这一篇介绍Hive的安装及操作。版本是Hive3.1.2。
在Hadoop篇里，我用了5台虚拟机来搭建集群，但是我的电脑只有8G内存，虚拟机启动之后卡到没法操作，把自己坑惨了。
Hive的运行是基于Hadoop集群的，为了操作的流畅性，我把Hadoop集群全部重新搭建了，一共只有3台虚拟机。
| 主机 | NN | RM | ZKFC | DN | NM | JN | ZK | MySQL | Hive服务端 | Hive客户端 |
| server01 | ? | ? | ? | ? | ? | ? | ? | |||
| server02 | ? | ? | ? | ? | ? | ? | ? | ? | ||
| server03 | ? | ? | ? | ? | ? | ? |
一般我们使用关系型数据库来存储Hive的元数据，mysql用的比较多。
先检查一下是否有mysql的yum源，注意我们要安装的是mysql-server。下图中没有这个源，需要下载。
[root@server03 ~]# yum list|grep mysql akonadi-mysql.x86_64 1.9.2-4.el7 base apr-util-mysql.x86_64 1.5.2-6.el7 base dovecot-mysql.x86_64 1:2.2.36-6.el7_8.1 updates freeradius-mysql.x86_64 3.0.13-10.el7_6 base libdbi-dbd-mysql.x86_64 0.8.3-16.el7 base mysql-connector-java.noarch 1:5.1.25-3.el7 base mysql-connector-odbc.x86_64 5.2.5-8.el7 base pcp-pmda-mysql.x86_64 4.3.2-7.el7_8 updates php-mysql.x86_64 5.4.16-48.el7 base php-mysqlnd.x86_64 5.4.16-48.el7 base qt-mysql.i686 1:4.8.7-8.el7 base qt-mysql.x86_64 1:4.8.7-8.el7 base qt5-qtbase-mysql.i686 5.9.7-2.el7 base qt5-qtbase-mysql.x86_64 5.9.7-2.el7 base redland-mysql.x86_64 1.0.16-6.el7 base rsyslog-mysql.x86_64 8.24.0-52.el7_8.2 updates
下载mysql的repo源
[root@server03 ~]# wget http://repo.mysql.com/mysql-community-release-el7-5.noarch.rpm --2020-10-12 22:36:02-- http://repo.mysql.com/mysql-community-release-el7-5.noarch.rpm 正在解析主机 repo.mysql.com (repo.mysql.com)... 23.45.57.22 正在连接 repo.mysql.com (repo.mysql.com)|23.45.57.22|:80... 已连接。 已发出 HTTP 请求，正在等待回应... 200 OK 长度：6140 (6.0K) [application/x-redhat-package-manager] 正在保存至: “mysql-community-release-el7-5.noarch.rpm” 100%[======================================================================================>] 6,140 --.-K/s 用时 0.002s 2020-10-12 22:36:02 (3.80 MB/s) - 已保存 “mysql-community-release-el7-5.noarch.rpm” [6140/6140])
安装mysql-community-release-el7-5.noarch.rpm包
[root@server03 home]# rpm -ivh mysql-community-release-el7-5.noarch.rpm 准备中... ################################# [100%] 正在升级/安装... 1:mysql-community-release-el7-5 ################################# [100%]
安装mysql
[root@server03 home]# yum -y install mysql-server 已加载插件：fastestmirror Loading mirror speeds from cached hostfile * base: mirrors.ustc.edu.cn * extras: mirrors.cn99.com * updates: mirrors.njupt.edu.cn 正在解决依赖关系 --> 正在检查事务 ---> 软件包 mysql-community-server.x86_64.0.5.6.49-2.el7 将被 安装 --> 正在处理依赖关系 mysql-community-common(x86-64) = 5.6.49-2.el7，它被软件包 mysql-community-server-5.6.49-2.el7.x86_64 需要
........................................................... perl-Scalar-List-Utils.x86_64 0:1.27-248.el7 perl-Socket.x86_64 0:2.010-5.el7 perl-Storable.x86_64 0:2.45-3.el7 perl-Text-ParseWords.noarch 0:3.29-4.el7 perl-Time-HiRes.x86_64 4:1.9725-3.el7 perl-Time-Local.noarch 0:1.2300-2.el7 perl-constant.noarch 0:1.27-2.el7 perl-libs.x86_64 4:5.16.3-295.el7 perl-macros.x86_64 4:5.16.3-295.el7 perl-parent.noarch 1:0.225-244.el7 perl-podlators.noarch 0:2.5.1-3.el7 perl-threads.x86_64 0:1.87-4.el7 perl-threads-shared.x86_64 0:1.43-6.el7 替代: mariadb-libs.x86_64 1:5.5.64-1.el7 完毕！
