Sqoop-1.4.7-部署与常见案例

该文章是基于 Hadoop2.7.6_01_部署 、 Hive-1.2.1_01_安装部署 进行的

1. 前言

在一个完整的大数据处理系统中，除了hdfs+mapreduce+hive组成分析系统的核心之外，还需要数据采集、结果数据导出、任务调度等不可或缺的辅助系统，而这些辅助工具在hadoop生态体系中都有便捷的开源框架，如图所示：

1.1. 概述

sqoop是apache旗下一款“Hadoop和关系数据库服务器之间传送数据”的工具。

　　导入数据：MySQL，Oracle导入数据到Hadoop的HDFS、HIVE、HBASE等数据存储系统；

　　导出数据：从Hadoop的文件系统中导出数据到关系数据库

1.3. 工作机制

将导入或导出命令翻译成mapreduce程序来实现

在翻译出的mapreduce中主要是对inputformat和outputformat进行定制

2. Sqoop的安装部署

2.1. 软件部署

 [yun@mini01 software]$ pwd
 /app/software
 [yun@mini01 software]$ tar xf sqoop-1.4..bin__hadoop-2.6..tar.gz
 [yun@mini01 software]$ mv sqoop-1.4..bin__hadoop-2.6. /app/sqoop-1.4.
 [yun@mini01 software]$ cd /app/
 [yun@mini01 ~]$ ln -s sqoop-1.4./ sqoop
 [yun@mini01 ~]$ ll
 total
 …………
 lrwxrwxrwx   yun yun     Aug   : sqoop -> sqoop-1.4./
 drwxr-xr-x   yun yun    Dec    sqoop-1.4.

2.2. 配置修改

 [yun@mini01 conf]$ pwd
 /app/sqoop/conf
 [yun@mini01 conf]$ ll
 total
 -rw-rw-r--  yun yun  Dec    oraoop-site-template.xml
 -rw-rw-r--  yun yun  Dec    sqoop-env-template.cmd
 -rwxr-xr-x  yun yun  Dec    sqoop-env-template.sh
 -rw-rw-r--  yun yun  Dec    sqoop-site-template.xml
 -rw-rw-r--  yun yun  Dec    sqoop-site.xml
 [yun@mini01 conf]$ cp -a sqoop-env-template.sh sqoop-env.sh
 [yun@mini01 conf]$ cat sqoop-env.sh
 # Licensed to the Apache Software Foundation (ASF) under one or more
 ………………

 # Set Hadoop-specific environment variables here.

 #Set path to where bin/hadoop is available  # 修改的配置
 export HADOOP_COMMON_HOME=${HADOOP_HOME}

 #Set path to where hadoop-*-core.jar is available  # 修改的配置
 export HADOOP_MAPRED_HOME=${HADOOP_HOME}

 #set the path to where bin/hbase is available
 #export HBASE_HOME=

 #Set the path to where bin/hive is available  # 修改的配置
 export HIVE_HOME=${HIVE_HOME}

 #Set the path for where zookeper config dir is
 #export ZOOCFGDIR=

2.3. 加入mysql的jdbc驱动包

 [yun@mini01 software]$ pwd
 /app/software
 [yun@mini01 software]$ cp -a mysql-connector-java-5.1..jar /app/sqoop/lib 

2.4. 加入hive的执行包

 [yun@mini01 lib]$ pwd
 /app/hive/lib
 [yun@mini01 lib]$ cp -a hive-exec-1.2..jar /app/sqoop/lib/

避免出现

2.5. 验证启动

 [yun@mini01 bin]$ pwd
 /app/sqoop/bin
 [yun@mini01 bin]$ ./sqoop-version
 // :: INFO sqoop.Sqoop: Running Sqoop version: 1.4.
 Sqoop 1.4.
 git commit id 2328971411f57f0cb683dfb79d19d4d19d185dd8
 Compiled by maugli on Thu Dec  :: STD
 [yun@mini01 bin]$ ./sqoop help  # 查看帮助
 // :: INFO sqoop.Sqoop: Running Sqoop version: 1.4.
 usage: sqoop COMMAND [ARGS]

 Available commands:
   codegen            Generate code to interact with database records
   create-hive-table  Import a table definition into Hive
   eval               Evaluate a SQL statement and display the results
   export             Export an HDFS directory to a database table
   help               List available commands
   import             Import a table from a database to HDFS
   import-all-tables  Import tables from a database to HDFS
   import-mainframe   Import datasets from a mainframe server to HDFS
   job                Work with saved jobs
   list-databases     List available databases on a server
   list-tables        List available tables in a database
   merge              Merge results of incremental imports
   metastore          Run a standalone Sqoop metastore
   version            Display version information

 See 'sqoop help COMMAND' for information on a specific command.

