簡介：
本篇文章主要介紹sqoop的基本語法及簡單使用方法誓酒。
1.查看命令幫助
[hadoop@hadoop000 ~]$ sqoop help
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.
# 這里提示我們使用sqoop help command（要查詢的命令）進行該命令的詳細(xì)查詢
2.list-databases
# 查看list-databases命令幫助
[hadoop@hadoop000 ~]$ sqoop help list-databases
usage: sqoop list-databases [GENERIC-ARGS] [TOOL-ARGS]
Common arguments:
--connect <jdbc-uri> Specify JDBC connect
string
--connection-manager <class-name> Specify connection manager
class name
--connection-param-file <properties-file> Specify connection
parameters file
--driver <class-name> Manually specify JDBC
driver class to use
--hadoop-home <hdir> Override
$HADOOP_MAPRED_HOME_ARG
--hadoop-mapred-home <dir> Override
$HADOOP_MAPRED_HOME_ARG
--help Print usage instructions
-P Read password from console
--password <password> Set authentication
password
--password-alias <password-alias> Credential provider
password alias
--password-file <password-file> Set authentication
password file path
--relaxed-isolation Use read-uncommitted
isolation for imports
--skip-dist-cache Skip copying jars to
distributed cache
--username <username> Set authentication
username
--verbose Print more information
while working
# 簡單使用
[hadoop@oradb3 ~]$ sqoop list-databases \
> --connect jdbc:mysql://localhost:3306 \
> --username root \
> --password 123456
# 結(jié)果
information_schema
mysql
performance_schema
slow_query_log
sys
test
3.list-tables
# 命令幫助
[hadoop@hadoop000 ~]$ sqoop help list-tables
usage: sqoop list-tables [GENERIC-ARGS] [TOOL-ARGS]
Common arguments:
--connect <jdbc-uri> Specify JDBC connect
string
--connection-manager <class-name> Specify connection manager
class name
--connection-param-file <properties-file> Specify connection
parameters file
--driver <class-name> Manually specify JDBC
driver class to use
--hadoop-home <hdir> Override
$HADOOP_MAPRED_HOME_ARG
--hadoop-mapred-home <dir> Override
$HADOOP_MAPRED_HOME_ARG
--help Print usage instructions
-P Read password from console
--password <password> Set authentication
password
--password-alias <password-alias> Credential provider
password alias
--password-file <password-file> Set authentication
password file path
--relaxed-isolation Use read-uncommitted
isolation for imports
--skip-dist-cache Skip copying jars to
distributed cache
--username <username> Set authentication
username
--verbose Print more information
while working
# 使用方法
[hadoop@hadoop000 ~]$ sqoop list-tables \
> --connect jdbc:mysql://localhost:3306/test \
> --username root \
> --password 123456
# 結(jié)果
t_order
test0001
test_1013
test_dyc
test_tb
4.將mysql導(dǎo)入HDFS中(import)
(默認(rèn)導(dǎo)入當(dāng)前用戶目錄下/user/用戶名/表名)
說到這里擴展一個小知識點：
- hadoop fs -ls 顯示的是當(dāng)前的用戶目錄 即/user/hadoop
hadoop fs -ls / 顯示的是HDFS根目錄
# 查看命令幫助
[hadoop@hadoop000 ~]$ sqoop help import
# 執(zhí)行import
[hadoop@hadoop000 ~]$ sqoop import \
> --connect jdbc:mysql://localhost:3306/test \
> --username root \
> --password 123456 \
> --table students
這時很可能會出現(xiàn)這個錯誤
Exception in thread "main" java.lang.NoClassDefFoundError: org/json/JSONObject
這里我們需要導(dǎo)入java-json.jar包 下載地址 把java-json.jar添加到../sqoop/lib目錄下即可
# 再次執(zhí)行 import導(dǎo)入
[hadoop@hadoop000 ~]$ sqoop import \
> --connect jdbc:mysql://localhost:3306/test \
> --username root \
> --password 123456 \
> --table students
18/07/04 13:28:35 INFO sqoop.Sqoop: Running Sqoop version: 1.4.6-cdh5.7.0
18/07/04 13:28:35 WARN tool.BaseSqoopTool: Setting your password on the command-line is insecure. Consider using -P instead.