启动mysql
[root@server03 home]# service mysqld start
Redirecting to /bin/systemctl start mysqld.service
需要更改权限才能实现远程连接MYSQL数据库。可以通过以下方式来修改权限。
[root@server03 ~]# mysql -uroot Welcome to the MySQL monitor. Commands end with ; or \g. Your MySQL connection id is 7 Server version: 5.6.49 MySQL Community Server (GPL) Copyright (c) 2000, 2020, Oracle and/or its affiliates. All rights reserved. Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other names may be trademarks of their respective owners. Type ‘help;‘ or ‘\h‘ for help. Type ‘\c‘ to clear the current input statement. mysql> use mysql; Reading table information for completion of table and column names You can turn off this feature to get a quicker startup with -A Database changed mysql> select host,user,password from user; +-----------+------+-------------------------------------------+ | host | user | password | +-----------+------+-------------------------------------------+ | localhost | root | | | server03 | root | | | 127.0.0.1 | root | | | ::1 | root | | | localhost | | | | server03 | | | +-----------+------+-------------------------------------------+ 6 rows in set (0.00 sec) mysql> grant all privileges on *.* to ‘root‘@‘%‘ identified by ‘111111‘ with grant option; Query OK, 0 rows affected (0.00 sec) mysql> flush privileges; Query OK, 0 rows affected (0.00 sec) mysql> select host,user,password from user; +-----------+------+-------------------------------------------+ | host | user | password | +-----------+------+-------------------------------------------+ | localhost | root | | | server03 | root | | | 127.0.0.1 | root | | | ::1 | root | | | localhost | | | | server03 | | | | % | root | *A46E551C358E21DEFC306B31BB84ADBFC2A75AAB | +-----------+------+-------------------------------------------+ 7 rows in set (0.00 sec)
至此，mysql配置完毕。
这里采用最常用的远程元数据模式来安装Hive。server03上安装Hive服务端，server02上安装Hive客户端。
在server03上解压hive压缩包到/usr目录下，并将用户和组修改为hadoop。将conf目录下的hive-default.xml.template复制一份，命名为hive-site.xml。注意，这个名字必须是hive-site.xml，跟hdfs类似。
[hadoop@server03 conf]$ pwd /usr/apache-hive-3.1.2/conf [hadoop@server03 conf]$ cp hive-default.xml.template hive-site.xml
修改hive-site.xml文件中的配置如下
<configuration> <property> <name>hive.metastore.warehouse.dir</name> <value>/opt/hive_remote/warehouse</value> </property> <property> <name>javax.jdo.option.ConnectionURL</name> <value>jdbc:mysql://server03:3306/hive_remote?createDatabaseIfNotExist=true</value> </property> <property> <name>javax.jdo.option.ConnectionDriverName</name> <value>com.mysql.jdbc.Driver</value> </property> <property> <name>javax.jdo.option.ConnectionUserName</name> <value>root</value> </property> <property> <name>javax.jdo.option.ConnectionPassword</name> <value>111111</value> </property> </configuration>
第一个配置项是hdfs里的路径，我们不需要创建这个目录，元数据初始化的时候会自动在hdfs创建/opt/hive_remote/warehouse
配置hive环境变量
JAVA_HOME=/usr/java/jdk-11.0.8/
ZOOKEEPER_HOME=/usr/apache-zookeeper-3.5.8/
HADOOP_HOME=/usr/hadoop-3.3.0
HIVE_HOME=/usr/apache-hive-3.1.2