3. 数据库信息

 # 建库
 CREATE DATABASE sqoop_test DEFAULT CHARACTER SET utf8 ;
 # 建账号  数据库在mini03机器上
 grant all on sqoop_test.* to sqoop_test@'%' identified by 'sqoop_test';
 grant all on sqoop_test.* to sqoop_test@'mini03' identified by 'sqoop_test';
 # 刷新权限
 flush privileges;

3.1. 表信息

表emp:

id	name	deg	salary	dept
1201	gopal	manager	50000	TP
1202	manisha	Proof reader	50000	TP
1203	khalil	php dev	30000	AC
1204	prasanth	php dev	30000	AC
1205	kranthi	admin	20000	TP

表emp_add:

id	name	deg	salary
1201	288A	vgiri	jublee
1202	108I	aoc	sec-bad
1203	144Z	pgutta	hyd
1204	78B	old city	sec-bad
1205	720X	hitec	sec-bad

表emp_conn:

id	name	deg
1201	2356742	gopal@tp.com
1202	1661663	manisha@tp.com
1203	8887776	khalil@ac.com
1204	9988774	prasanth@ac.com
1205	1231231	kranthi@tp.com

4. Sqoop的数据导入

“导入工具”导入单个表从RDBMS到HDFS。表中的每一行被视为HDFS的记录。所有记录都存储为文本文件的文本数据（或者Avro、sequence文件等二进制数据）

 $ sqoop import (generic-args) (import-args)
 $ sqoop-import (generic-args) (import-args)

4.1. 导入表数据到HDFS

 [yun@mini01 sqoop]$ pwd
 /app/sqoop
 [yun@mini01 sqoop]$ bin/sqoop import   \
   --connect jdbc:mysql://mini03:3306/sqoop_test   \
   --username sqoop_test  \
   --password sqoop_test   \
   --table emp   \
   --m
 // :: INFO sqoop.Sqoop: Running Sqoop version: 1.4.
 // :: WARN tool.BaseSqoopTool: Setting your password on the command-line is insecure. Consider using -P instead.
 ………………
 // :: INFO mapreduce.Job:  map % reduce %
 // :: INFO mapreduce.Job: Job job_1533196573365_0001 completed successfully
 ………………
 // :: INFO mapreduce.ImportJobBase: Transferred  bytes in 47.2865 seconds (3.1933 bytes/sec)
 // :: INFO mapreduce.ImportJobBase: Retrieved  records.

查看导入的数据

 [yun@mini02 ~]$ hadoop fs -ls /user/yun/emp
 Found  items
 -rw-r--r--    yun supergroup           -- : /user/yun/emp/_SUCCESS
 -rw-r--r--    yun supergroup         -- : /user/yun/emp/part-m-
 [yun@mini02 ~]$ hadoop fs -cat /user/yun/emp/part-m-
 ,gopal,manager,,TP
 ,manisha,Proof reader,,TP
 ,khalil,php dev,,AC
 ,prasanth,php dev,,AC
 ,kranthi,admin,,TP

4.2. 导入表到HDFS指定目录

 [yun@mini01 sqoop]$ pwd
 /app/sqoop
 [yun@mini01 sqoop]$ bin/sqoop import --connect jdbc:mysql://mini03:3306/sqoop_test \
   --username sqoop_test --password sqoop_test \
   --target-dir /sqoop_test/table_emp/queryresult \
   --table emp --num-mappers 

　　注意：如果没有目录，那么会创建

查看导入的数据

 [yun@mini02 ~]$ hadoop fs -ls /sqoop_test/table_emp/queryresult
 Found  items
 -rw-r--r--    yun supergroup           -- : /sqoop_test/table_emp/queryresult/_SUCCESS
 -rw-r--r--    yun supergroup         -- : /sqoop_test/table_emp/queryresult/part-m-
 [yun@mini02 ~]$ hadoop fs -cat /sqoop_test/table_emp/queryresult/part-m-
 ,gopal,manager,,TP
 ,manisha,Proof reader,,TP
 ,khalil,php dev,,AC
 ,prasanth,php dev,,AC
 ,kranthi,admin,,TP

