, reduce = 0%
2011-07-12 14:48:56,653 null map = 100%, reduce = 0%
2011-07-12 14:48:57,659 null map = 100%, reduce = 100%
Ended Job = job_local_0001
OK
val_100   100   99
val_100   100   99
val_103   103   102
val_103   103   102
val_105   105   104
val_105   105   104
val_105   105   104
val_11   11   10
val_111   111   110
val_118   118   117
Time taken: 8.686 seconds

============================================================

搭建Hive平台

Hive是一个基于Hadoop的数据仓库平台。通过hive，我们可以方便地进行ETL的工作。hive定义了一个类似于SQL的查询语言：HQL，能够将用户编写的QL转化为相应的Mapreduce程序基于Hadoop执行。

本文讲解如何搭建一个Hive平台。假设我们有3台机器：hadoop1,hadoop2,hadoop3。并且都安装好了Hadoop-0.19.2（hive支持的Hadoop版本很多），hosts文件配置正确。Hive部署在hadoop1机器上。

最简单，最快速的部署方案

在Hadoop-0.19.2中自带了hive的文件。版本为0.3.0。

我们首先启动Hadoop：sh $HADOOP_HOME/bin/start-all.sh

然后启动hive即可：sh $HADOOP_HOME/contrib/hive/bin/hive

这个时候，我们的Hive的命令行接口就启动起来了，你可以直接输入命令来执行相应的hive应用了。

这种部署方式使用derby的嵌入式模式，虽然简单快速，但是无法提供多用户同时访问，所以只能用于简单的测试，无法实际应用于生产环境。所以，我们要修改hive的默认配置，提高可用性。

搭建多用户的，提供web界面的部署方案

目前只用比较多hive版本是hive-0.4.1。我们将使用这个版本来搭建hive平台。

首先，下载hive-0.4.1：svn co http://svn.apache.org/repos/asf/hadoop/hive/tags/release-0.4.1/ hive-0.4.1

然后，修改下载文件里面的编译选项文件shims/ivy.xml，将其修改为如下内容（对应的Hadoop版本为0.19.2）

<ivy-module version="2.0">
    <info organisation="org.apache.hadoop.hive" module="shims"/>
    <dependencies>
        <dependency org="hadoop" name="core" rev="0.19.2">
          <artifact name="hadoop" type="source" ext="tar.gz"/>
        </dependency>
        <conflict manager="all" />
    </dependencies>
</ivy-module>

接下来，我们使用ant去编译hive: ant package

编译成功后，我们会发现在build/dist目录中就是编译成功的文件。将这个目录设为$HIVE_HOME

修改conf/hive-default.xml文件，主要修改内容如下：

<property>
<name>javax.jdo.option.ConnectionURL</name>
<value>jdbc:derby://hadoop1:1527/metastore_db;create=true</value>
<description>JDBC connect string for a JDBC metastore</description>
</property>
<property>
<name>javax.jdo.option.ConnectionDriverName</name>
<value>org.apache.derby.jdbc.ClientDriver</value>
<description>Driver class name for a JDBC metastore</description>
</property>

在hadoop1机器上下载并安装apache derby数据库：wget http://labs.renren.com/apache-mirror/db/derby/db-derby-10.5.3.0/db-derby-10.5.3.0-bin.zip

解压derby后，设置$DERBY_HOME

然后启动derby的network Server：sh $DERBY_HOME/bin/startNetworkServer -h 0.0.0.0

接下来，将$DERBY_HOME/lib目录下的derbyclient.jar与derbytools.jar文件copy到$HIVE_HOME/lib目录下。

启动Hadoop：sh $HADOOP_HOME/bin/start-all.sh

最后，启动hive的web界面：sh $HIVE_HOME/bin/hive --service hwi

这样，我们的hive就部署完成了。我们可以直接在浏览器中输入: http://hadoop1:9999/hwi/ 进行访问了（如果不行话，请将hadoop1替换为实际的ip地址，如：http://10.210.152.17:9999/hwi/）。

这种部署方式使用derby的c/s模式，允许多用户同时访问，同时提供web界面，方便使用。推荐使用这种部署方案。

关注Hive的schema

我们上面谈到的2中部署方案都是使用derby数据库来保存hive中的schema信息。我们也可以使用其他的数据库来保存schema信息，如mysql。

可以参考这篇文章了解如果使用mysql来替换derby：http://www.mazsoft.com/blog/post/2010/02/01/Setting-up-HadoopHive-to-use-MySQL-as-metastore.aspx

我们也可以使用HDFS来保存schema信息，具体的做法是修改conf/hive-default.xml，修改内容如下：

<property>
<name>hive.metastore.rawstore.impl</name>
<value>org.apache.hadoop.hive.metastore.FileStore</value>
<description>Name of the class that implements org.apache.hadoop.hive.metastore.rawstore interface. This class is used to store and retrieval of raw metadata objects such as table, database</description>
</property>