PATH=$PATH:$JAVA_HOME/bin:$ZOOKEEPER_HOME/bin:$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$HIVE_HOME/bin
在网上下载mysql驱动包mysql-connector-java-bin.jar，放在hive的lib目录下。在server03上执行元数据初始化操作。
[hadoop@server03 ~]$ schematool -dbType mysql -initSchema SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/apache-hive-3.1.2/lib/log4j-slf4j-impl-2.10.0.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/hadoop-3.3.0/share/hadoop/common/lib/slf4j-log4j12-1.7.25.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.apache.logging.slf4j.Log4jLoggerFactory] Exception in thread "main" java.lang.NoSuchMethodError: com.google.common.base.Preconditions.checkArgument(ZLjava/lang/String;Ljava/lang/Object;)V at org.apache.hadoop.conf.Configuration.set(Configuration.java:1380) at org.apache.hadoop.conf.Configuration.set(Configuration.java:1361) at org.apache.hadoop.mapred.JobConf.setJar(JobConf.java:536) at org.apache.hadoop.mapred.JobConf.setJarByClass(JobConf.java:554) at org.apache.hadoop.mapred.JobConf.<init>(JobConf.java:448) at org.apache.hadoop.hive.conf.HiveConf.initialize(HiveConf.java:5141) at org.apache.hadoop.hive.conf.HiveConf.<init>(HiveConf.java:5104) at org.apache.hive.beeline.HiveSchemaTool.<init>(HiveSchemaTool.java:96) at org.apache.hive.beeline.HiveSchemaTool.main(HiveSchemaTool.java:1473) at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.base/java.lang.reflect.Method.invoke(Method.java:566) at org.apache.hadoop.util.RunJar.run(RunJar.java:323) at org.apache.hadoop.util.RunJar.main(RunJar.java:236)
这里报错了。原因是hive里的guava.jar与hadoop里的guava.jar版本不一致。需要把hive里的低版本删除，再把hadoop里的高版本拷贝过来。
[hadoop@server03 ~]$ ll /usr/hadoop-3.3.0/share/hadoop/common/lib |grep guava
-rw-r--r--. 1 hadoop hadoop 2747878 7月 7 02:47 guava-27.0-jre.jar
-rw-r--r--. 1 hadoop hadoop 2199 7月 7 02:47 listenablefuture-9999.0-empty-to-avoid-conflict-with-guava.jar
[hadoop@server03 ~]$ ll /usr/apache-hive-3.1.2/lib/ |grep guava
-rw-r--r--. 1 hadoop hadoop 2308517 9月 27 2018 guava-19.0.jar
-rw-r--r--. 1 hadoop hadoop 971309 5月 21 2019 jersey-guava-2.25.1.jar
[hadoop@server03 ~]$ rm -f /usr/apache-hive-3.1.2/lib/guava-19.0.jar
[hadoop@server03 ~]$ cp /usr/hadoop-3.3.0/share/hadoop/common/lib/guava-27.0-jre.jar /usr/apache-hive-3.1.2/lib/
[hadoop@server03 ~]$ ll /usr/apache-hive-3.1.2/lib/ |grep guava
-rw-r--r--. 1 hadoop hadoop 2747878 10月 19 00:15 guava-27.0-jre.jar
-rw-r--r--. 1 hadoop hadoop 971309 5月 21 2019 jersey-guava-2.25.1.jar
再次执行初始化脚本
[hadoop@server03 ~]$ schematool -dbType mysql -initSchema SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/apache-hive-3.1.2/lib/log4j-slf4j-impl-2.10.0.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/hadoop-3.3.0/share/hadoop/common/lib/slf4j-log4j12-1.7.25.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.apache.logging.slf4j.Log4jLoggerFactory] Metastore connection URL: jdbc:mysql://server03:3306/hive_remote?createDatabaseIfNotExist=true Metastore Connection Driver : com.mysql.jdbc.Driver Metastore connection User: root Starting metastore schema initialization to 3.1.0 Initialization script hive-schema-3.1.0.mysql.sql Initialization script completed schemaTool completed