4.3. 导入关系表到HIVE

 [yun@mini01 sqoop]$ pwd
 /app/sqoop
 [yun@mini01 sqoop]$ bin/sqoop import --connect jdbc:mysql://mini03:3306/sqoop_test \
   --username sqoop_test --password sqoop_test \
   --table emp --hive-import \
   --num-mappers
 // :: INFO sqoop.Sqoop: Running Sqoop version: 1.4.
 // :: WARN tool.BaseSqoopTool: Setting your password on the command-line is insecure. Consider using -P instead.
 ………………
 // :: INFO mapreduce.ImportJobBase: Transferred  bytes in 20.6744 seconds (7.3037 bytes/sec)
 // :: INFO mapreduce.ImportJobBase: Retrieved  records.
 // :: INFO mapreduce.ImportJobBase: Publishing Hive/Hcat import job data to Listeners for table emp
 // :: INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM `emp` AS t LIMIT
 // :: WARN hive.TableDefWriter: Column salary had to be cast to a less precise type in Hive
 // :: INFO hive.HiveImport: Loading uploaded data into Hive
 // :: INFO hive.HiveImport:
 // :: INFO hive.HiveImport: Logging initialized using configuration in jar:file:/app/sqoop-1.4./lib/hive-exec-1.2..jar!/hive-log4j.properties
 // :: INFO hive.HiveImport: OK
 // :: INFO hive.HiveImport: Time taken: 1.677 seconds
 // :: INFO hive.HiveImport: Loading data to table default.emp
 // :: INFO hive.HiveImport: Table default.emp stats: [numFiles=, totalSize=]
 // :: INFO hive.HiveImport: OK
 // :: INFO hive.HiveImport: Time taken: 0.629 seconds
 // :: INFO hive.HiveImport: Hive import complete.
 // :: INFO hive.HiveImport: Export directory is contains the _SUCCESS file only, removing the directory.

查看导入的数据

 hive (default)> show tables;
 OK
 emp
 Time taken: 0.031 seconds, Fetched:  row(s)
 hive (default)> select * from emp;
 OK
     gopal    manager    50000.0    TP
     manisha    Proof reader    50000.0    TP
     khalil    php dev    30000.0    AC
     prasanth    php dev    30000.0    AC
     kranthi    admin    20000.0    TP
 Time taken: 0.489 seconds, Fetched:  row(s)

4.4. 导入表到HIVE指定库指定表

 [yun@mini01 sqoop]$ pwd
 /app/sqoop
 [yun@mini01 sqoop]$ bin/sqoop import --connect jdbc:mysql://mini03:3306/sqoop_test \
   --username sqoop_test --password sqoop_test \
   --table emp \
   --delete-target-dir \
   --fields-terminated-by '\t' \
   --hive-import \
   --hive-database sqoop_test \
   --hive-table hive_emp \
   --num-mappers 

注意：hive的库sqoop_test，必须先建立。否则会报：FAILED: SemanticException [Error 10072]: Database does not exist: sqoop_test

查看导入的数据

 : jdbc:hive2://mini01:10000> use sqoop_test;
 No rows affected (0.049 seconds)
 : jdbc:hive2://mini01:10000> show tables;
 +-----------+--+
 | tab_name  |
 +-----------+--+
 | hive_emp  |
 +-----------+--+
  row selected (0.076 seconds)
 : jdbc:hive2://mini01:10000> select * from hive_emp;
 +--------------+----------------+---------------+------------------+----------------+--+
 | hive_emp.id  | hive_emp.name  | hive_emp.deg  | hive_emp.salary  | hive_emp.dept  |
 +--------------+----------------+---------------+------------------+----------------+--+
 |          | gopal          | manager       | 50000.0          | TP             |
 |          | manisha        | Proof reader  | 50000.0          | TP             |
 |          | khalil         | php dev       | 30000.0          | AC             |
 |          | prasanth       | php dev       | 30000.0          | AC             |
 |          | kranthi        | admin         | 20000.0          | TP             |
 +--------------+----------------+---------------+------------------+----------------+--+
  rows selected (0.162 seconds)

4.5. 导入表数据子集

4.5.1. where子句的导入

 [yun@mini01 sqoop]$ pwd
 /app/sqoop
 [yun@mini01 sqoop]$ bin/sqoop import --connect jdbc:mysql://mini03:3306/sqoop_test \
   --username sqoop_test --password sqoop_test \
   --table emp_add \
   --where "city ='sec-bad'" \
   --target-dir /sqoop_test/table_emp/queryresult2 \
   --num-mappers 

查看导入的数据

 [yun@mini02 ~]$ hadoop fs -ls /sqoop_test/table_emp/queryresult2
 Found  items
 -rw-r--r--    yun supergroup           -- : /sqoop_test/table_emp/queryresult2/_SUCCESS
 -rw-r--r--    yun supergroup          -- : /sqoop_test/table_emp/queryresult2/part-m-
 [yun@mini02 ~]$ hadoop fs -cat /sqoop_test/table_emp/queryresult2/part-m-
 ,108I,aoc,sec-bad
 ,78B,old city,sec-bad
 ,720X,hitec,sec-bad

