系统环境:centos7
软件版本:jdk1.8、zookeeper3.4.8、hadoop2.8.5、hive1.1.0
一、下载安装
下载hive安装包,上传到linux服务器上,
解压安装包
tar -zxvf apache-hive-1.1.0-bin.tar.gz -C /home/local/
重命名文件
mv apache-hive-1.1.0-bin/ hive
二、Hive部署
配置Hive环境
vi /etc/profile
添加如下配置
#hive
export HIVE_HOME=/home/local/hive
export PATH=$PATH:${HIVE_HOME}/bin
数据库设置
mysql -uroot -p
#创建数据库
create database metastore;
#关闭只读
set global read_only=0;
#设置密码安全策略
set global validate_password_policy=0;
#设置密码长度
set global validate_password_length=4;
#更改密码
ALTER USER 'root'@'localhost' IDENTIFIED BY 'root';
#数据库授权
grant all on metastore.* to hive@'%' identified by 'hive';
GRANT ALL PRIVILEGES ON *.* TO 'root'@'%' IDENTIFIED BY 'root';
#刷新权限
flush privileges;
#退出mysql
exit
上传mysql-connector到Hive库目录
cp mysql-connector-java-5.1.49.jar /home/local/hive/lib/
修改hive配置文件
复制hive-env.sh
cp /home/local/hive/conf/hive-env.sh.template /home/local/hive/conf/hive-env.sh
修改hive-env.sh文件,参考如下内容:
vim /home/local/hive/conf/hive-env.sh
export JAVA_HOME=/home/local/java
export HADOOP_HOME=/home/local/hadoop
export HIVE_HOME=/home/local/hive
export HIVE_CONF_DIR=/home/local/hive/conf
复制hive-site.xml
cp /home/local/hive/conf/hive-default.xml.template /home/local/hive/conf/hive-site.xml
修改hive-site.xml
property>
<name>hive.exec.scratchdir</name>
<value>/user/hive/tmp</value>
</property>
<property>
<name>hive.metastore.warehouse.dir</name>
<value>/user/hive/warehouse</value>
</property>
<property>
<name>hive.querylog.location</name>
<value>/user/hive/log</value>
</property>
<property>
<name>javax.jdo.option.ConnectionURL</name>
<value>jdbc:mysql://master:3306/metastore?createDatabaseIfNotExist=true&characterEncoding=UTF-8&useSSL=false</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>root</value>
</property>
初始化hive元数据库
schematool -dbType mysql -initSchema root root
分发hive至每个节点服务器
for i in {1..2};do scp -r /home/local/hive/ root@slave${i}:/home/local/;done
三、Hive测试
启动Hive
启动mysql
systemctl start mysql
hive
hive创建数据库
create database testdb;
创建表
drop table id exists testdb.score;
create table testdb.score
(name string comment "名称",
gender string comment "性别",
score int comment "分数")
row format delimited fields terminated by '\t'
lines terminated by '\n'
stored as textfile;
删除表
drop table if exists table;
本地文件导入hive
load data local inpath '' into table ''
hdfs文件导入hive
在/opt/hive/data/目录下新建stu_score.txt文件,添加如下内容
jone male 30
mike male 40
sunny female 50
文件上传到hdfs上
hdfs dfs -put /opt/hive/data/stu_score.txt /hive/warehouse
hdfs上传到hive表中
load data inpath '/hive/warehouse/stu_score.txt' into table score;
exit;
Java程序将hdfs文件上传到hive
导入依赖
<!--hadoop-->
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-client</artifactId>
<version>2.8.5</version>
</dependency>
<!--hive-->
<dependency>
<groupId>org.apache.hive</groupId>
<artifactId>hive-jdbc</artifactId>
<version>1.1.0</version>
</dependency>
编写测试用例
public class HiveTest {
public static void main(String[] args) throws SQLException {
HdfsFileUploader("input/stu_score.txt", "/hive/warehouse");
HdfsToHiveUploader("/hive/warehouse/stu_score.txt", "score");
}
/**
* 使用sql查询数据库 统计时间 测试性能
*
* @param sql
* @return
*/
public static List getDataList(String sql) {
long start = System.currentTimeMillis();
List list = new ArrayList();
Connection connection = getConnection();
try {
Statement statement = connection.createStatement();
ResultSet resultSet = statement.executeQuery(sql);
ResultSetMetaData metaData = resultSet.getMetaData();
int columnCount = metaData.getColumnCount();
while (resultSet.next()) {
Map map = new HashMap<>();
for (int i = 1; i <= columnCount; i++) {
// 获取字段名称 metaData.getColumnName(i)
map.put(metaData.getColumnName(i), resultSet.getObject(i));
}
list.add(map);
// Process the result set
}
resultSet.close();
statement.close();
connection.close();
} catch (Exception e) {
e.printStackTrace();
}
long end = System.currentTimeMillis();
System.out.println("本次查询耗时" + (end - start) / 1000 + "秒");
return list;
}
/**
* 获取数据库连接对象
*
* @return
*/
public static Connection getConnection() {
Connection connection = null;
try {
// Load Hive JDBC driver
Class.forName("org.apache.hive.jdbc.HiveDriver");
// Establish connection to Hive
connection = DriverManager.getConnection("jdbc:hive2://192.168.245.200:10000/testdb", "root", "root");
} catch (Exception e) {
e.printStackTrace();
}
return connection;
}
/**
* 本地文件上传到hdfs
*/
public static void HdfsFileUploader(String localFilePath, String hdfsFilePath) {
Configuration configuration = new Configuration();
configuration.set("fs.defaultFS", "hdfs://192.168.245.200:9000"); // 设置HDFS的URL
try {
FileSystem fileSystem = FileSystem.get(configuration);
fileSystem.copyFromLocalFile(new Path(localFilePath), new Path(hdfsFilePath));
System.out.println("File uploaded to HDFS successfully.");
} catch (IOException e) {
e.printStackTrace();
}
}
/**
* hdfs上传到hive
*/
public static void HdfsToHiveUploader(String hdfsFilePath, String tableName) {
try {
Connection connection = getConnection();
Statement statement = connection.createStatement();
// Create temporary external table in Hive
String createTableQuery = "CREATE EXTERNAL TABLE IF NOT EXISTS " + tableName + " (City STRING, City_Admaster STRING,City_EN STRING,Province STRING,Province_EN STRING,Region STRING,Tier STRING) " + "ROW FORMAT DELIMITED FIELDS TERMINATED BY ',' " + "STORED AS TEXTFILE " + "LOCATION '/hive/warehouse/china_city_list'";
statement.execute(createTableQuery);
// Load data from HDFS to Hive table
String loadDataQuery = "LOAD DATA INPATH '" + hdfsFilePath + "' OVERWRITE INTO TABLE " + tableName;
statement.execute(loadDataQuery);
System.out.println("File uploaded from HDFS to Hive successfully.");
// Close connection and statement
statement.close();
connection.close();
} catch (Exception e) {
e.printStackTrace();
}
}
}
四、总结
本次实验中,Hive采用MySQL数据库保存Hive的元数据,而不是采用Hive自带的derby来存储元数据。
Hive常用的HiveQL操作命令主要包括:数据定义、数据操作。
Hive实现最大的优势是,对于非程序员,不用学习编写Java MapReduce代码,也可以完成MapReduce任务。