Hive服务端就搭建成功了，下面启动服务。
[hadoop@server03 ~]$ hive --service metastore 2020-10-21 08:14:44: Starting Hive Metastore Server SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/apache-hive-3.1.2/lib/log4j-slf4j-impl-2.10.0.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/hadoop-3.3.0/share/hadoop/common/lib/slf4j-log4j12-1.7.25.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.apache.logging.slf4j.Log4jLoggerFactory]
这是一个阻塞式窗口，再开一个窗口看看进程。为了看得更清楚，先不启动Hadoop集群。
[hadoop@server03 ~]$ jps 3910 Jps 3656 RunJar
RunJar就是Hive的服务端进程。
在server02上安装Hive客户端。客户端做的事情与服务端差不多，只是不需要执行元数据初始化的动作。修改hive-site.xml配置如下：
<property>
<name>hive.metastore.warehouse.dir</name>
<value>/user/hive_remote/warehouse</value>
</property>
<property>
<name>hive.metastore.uris</name>
<value>thrift://node03:9083</value>
</property>
执行hive，进入到hive的cli窗口
[hadoop@server02 ~]$ hive which: no hbase in (/usr/local/bin:/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/usr/java/jdk-11.0.8/bin:/usr/apache-zookeeper-3.5.8/bin:/usr/hadoop-3.3.0/bin:/usr/hadoop-3.3.0/sbin:/usr/apache-hive-3.1.2/bin:/home/hadoop/.local/bin:/home/hadoop/bin) SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/apache-hive-3.1.2/lib/log4j-slf4j-impl-2.10.0.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/hadoop-3.3.0/share/hadoop/common/lib/slf4j-log4j12-1.7.25.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.apache.logging.slf4j.Log4jLoggerFactory] Hive Session ID = 52f8c152-cd38-4293-b2b6-3efc17546299 Exception in thread "main" java.lang.ClassCastException: class jdk.internal.loader.ClassLoaders$AppClassLoader cannot be cast to class java.net.URLClassLoader (jdk.internal.loader.ClassLoaders$AppClassLoader and java.net.URLClassLoader are in module java.base of loader ‘bootstrap‘) at org.apache.hadoop.hive.ql.session.SessionState.<init>(SessionState.java:413) at org.apache.hadoop.hive.ql.session.SessionState.<init>(SessionState.java:389) at org.apache.hadoop.hive.cli.CliSessionState.<init>(CliSessionState.java:60) at org.apache.hadoop.hive.cli.CliDriver.run(CliDriver.java:705) at org.apache.hadoop.hive.cli.CliDriver.main(CliDriver.java:683) at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.base/java.lang.reflect.Method.invoke(Method.java:566) at org.apache.hadoop.util.RunJar.run(RunJar.java:323) at org.apache.hadoop.util.RunJar.main(RunJar.java:236)
执行报错了。到网上查这个错误信息，得出Hive3.1.2只支持jdk8。之前搭建Hadoop集群用的jdk11，哦豁，这下悲剧了。
没办法，只能再装一个jdk8了。
[root@server01 home]# rpm -ivh jdk-8u271-linux-x64.rpm 警告：jdk-8u271-linux-x64.rpm: 头V3 RSA/SHA256 Signature, 密钥 ID ec551f03: NOKEY 准备中... ################################# [100%] 正在升级/安装... 1:jdk1.8-2000:1.8.0_271-fcs ################################# [100%] Unpacking JAR files... tools.jar... plugin.jar... javaws.jar... deploy.jar... rt.jar... jsse.jar... charsets.jar... localedata.jar...
