Spark2.x学习笔记：Spark SQL快速入门

Spark SQL快速入门

本地表

（1）准备数据

[root@node1 ~]# mkdir /tmp/data
[root@node1 ~]# cat data/ml-1m/users.dat |tr -s "::" "," >> /tmp/data/users.dat
[root@node1 ~]# tail - /tmp/data/users.dat
,F,,,
,F,,,
,F,,,
,F,,,
,M,,,
[root@node1 ~]#

（2）确认HDFS已经启动 
（3）进入spark-sql

[root@node1 ~]# spark-sql
// :: WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
// :: WARN ObjectStore: Version information not found in metastore. hive.metastore.schema.verification is not enabled so recording the schema version 1.2.
// :: WARN ObjectStore: Failed to get database default, returning NoSuchObjectException
// :: WARN ObjectStore: Failed to get database global_temp, returning NoSuchObjectException
spark-sql>

（4）创建表

spark-sql> create external table user(
         > userid int,
         > gender string,
         > age int,
         > occupation string,
         > zipcode int
         > )
         > row format delimited fields terminated by ','
         > stored as textfile
         > location 'file:///tmp/data';
Time taken: 4.032 seconds
spark-sql>

（5）查看表结构

spark-sql> show create table user;
CREATE EXTERNAL TABLE `user`(`userid` int, `gender` string, `age` int, `occupation` string, `zipcode` int)
ROW FORMAT SERDE 'org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe'
WITH SERDEPROPERTIES (
  'field.delim' = ',',
  'serialization.format' = ','
)
STORED AS
  INPUTFORMAT 'org.apache.hadoop.mapred.TextInputFormat'
  OUTPUTFORMAT 'org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat'
LOCATION 'file:/tmp/data'
TBLPROPERTIES (
  'rawDataSize' = '-1',
  'numFiles' = '',
  'transient_lastDdlTime' = '',
  'totalSize' = '',
  'COLUMN_STATS_ACCURATE' = 'false',
  'numRows' = '-1'
)

Time taken: 1.816 seconds, Fetched  row(s)
spark-sql> 

（6）查询本地数据表

spark-sql> select * from user limit ;
   F
   M
   M
   M
   M
   F
   M
   M
   M
  F
Time taken: 2.95 seconds, Fetched  row(s)
spark-sql>

使用spark-sql处理Hive MeteStore中的表

注意，默认HDFS已经启动。

（1）hive-site.xml 
hive可以通过服务的形式对外提供元数据读写操作，修改配置文件 $HIVE_HOME/conf/hive-site.xml，增加如下内容

<property>
         <name>hive.metastore.uris</name>
         <value>thrift://node:9083</value>
</property>

其中，node是启动metastore服务所在的节点。 
我的hive-site.xml内容如下：

[root@node3 ~]# cat /opt/hive-2.1./conf/hive-site.xml
<?xml version="1.0" encoding="UTF-8"?>
<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
<configuration>
     <property>
          <name>hive.metastore.uris</name>
          <value>thrift://node3:9083</value>
          <description>用于metastore客户端连接远程metestore</description>
    </property>
    <property>
          <name>javax.jdo.option.ConnectionURL</name>
          <value>jdbc:mysql://node3:3306/hive?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>
          <!--mysql数据库用户名-->
          <value>root</value>
    </property>
    <property>
          <name>javax.jdo.option.ConnectionPassword</name>
          <value></value>
    </property>
    <property>
          <name>mapred.reduce.tasks</name>
          <value></value>
    </property>

</configuration>
[root@node3 ~]#

（2）启动hive metastore

[root@node3 ~]# hive --service hiveserver2  >/dev/null >/dev/null &
[root@node3 ~]# hive --service metastore  >/dev/null  >&  &

[root@node3 ~]# jobs
[]-  Running                 hive --service hiveserver2 > /dev/null > /dev/null &
[]+  Running                 hive --service metastore > /dev/null >& &
[root@node3 ~]#

（3）将hive的配置文件复制到spark

[root@node3 ~]# scp /opt/hive-2.1./conf/hive-site.xml  node1:/opt/spark-2.2./conf

（4）启动spark-sql

[root@node1 ~]# spark-sql
// :: WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
// :: WARN ObjectStore: Version information not found in metastore. hive.metastore.schema.verification is not enabled so recording the schema version 1.2.
// :: WARN ObjectStore: Failed to get database default, returning NoSuchObjectException
spark-sql>

（5）查看Hive数据表

spark-sql> show tables;
default copyemp false
default demo    false
default dept    false
default dual    false
default emp false
default empbak  false
default employees   false
default mytb    false
default users   false
Time taken: 4.908 seconds, Fetched  row(s)
spark-sql> select * from users;
   aa
   bb
   cc
Time taken: 3.196 seconds, Fetched  row(s)
spark-sql>