18/07/04 13:28:35 INFO manager.MySQLManager: Preparing to use a MySQL streaming resultset.
18/07/04 13:28:35 INFO tool.CodeGenTool: Beginning code generation
18/07/04 13:28:35 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM `students` AS t LIMIT 1
18/07/04 13:28:35 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM `students` AS t LIMIT 1
18/07/04 13:28:35 INFO orm.CompilationManager: HADOOP_MAPRED_HOME is /home/hadoop/app/hadoop-2.6.0-cdh5.7.0
18/07/04 13:28:37 INFO orm.CompilationManager: Writing jar file: /tmp/sqoop-hadoop/compile/3024b8df04f623e8c79ed9b5b30ace75/students.jar
18/07/04 13:28:37 WARN manager.MySQLManager: It looks like you are importing from mysql.
18/07/04 13:28:37 WARN manager.MySQLManager: This transfer can be faster! Use the --direct
18/07/04 13:28:37 WARN manager.MySQLManager: option to exercise a MySQL-specific fast path.
18/07/04 13:28:37 INFO manager.MySQLManager: Setting zero DATETIME behavior to convertToNull (mysql)
18/07/04 13:28:37 INFO mapreduce.ImportJobBase: Beginning import of students
18/07/04 13:28:38 INFO Configuration.deprecation: mapred.jar is deprecated. Instead, use mapreduce.job.jar
18/07/04 13:28:39 INFO Configuration.deprecation: mapred.map.tasks is deprecated. Instead, use mapreduce.job.maps
18/07/04 13:28:39 INFO client.RMProxy: Connecting to ResourceManager at /0.0.0.0:8032
18/07/04 13:28:41 INFO db.DBInputFormat: Using read commited transaction isolation
18/07/04 13:28:41 INFO db.DataDrivenDBInputFormat: BoundingValsQuery: SELECT MIN(`id`), MAX(`id`) FROM `students`
18/07/04 13:28:41 INFO db.IntegerSplitter: Split size: 0; Num splits: 4 from: 1001 to: 1003
18/07/04 13:28:41 INFO mapreduce.JobSubmitter: number of splits:3
18/07/04 13:28:42 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1530598609758_0015
18/07/04 13:28:42 INFO impl.YarnClientImpl: Submitted application application_1530598609758_0015
18/07/04 13:28:42 INFO mapreduce.Job: The url to track the job: http://oradb3:8088/proxy/application_1530598609758_0015/
18/07/04 13:28:42 INFO mapreduce.Job: Running job: job_1530598609758_0015
18/07/04 13:28:52 INFO mapreduce.Job: Job job_1530598609758_0015 running in uber mode : false
18/07/04 13:28:52 INFO mapreduce.Job: map 0% reduce 0%
18/07/04 13:28:58 INFO mapreduce.Job: map 33% reduce 0%
18/07/04 13:28:59 INFO mapreduce.Job: map 67% reduce 0%
18/07/04 13:29:00 INFO mapreduce.Job: map 100% reduce 0%
18/07/04 13:29:00 INFO mapreduce.Job: Job job_1530598609758_0015 completed successfully
18/07/04 13:29:00 INFO mapreduce.Job: Counters: 30
...
18/07/04 13:29:00 INFO mapreduce.ImportJobBase: Transferred 40 bytes in 21.3156 seconds (1.8766 bytes/sec)
18/07/04 13:29:00 INFO mapreduce.ImportJobBase: Retrieved 3 records.