安装完毕后，看看java目录，现在有两个版本的java
[root@server01 java]# cd /usr/java [root@server01 java]# ll 总用量 0 lrwxrwxrwx. 1 root root 16 9月 23 18:21 default -> /usr/java/latest drwxr-xr-x. 9 root root 128 9月 23 18:21 jdk-11.0.8 drwxr-xr-x. 9 root root 286 10月 23 13:04 jdk1.8.0_271-amd64 lrwxrwxrwx. 1 root root 20 9月 23 18:21 latest -> /usr/java/jdk-11.0.8
先把目录名字修改一下，太长了不方便
[root@server01 java]# mv jdk1.8.0_271-amd64/ jdk1.8.0 [root@server01 java]# ll 总用量 0 lrwxrwxrwx. 1 root root 16 9月 23 18:21 default -> /usr/java/latest drwxr-xr-x. 9 root root 128 9月 23 18:21 jdk-11.0.8 drwxr-xr-x. 9 root root 286 10月 23 13:04 jdk1.8.0 lrwxrwxrwx. 1 root root 20 9月 23 18:21 latest -> /usr/java/jdk-11.0.8
接着修改环境变量/etc/profile和hadoop-env.sh里的配置，执行source /etc/profile让配置生效
#JAVA_HOME=/usr/java/jdk-11.0.8 JAVA_HOME=/usr/java/jdk1.8.0
再切换java版本
[root@server01 java]# alternatives --install /usr/bin/java java /usr/java/jdk1.8.0/bin/java 1400 [root@server01 java]# alternatives --config java 共有 3 个提供“java”的程序。 选项 命令 ----------------------------------------------- *+ 1 /usr/java/jdk-11.0.8/bin/java 2 /usr/java/jdk1.8.0_271-amd64/bin/java 3 /usr/java/jdk1.8.0/bin/java 按 Enter 保留当前选项[+]，或者键入选项编号：3
第2个版本是安装rpm包的时候自动生成的，我们的路径修改之后应该把它删掉
[root@server01 java]# alternatives --remove java /usr/java/jdk1.8.0_271-amd64/bin/java [root@server01 java]# alternatives --config java 共有 2 个提供“java”的程序。 选项 命令 ----------------------------------------------- * 1 /usr/java/jdk-11.0.8/bin/java + 2 /usr/java/jdk1.8.0/bin/java 按 Enter 保留当前选项[+]，或者键入选项编号：2
这个时候就只有2条记录了。我们来检查一下版本切换生效没有。
[root@server01 java]# java -version java version "1.8.0_271" Java(TM) SE Runtime Environment (build 1.8.0_271-b09) Java HotSpot(TM) 64-Bit Server VM (build 25.271-b09, mixed mode) [root@server01 java]# which java /usr/bin/java [root@server01 java]# ll /usr/bin/java lrwxrwxrwx. 1 root root 22 10月 23 13:11 /usr/bin/java -> /etc/alternatives/java [root@server01 java]# ll /etc/alternatives/java lrwxrwxrwx. 1 root root 27 10月 23 13:11 /etc/alternatives/java -> /usr/java/jdk1.8.0/bin/java
完美完成Java版本切换。接下来我们就能继续Hive环境搭建了。
启动zookeeper集群和hadoop集群就不用说了，轻车熟路。
在server03上启动hive元数据服务，再次提示，这是个阻塞式窗口，不用等它退出到命令行界面
[hadoop@server03 bin]$ hive --service metastore 2020-10-23 15:48:01: Starting Hive Metastore Server SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/apache-hive-3.1.2/lib/log4j-slf4j-impl-2.10.0.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/hadoop-3.3.0/share/hadoop/common/lib/slf4j-log4j12-1.7.25.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.apache.logging.slf4j.Log4jLoggerFactory]
在server02上启动hive客户端，出现hive提示符就可以尽情玩耍了。
[hadoop@server02 ~]$ hive which: no hbase in (/usr/local/bin:/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/usr/java/jdk1.8.0/bin:/usr/apache-zookeeper-3.5.8/bin:/usr/hadoop-3.3.0/bin:/usr/hadoop-3.3.0/sbin:/usr/apache-hive-3.1.2/bin:/home/hadoop/.local/bin:/home/hadoop/bin) SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/apache-hive-3.1.2/lib/log4j-slf4j-impl-2.10.0.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/hadoop-3.3.0/share/hadoop/common/lib/slf4j-log4j12-1.7.25.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.apache.logging.slf4j.Log4jLoggerFactory] Hive Session ID = d252045d-6dc8-41ee-9808-c324521a880f Logging initialized using configuration in jar:file:/usr/apache-hive-3.1.2/lib/hive-common-3.1.2.jar!/hive-log4j2.properties Async: true Hive-on-MR is deprecated in Hive 2 and may not be available in the future versions. Consider using a different execution engine (i.e. spark, tez) or using Hive 1.X releases. Hive Session ID = c39b1a09-a43a-4ccb-bcf9-9608969cbd12 hive>