4.5.2. query按需导入

 [yun@mini01 sqoop]$ pwd
 /app/sqoop
 [yun@mini01 sqoop]$ bin/sqoop import --connect jdbc:mysql://mini03:3306/sqoop_test \
   --username sqoop_test --password sqoop_test \
   --query 'select id,name,deg from emp WHERE  id>1203 and $CONDITIONS' \
   --split-by id \
   --fields-terminated-by '\t' \
   --target-dir /sqoop_test/table_emp/queryresult4 \
   --num-mappers 

查看导入的数据

 [yun@mini02 ~]$ hadoop fs -cat /sqoop_test/table_emp/queryresult4/part-m-
     prasanth    php dev
     kranthi    admin

4.6. 增量导入

所需参数

 --check-column (col)
 --incremental (mode)        # mode include append and lastmodified
 --last-value (value)

示例

 [yun@mini01 sqoop]$ pwd
 /app/sqoop
 [yun@mini01 sqoop]$ bin/sqoop import --connect jdbc:mysql://mini03:3306/sqoop_test \
   --username sqoop_test --password sqoop_test \
   --table emp \
   --incremental append \
   --check-column id \
   --last-value  \
   --fields-terminated-by '\t' \
   --target-dir /sqoop_test/table_emp/queryresult4 \
   --num-mappers 

查看导入的数据

 [yun@mini02 ~]$ hadoop fs -ls /sqoop_test/table_emp/queryresult4/
 Found  items
 -rw-r--r--    yun supergroup           -- : /sqoop_test/table_emp/queryresult4/_SUCCESS
 -rw-r--r--    yun supergroup          -- : /sqoop_test/table_emp/queryresult4/part-m-
 -rw-r--r--    yun supergroup          -- : /sqoop_test/table_emp/queryresult4/part-m-
 [yun@mini02 ~]$
 [yun@mini02 ~]$ hadoop fs -cat /sqoop_test/table_emp/queryresult4/part-m-
     prasanth    php dev
     kranthi    admin
 [yun@mini02 ~]$ hadoop fs -cat /sqoop_test/table_emp/queryresult4/part-m-
     khalil    php dev        AC
     prasanth    php dev        AC
     kranthi    admin        TP

5. Sqoop的数据导出

将数据从HDFS导出到RDBMS数据库

导出前，目标表必须存在于目标数据库中。

• 默认操作是从将文件中的数据使用INSERT语句插入到表中
• 更新模式下，是生成UPDATE语句更新表数据

 $ sqoop export (generic-args) (export-args)
 $ sqoop-export (generic-args) (export-args)

5.1. 示例

数据

 [yun@mini02 ~]$ hadoop fs -ls /sqoop_test/table_emp/queryresult
 Found  items
 -rw-r--r--    yun supergroup           -- : /sqoop_test/table_emp/queryresult/_SUCCESS
 -rw-r--r--    yun supergroup         -- : /sqoop_test/table_emp/queryresult/part-m-
 [yun@mini02 ~]$ hadoop fs -cat /sqoop_test/table_emp/queryresult/part-m-
 ,gopal,manager,,TP
 ,manisha,Proof reader,,TP
 ,khalil,php dev,,AC
 ,prasanth,php dev,,AC
 ,kranthi,admin,,TP

1、首先需要手动创建mysql中的目标表

 MariaDB [(none)]> use sqoop_test;
 Database changed

 MariaDB [sqoop_test]> CREATE TABLE employee (
           id INT NOT NULL PRIMARY KEY,
           name VARCHAR(),
           deg VARCHAR(),
           salary INT,
           dept VARCHAR());
 Query OK,  rows affected (0.00 sec)

 MariaDB [sqoop_test]> show tables;
 +----------------------+
 | Tables_in_sqoop_test |
 +----------------------+
 | emp                  |
 | emp_add              |
 | emp_conn             |
 | employee             |
 +----------------------+
  rows in set (0.00 sec)

2、然后执行导出命令

 [yun@mini01 sqoop]$ pwd
 /app/sqoop
 [yun@mini01 sqoop]$ bin/sqoop export \
   --connect jdbc:mysql://mini03:3306/sqoop_test \
   --username sqoop_test --password sqoop_test \
   --table employee \
   --export-dir /sqoop_test/table_emp/queryresult/

3、验证表mysql命令行

 MariaDB [sqoop_test]> select * from employee;
 +------+----------+--------------+--------+------+
 | id   | name     | deg          | salary | dept |
 +------+----------+--------------+--------+------+
 |  | gopal    | manager      |   | TP   |
 |  | manisha  | Proof reader |   | TP   |
 |  | khalil   | php dev      |   | AC   |
 |  | prasanth | php dev      |   | AC   |
 |  | kranthi  | admin        |   | TP   |
 +------+----------+--------------+--------+------+
  rows in set (0.00 sec)