# 生成的日志信息大家一定要好好理解
# 查看HDFS上的文件
[hadoop@hadoop000 ~]$ hadoop fs -ls /user/hadoop/students
Found 4 items
-rw-r--r-- 1 hadoop supergroup 0 2018-07-04 13:28 /user/hadoop/students/_SUCCESS
-rw-r--r-- 1 hadoop supergroup 13 2018-07-04 13:28 /user/hadoop/students/part-m-00000
-rw-r--r-- 1 hadoop supergroup 13 2018-07-04 13:28 /user/hadoop/students/part-m-00001
-rw-r--r-- 1 hadoop supergroup 14 2018-07-04 13:28 /user/hadoop/students/part-m-00002
[hadoop@hadoop000 ~]$ hadoop fs -cat /user/hadoop/students/"part*"
1001,lodd,23
1002,sdfs,21
1003,sdfsa,24
我們還可以加一些其他參數(shù) 使導(dǎo)入過程更加可控
-m指定啟動map進程個數(shù)，默認(rèn)是4個
--delete-target-dir刪除目標(biāo)目錄
--mapreduce-job-name指定mapreduce的job的名字
--target-dir導(dǎo)入到指定目錄
--fields-terminated-by指定字段之間的分隔符
--null-string含義是 string類型的字段，當(dāng)Value是NULL徽曲，替換成指定的字符
--null-non-string含義是非string類型的字段，當(dāng)Value是NULL，替換成指定字符
--columns導(dǎo)入表中的部分字段
--where按條件導(dǎo)入數(shù)據(jù)
--query按照sql語句進行導(dǎo)入 使用--query關(guān)鍵字，就不能使用--table和--columns
--options-file在文件中執(zhí)行
# 執(zhí)行導(dǎo)入
[hadoop@hadoop000 ~]$ sqoop import \
> --connect jdbc:mysql://localhost:3306/test \
> --username root --password 123456 \
> --mapreduce-job-name FromMySQL2HDFS \
> --delete-target-dir \
> --table students \
> -m 1
# HDFS中查看
[hadoop@hadoop000 ~]$ hadoop fs -ls /user/hadoop/students
Found 2 items
-rw-r--r-- 1 hadoop supergroup 0 2018-07-04 13:53 /user/hadoop/students/_SUCCESS
-rw-r--r-- 1 hadoop supergroup 40 2018-07-04 13:53 /user/hadoop/students/part-m-00000
[hadoop@oradb3 ~]$ hadoop fs -cat /user/hadoop/students/"part*"
1001,lodd,23
1002,sdfs,21
1003,sdfsa,24
# 使用where 參數(shù)
[hadoop@hadoop000 ~]$ sqoop import \
> --connect jdbc:mysql://localhost:3306/test \
> --username root --password 123456 \
> --table students \
> --mapreduce-job-name FromMySQL2HDFS2 \
> --delete-target-dir \
> --fields-terminated-by '\t' \
> -m 1 \
> --null-string 0 \
> --columns "name" \
> --target-dir STU_COLUMN_WHERE \
> --where 'id<1002'
# HDFS 結(jié)果
[hadoop@hadoop000 ~]$ hadoop fs -cat STU_COLUMN_WHERE/"part*"
lodd
# 使用query 參數(shù)
[hadoop@hadoop000 ~]$ sqoop import \
> --connect jdbc:mysql://localhost:3306/test \
> --username root --password 123456 \
> --mapreduce-job-name FromMySQL2HDFS3 \
> --delete-target-dir \
> --fields-terminated-by '\t' \
> -m 1 \
> --null-string 0 \
> --target-dir STU_COLUMN_QUERY \
> --query "select * from students where id>1001 and \$CONDITIONS"
# HDFS查看
[hadoop@hadoop000 ~]$ hadoop fs -cat STU_COLUMN_QUERY/"part*"
1002 sdfs 21
1003 sdfsa 24
# 使用options-file參數(shù)
[hadoop@hadoop000 ~]$ vi sqoop-import-hdfs.txt
import
--connect
jdbc:mysql://localhost:3306/test
--username
root
--password
123456
--table
students
--target-dir
STU_option_file
# 執(zhí)行導(dǎo)入
[hadoop@hadoop000 ~]$ sqoop --options-file /home/hadoop/sqoop-import-hdfs.txt