数据库操作，简直跟mysql一模一样
hive> show databases; OK default Time taken: 4.862 seconds, Fetched: 1 row(s) hive> create database test; OK Time taken: 4.09 seconds hive> show databases; OK default test Time taken: 0.099 seconds, Fetched: 2 row(s) hive> use test; OK Time taken: 0.138 seconds hive>
这里我们再试验一下Hadoop里的例子，数据稍微做一下调整，存放在/opt/hive_remote/warehouse/data1中。
气温文件，年-月-日-温度 1949-10-01-34 1949-10-01-38 1949-10-02-36 1950-01-01-32 1950-10-01-37 1951-12-01-23 1950-10-02-41 1950-10-03-27 1951-07-01-45 1951-07-02-46 1951-07-03-47
在hive里创建一张表来保存数据。
hive> create table hive1( > year string, > month string, > day string, > temperature smallint > ) > row format delimited > fields terminated by ‘-‘; OK Time taken: 3.818 seconds
把数据导入表里
hive> load data local inpath ‘/opt/hive_remote/warehouse/data1‘into table hive1; Loading data to table test.hive1 OK Time taken: 15.852 seconds hive> select * from hive1; OK 1949 10 01 34 1949 10 01 38 1949 10 02 36 1950 01 01 32 1950 10 01 37 1951 12 01 23 1950 10 02 41 1950 10 03 27 1951 07 01 45 1951 07 02 46 1951 07 03 47 Time taken: 12.108 seconds, Fetched: 11 row(s)
好了，数据准备完毕。还是那个需求，从中找出每个月温度最高的两天所对应的气温。先建一张中间表，把一天内有多条记录的去掉，只保留最高温。
hive> create table hive1_mid( > year string, > month string, > day string, > temperature smallint > ); OK Time taken: 1.627 seconds hive> insert into hive1_mid select year,month,day,max(temperature) from hive1 group by year,month,day; Query ID = hadoop_20201101093442_c624ad0c-4b55-406d-b49e-e8aa0315e7a2 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_1604193604905_0002, Tracking URL = http://server01:8088/proxy/application_1604193604905_0002/ Kill Command = /usr/hadoop-3.3.0/bin/mapred job -kill job_1604193604905_0002 Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 1 2020-11-01 09:35:35,506 Stage-1 map = 0%, reduce = 0% 2020-11-01 09:36:41,001 Stage-1 map = 0%, reduce = 0% 2020-11-01 09:36:50,906 Stage-1 map = 100%, reduce = 0%, Cumulative CPU 2.59 sec 2020-11-01 09:37:52,195 Stage-1 map = 100%, reduce = 0%, Cumulative CPU 2.59 sec 2020-11-01 09:38:00,629 Stage-1 map = 100%, reduce = 67%, Cumulative CPU 4.87 sec 2020-11-01 09:38:05,778 Stage-1 map = 100%, reduce = 100%, Cumulative CPU 5.72 sec MapReduce Total cumulative CPU time: 5 seconds 720 msec Ended Job = job_1604193604905_0002 Loading data to table test.hive1_mid 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_1604193604905_0003, Tracking URL = http://server01:8088/proxy/application_1604193604905_0003/ Kill Command = /usr/hadoop-3.3.0/bin/mapred job -kill job_1604193604905_0003 Hadoop job information for Stage-3: number of mappers: 1; number of reducers: 1 2020-11-01 09:39:00,431 Stage-3 map = 0%, reduce = 0% 2020-11-01 09:40:08,674 Stage-3 map = 0%, reduce = 0% 2020-11-01 09:40:17,031 Stage-3 map = 100%, reduce = 0%, Cumulative CPU 