Spark SQL 连接 Hive

Table of Contents

环境

准备测试数据

依赖包准备

第一种 metastore service 已启动

第二种 Metastore Service 未启动

---

环境

Hadoop 2.10.0

Spark version： 2.4.5

Hive version: 3.1.2

 

准备测试数据

> beeline -u jdbc:hive2://localhost:10000 -n root

create table default.member_phone as select '00001' as member_srl, '136****3896' as phone_num;

 

依赖包准备

compile 'org.apache.spark:spark-sql_2.12:2.4.5'
compile 'org.apache.spark:spark-hive_2.12:2.4.5'

 

根据Hive是否启动metastore service有两种不同的连接方式。

第一种 metastore service 已启动

若已经启动metastore service，则spark sql直接做为metastore service的thrift client来访问metastore数据。

.config("hive.metastore.uris", "thrift://localhost:9083")

为SparkSession对象添加上面的metastore 服务器地址即可。

 

测试代码如下：

import org.apache.spark.sql.SparkSession;

public class SparkHive {

    public static void main(String[] args) {

        SparkSession session = null;

        try {
            session = SparkSession.builder()
                    .appName("spark hive")
                    .master("local[*]")
                    .config("hive.metastore.uris", "thrift://localhost:9083")
                    .enableHiveSupport()
                    .getOrCreate();

            session.sql("select * from default.member_phone")
                    .show();

        } finally {
            if (session != null) {
                session.close();
            }
        }

    }
}

 

第二种 Metastore Service 未启动

若metastore service 以embedded 模式作为hive server的一部分运行的，那它就不能独立提供metastore service。就没有9083端口的thrift metastore service. 那么这时就需要spark sql 自己根据metastore database的配置创建metastore server访问hive的meta数据。这里的metastore database 应该是remote模式运行的。

 

配置文件 resources/hive-site.xml

<configuration>

    <!-- database configuration -->
    <property>
        <name>javax.jdo.option.ConnectionDriverName</name>
        <value>com.mysql.jdbc.Driver</value>
        <description>Driver class name for a JDBC metastore</description>
    </property>

    <property>
        <name>javax.jdo.option.ConnectionURL</name>
        <value>jdbc:mysql://localhost:3306/metastore?useSSL=false</value>
        <description>
            JDBC connect string for a JDBC metastore.
            To use SSL to encrypt/authenticate the connection, provide database-specific SSL flag in the connection URL.
            For example, jdbc:postgresql://myhost/db?ssl=true for postgres database.
        </description>
    </property>

    <property>
        <name>javax.jdo.option.ConnectionUserName</name>
        <value>root</value>
        <description>Username to use against metastore database</description>
    </property>

    <property>
        <name>javax.jdo.option.ConnectionPassword</name>
        <value>abc123</value>
        <description>password to use against metastore database</description>
    </property>

</configuration>

加入mysql的driver jar

compile 'mysql:mysql-connector-java:5.1.41'

 

Java测试代码

import org.apache.spark.sql.SparkSession;

public class SparkHive {

    public static void main(String[] args) {

        SparkSession session = null;

        try {
            session = SparkSession.builder()
                    .appName("spark hive")
                    .master("local[*]")
                    .enableHiveSupport()
                    .getOrCreate();

            session.sql("select * from default.member_phone")
                    .show();

        } finally {
            if (session != null) {
                session.close();
            }
        }

    }
}

打印出数据库里的信息就表示访问成功。

+----------+-----------+
|member_srl|  phone_num|
+----------+-----------+
|      0001|136****3896|
+----------+-----------+

 

hive的不同启动模式参考：https://blog.csdn.net/adorechen/article/details/104934057

 

常见错误问题

Unable to instantiate SparkSession with Hive support because Hive classes are not found.

没有找到spark-hive相关的包，加入spark-hive依赖。

compile 'org.apache.spark:spark-hive_2.12:2.4.5'

 

参考文章：

http://spark.apache.org/docs/latest/sql-data-sources-hive-tables.html