# HDFS查看
[hadoop@hadoop000 ~]$ hadoop fs -cat STU_option_file/"part*"
1001,lodd,23
1002,sdfs,21
1003,sdfsa,24
5.eval
查看幫助命令對與該命令的解釋為： Evaluate a SQL statement and display the results腿堤，也就是說執(zhí)行一個SQL語句并查詢出結(jié)果。
# 查看命令幫助
[hadoop@hadoop000 ~]$ sqoop help eval
usage: sqoop eval [GENERIC-ARGS] [TOOL-ARGS]
Common arguments:
--connect <jdbc-uri> Specify JDBC connect
string
--connection-manager <class-name> Specify connection manager
class name
--connection-param-file <properties-file> Specify connection
parameters file
--driver <class-name> Manually specify JDBC
driver class to use
--hadoop-home <hdir> Override
$HADOOP_MAPRED_HOME_ARG
--hadoop-mapred-home <dir> Override
$HADOOP_MAPRED_HOME_ARG
--help Print usage instructions
-P Read password from console
--password <password> Set authentication
password
--password-alias <password-alias> Credential provider
password alias
--password-file <password-file> Set authentication
password file path
--relaxed-isolation Use read-uncommitted
isolation for imports
--skip-dist-cache Skip copying jars to
distributed cache
--username <username> Set authentication
username
--verbose Print more information
while working
SQL evaluation arguments:
-e,--query <statement> Execute 'statement' in SQL and exit
# 執(zhí)行
[hadoop@hadoop000 ~]$ sqoop eval \
> --connect jdbc:mysql://localhost:3306/test \
> --username root --password 123456 \
> --query "select * from students"
18/07/04 14:28:44 INFO sqoop.Sqoop: Running Sqoop version: 1.4.6-cdh5.7.0
18/07/04 14:28:44 WARN tool.BaseSqoopTool: Setting your password on the command-line is insecure. Consider using -P instead.
18/07/04 14:28:44 INFO manager.MySQLManager: Preparing to use a MySQL streaming resultset.
----------------------------------------------------
| id | name | age |
----------------------------------------------------
| 1001 | lodd | 23 |
| 1002 | sdfs | 21 |
| 1003 | sdfsa | 24 |
----------------------------------------------------
6.export (HDFS數(shù)據(jù)導(dǎo)出到MySQL或Hive中的數(shù)據(jù)導(dǎo)入到MySQL)
常用參數(shù)：
--table指定導(dǎo)出表的名稱
--input-fields-terminated-by指定hdfs上文件的分隔符，默認(rèn)是逗號
--export-dir導(dǎo)出數(shù)據(jù)的目錄
--columns指定導(dǎo)出的字段
在執(zhí)行導(dǎo)出語句前mysql要先創(chuàng)建表（不創(chuàng)建表會報錯）：
# HDFS原文件
[hadoop@hadoop000 ~]$ hadoop fs -cat /user/hadoop/students/part-m-00000
1001,lodd,23
1002,sdfs,21
1003,sdfsa,24
# export導(dǎo)出到mysql
[hadoop@hadoop000 ~]$ sqoop export \
> --connect jdbc:mysql://localhost:3306/test \
> --username root \
> --password 123456 \
> --table students_demo \
> --export-dir /user/hadoop/students/
18/07/04 14:46:20 INFO sqoop.Sqoop: Running Sqoop version: 1.4.6-cdh5.7.0
18/07/04 14:46:20 WARN tool.BaseSqoopTool: Setting your password on the command-line is insecure. Consider using -P instead.
18/07/04 14:46:20 INFO manager.MySQLManager: Preparing to use a MySQL streaming resultset.