1.67 sec 2020-11-01 09:40:52,565 Stage-3 map = 100%, reduce = 100%, Cumulative CPU 4.37 sec MapReduce Total cumulative CPU time: 4 seconds 370 msec Ended Job = job_1604193604905_0003 MapReduce Jobs Launched: Stage-Stage-1: Map: 1 Reduce: 1 Cumulative CPU: 5.72 sec HDFS Read: 16922 HDFS Write: 502 SUCCESS Stage-Stage-3: Map: 1 Reduce: 1 Cumulative CPU: 4.37 sec HDFS Read: 12824 HDFS Write: 327 SUCCESS Total MapReduce CPU Time Spent: 10 seconds 90 msec OK Time taken: 400.594 seconds hive> select * from hive1_mid; OK 1949 10 01 38 1949 10 02 36 1950 01 01 32 1950 10 01 37 1950 10 02 41 1950 10 03 27 1951 07 01 45 1951 07 02 46 1951 07 03 47 1951 12 01 23 Time taken: 7.142 seconds, Fetched: 10 row(s)
接下来的任务就简单了，从中间表中按照年月日分组，获取每组记录里温度最高的两条。
hive> select a.year,a.month,a.day,a.temperature from hive1_mid a > left join hive1_mid b > on a.year=b.year AND a.temperature>=b.temperature > group by a.year,a.month,a.day,a.temperature > having count(*)<=3 > order by a.year,a.month,a.day,a.temperature desc; Query ID = hadoop_20201101143949_50884d85-1bc6-4009-915f-9c7ce4f50284 Total jobs = 2 Execution completed successfully MapredLocal task succeeded 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_1604209171036_0001, Tracking URL = http://server01:8088/proxy/application_1604209171036_0001/ Kill Command = /usr/hadoop-3.3.0/bin/mapred job -kill job_1604209171036_0001 Hadoop job information for Stage-2: number of mappers: 1; number of reducers: 1 2020-11-01 14:43:31,432 Stage-2 map = 0%, reduce = 0% 2020-11-01 14:44:29,421 Stage-2 map = 100%, reduce = 0%, Cumulative CPU 2.6 sec 2020-11-01 14:45:30,094 Stage-2 map = 100%, reduce = 0%, Cumulative CPU 2.6 sec 2020-11-01 14:45:49,013 Stage-2 map = 100%, reduce = 67%, Cumulative CPU 3.58 sec 2020-11-01 14:46:29,415 Stage-2 map = 100%, reduce = 100%, Cumulative CPU 6.23 sec MapReduce Total cumulative CPU time: 6 seconds 230 msec Ended Job = job_1604209171036_0001 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_1604209171036_0002, Tracking URL = http://server01:8088/proxy/application_1604209171036_0002/ Kill Command = /usr/hadoop-3.3.0/bin/mapred job -kill job_1604209171036_0002 Hadoop job information for Stage-3: number of mappers: 1; number of reducers: 1 2020-11-01 14:48:37,771 Stage-3 map = 0%, reduce = 0% 2020-11-01 14:49:30,312 Stage-3 map = 100%, reduce = 0%, Cumulative CPU 2.31 sec 2020-11-01 14:50:14,041 Stage-3 map = 100%, reduce = 67%, Cumulative CPU 4.94 sec 2020-11-01 14:50:16,105 Stage-3 map = 100%, reduce = 100%, Cumulative CPU 5.18 sec MapReduce Total cumulative CPU time: 5 seconds 180 msec Ended Job = job_1604209171036_0002 MapReduce Jobs Launched: Stage-Stage-2: Map: 1 Reduce: 1 Cumulative CPU: 6.23 sec HDFS Read: 13325 HDFS Write: 336 SUCCESS Stage-Stage-3: Map: 1 Reduce: 1 Cumulative CPU: 5.18 sec HDFS Read: 8508 HDFS Write: 295 SUCCESS Total MapReduce CPU Time Spent: 11 seconds 410 msec OK 1949 10 01 38 1949 10 02 36 1950 01 01 32 1950 10 01 37 1950 10 03 27 1951 07 01 45 1951 07 02 46 1951 12 01 23 Time taken: 632.561 seconds, Fetched: 8 row(s)
上面的操作步骤跟hadoop高可用环境搭建里的map/reduce代码比起来是不是简洁多了？hive帮我们把sql语句转换成了对应的map/reduce任务，而不需要我们操心，对用户来说用好多了。
原文：https://www.cnblogs.com/burningblade/p/13642673.html