18/07/04 14:46:20 INFO tool.CodeGenTool: Beginning code generation
18/07/04 14:46:21 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM `students_demo` AS t LIMIT 1
18/07/04 14:46:21 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM `students_demo` AS t LIMIT 1
18/07/04 14:46:21 INFO orm.CompilationManager: HADOOP_MAPRED_HOME is /home/hadoop/app/hadoop-2.6.0-cdh5.7.0
18/07/04 14:46:24 INFO orm.CompilationManager: Writing jar file: /tmp/sqoop-hadoop/compile/fc7b53dd6eef701c0731c7a7c4a4b340/students_demo.jar
18/07/04 14:46:24 INFO mapreduce.ExportJobBase: Beginning export of students_demo
18/07/04 14:46:25 INFO Configuration.deprecation: mapred.jar is deprecated. Instead, use mapreduce.job.jar
18/07/04 14:46:25 INFO Configuration.deprecation: mapred.map.max.attempts is deprecated. Instead, use mapreduce.map.maxattempts
18/07/04 14:46:26 INFO Configuration.deprecation: mapred.reduce.tasks.speculative.execution is deprecated. Instead, use mapreduce.reduce.speculative
18/07/04 14:46:26 INFO Configuration.deprecation: mapred.map.tasks.speculative.execution is deprecated. Instead, use mapreduce.map.speculative
18/07/04 14:46:26 INFO Configuration.deprecation: mapred.map.tasks is deprecated. Instead, use mapreduce.job.maps
...
18/07/04 14:46:55 INFO mapreduce.ExportJobBase: Transferred 672 bytes in 29.3122 seconds (22.9256 bytes/sec)
18/07/04 14:46:55 INFO mapreduce.ExportJobBase: Exported 3 records.
# mysql中查看
mysql> select * from students_demo;
+------+-------+------+
| id | name | age |
+------+-------+------+
| 1001 | lodd | 23 |
| 1002 | sdfs | 21 |
| 1003 | sdfsa | 24 |
+------+-------+------+
3 rows in set (0.00 sec)
如果再導(dǎo)入一次會追加在表中
# 增加columns參數(shù)
[hadoop@hadoop000 ~]$ sqoop export \
> --connect jdbc:mysql://localhost:3306/test \
> --username root \
> --password 123456 \
> --table students_demo2 \
> --export-dir /user/hadoop/students/ \
> --columns id,name
# mysql結(jié)果
mysql> select * from students_demo2;
+------+-------+------+
| id | name | age |
+------+-------+------+
| 1001 | lodd | NULL |
| 1002 | sdfs | NULL |
| 1003 | sdfsa | NULL |
+------+-------+------+
3 rows in set (0.00 sec)
7.MySQL的中的數(shù)據(jù)導(dǎo)入到Hive中
常用參數(shù)：
--create-hive-table創(chuàng)建目標(biāo)表，如果有會報錯
--hive-database指定hive數(shù)據(jù)庫
--hive-import指定導(dǎo)入hive（沒有這個條件導(dǎo)入到hdfs中）
--hive-overwrite覆蓋
--hive-table指定hive中表的名字枣接，如果不指定使用導(dǎo)入的表的表名
--hive-partition-key指定Hive分區(qū)表字段
--hive-partition-value指定導(dǎo)入的分區(qū)值
首次導(dǎo)入可能會報錯如下：
18/07/04 15:06:26 ERROR hive.HiveConfig: Could not load org.apache.hadoop.hive.conf.HiveConf. Make sure HIVE_CONF_DIR is set correctly. 18/07/04 15:06:26 ERROR tool.ImportTool: Encountered IOException running import job: java.io.IOException: java.lang.ClassNotFoundException: org.apache.hadoop.hive.conf.HiveConf
解決方法：到hive目錄的lib下拷貝幾個jar包，問題就解決了
# 報錯解決方法
[hadoop@hadoop000 lib]$ pwd
/home/hadoop/app/hive-1.1.0-cdh5.7.0/lib
[hadoop@hadoop000 lib]$ cp hive-common-1.1.0-cdh5.7.0.jar /home/hadoop/app/sqoop-1.4.6-cdh5.7.0/lib/
[hadoop@hadoop000 lib]$ cp hive-shims* /home/hadoop/app/sqoop-1.4.6-cdh5.7.0/lib/
# 報錯解決后執(zhí)行導(dǎo)入
[hadoop@hadoop000 ~]$ sqoop import \
> --connect jdbc:mysql://localhost:3306/test \
> --username root --password 123456 \
> --table students \
> --create-hive-table \
> --hive-database hive \
> --hive-import \
> --hive-overwrite \
> --hive-table stu_import \
> --mapreduce-job-name FromMySQL2HIVE \
> --delete-target-dir \
> --fields-terminated-by '\t' \
> -m 1 \
> --null-non-string 0
# Hive中查看
hive> show tables;
OK
stu_import
Time taken: 0.051 seconds, Fetched: 1 row(s)
hive> select * from stu_import;
OK
1001 lodd 23
1002 sdfs 21
1003 sdfsa 24
Time taken: 0.969 seconds, Fetched: 3 row(s)
建議：導(dǎo)入Hive不建議大家使用–create-hive-table參數(shù),建議事先創(chuàng)建好hive表酗洒；因為自動創(chuàng)建的表字段類型可能并不是我們想要的。
# 增加partition參數(shù)
[hadoop@hadoop000 ~]$ sqoop import \
> --connect jdbc:mysql://localhost:3306/test \
> --username root --password 123456 \
> --table students \
> --create-hive-table \
> --hive-database hive \
> --hive-import \
> --hive-overwrite \
> --hive-table stu_import2 \
> --mapreduce-job-name FromMySQL2HIVE2 \
> --delete-target-dir \
> --fields-terminated-by '\t' \
> -m 1 \
> --null-non-string 0 \
> --hive-partition-key dt \
> --hive-partition-value "2018-08-08"
# Hive中查看
hive> select * from stu_import2;
OK
1001 lodd 23 2018-08-08
1002 sdfs 21 2018-08-08
1003 sdfsa 24 2018-08-08
Time taken: 0.192 seconds, Fetched: 3 row(s)
8.sqoop job的使用
sqoop job可以將執(zhí)行的語句變成一個job枷遂，并不是在創(chuàng)建語句的時候執(zhí)行樱衷，你可以查看該job，可以任何時候執(zhí)行該job酒唉，也可以刪除job矩桂，這樣就方便我們進行任務(wù)的調(diào)度。
--create<job-id> 創(chuàng)建一個新的job.
--delete<job-id> 刪除job
--exec<job-id> 執(zhí)行job
--show<job-id> 顯示job的參數(shù)
--list列出所有的job
# 創(chuàng)建job
[hadoop@hadoop000 ~]$ sqoop job --create person_job1 -- import --connect jdbc:mysql://localhost:3306/test \
> --username root \
> --password 123456 \
> --table students_demo \
> -m 1 \
> --delete-target-dir
# 查看job
[hadoop@hadoop000 ~]$ sqoop job --list
Available jobs:
person_job1
# 執(zhí)行job 會提示輸入mysql root用戶密碼
[hadoop@hadoop000 ~]$ sqoop job --exec person_job1
# HDFS查看
[hadoop@hadoop000 lib]$ hadoop fs -ls /user/hadoop/students_demo
Found 2 items
-rw-r--r-- 1 hadoop supergroup 0 2018-07-04 15:34 /user/hadoop/students_demo/_SUCCESS
-rw-r--r-- 1 hadoop supergroup 40 2018-07-04 15:34 /user/hadoop/students_demo/part-m-00000
我們發(fā)現(xiàn)執(zhí)行person_job的時候痪伦，需要輸入數(shù)據(jù)庫的密碼耍鬓，怎么樣能不輸入密碼呢
配置sqoop-site.xml即可解決
# 將sqoop.metastore.client.record.password參數(shù)的注釋去掉 或者再添加一下
[hadoop@hadoop000 conf]$ pwd
/home/hadoop/app/sqoop-1.4.6-cdh5.7.0/conf
[hadoop@hadoop000 conf]$ vi sqoop-site.xml
<property>
<name>sqoop.metastore.client.record.password</name>
<value>true</value>
<description>If true, allow saved passwords in the metastore.
</description>
</property>
參考文章：https://blog.csdn.net/yu0_zhang0/article/details/79069251
