Sqoop2的安装和使用
2017-04-10 本文已影响1150人
玄月府的小妖在debug
- 解压并安装
sudo tar xvfz sqoop-1.99.6-bin-hadoop200.tar.gz
sudo mv sqoop-1.99.6-bin-hadoop200 sqoop/
sudo chmod -R 775 sqoop/
sudo chown -R zhonglan sqoop/
- 修改环境变量
zhonglan@Master:~$ sudo vim /etc/profile
#sqoop
export SQOOP_HOME=$HOME/sqoop
export PATH=$SQOOP_HOME/bin:$PATH
export CATALINA_BASE=$SQOOP_HOME/server
export LOGDIR=$SQOOP_HOME/logs
hadoop@Master:~$ source /etc/profile
- 修改sqoop的环境变量
zhonglan@Master:/$ sudo vim /home/zhonglan/sqoop/server/conf/sqoop.properties
#修改指向我的hadoop安装目录
org.apache.sqoop.submission.engine.mapreduce.configuration.directory= /home/zhonglan/hadoop/etc/hadoop
#catalina.properties
zhonglan@Master:/$ sudo vim /home/zhonglan/sqoop/server/catalina.properties
common.loader=/home/zhonglan/hadoop/share/hadoop/common/*.jar,/home/zhonglan/hadoop/share/hadoop/common/lib/*.jar,/home/zhonglan/hadoop/share/hadoop/hdfs/*.jar,/home/zhonglan/hadoop/share/hadoop/hdfs/lib/*.jar,/home/zhonglan/hadoop/share/hadoop/mapreduce/*.jar,/home/zhonglan/hadoop/share/hadoop/mapreduce/lib/*.jar,/home/zhonglan/hadoop/share/hadoop/tools/*.jar,/home/zhonglan/hadoop/share/hadoop/tools/lib/*.jar,/home/zhonglan/hadoop/share/hadoop/yarn/*.jar,/home/zhonglan/hadoop/share/hadoop/yarn/lib/*.jar,/home/zhonglan/hadoop/share/hadoop/httpfs/tomcat/lib/*.jar,
下载mysql驱动包,mysql-connector-java-5.1.27.jar
把jar包丢到到$SQOOP_HOME/server/lib下面
sudo cp mysql-connector-java-5.1.27.jar $SQOOP_HOME/server/lib
有时,启动sqoop时可能会遇到找不到JAVA_HOME的情况,为了保险起见我们直接在配置文件中写入JAVA_HOME
在/home/zhonglan/sqoop/bin/sqoop.sh中,添加
export JAVA_HOME=/usr/lib/jvm/java-1.8.0-openjdk-1.8.0.111-1.b15.el7_2.x86_64
HADOOP_COMMON_HOME=/home/zhonglan/hadoop/share/hadoop/common
HADOOP_HDFS_HOME=/home/zhonglan/hadoop/share/hadoop/hdfs
HADOOP_MAPRED_HOME=/home/zhonglan/hadoop/share/hadoop/mapreduce
HADOOP_YARN_HOME=/home/zhonglan/hadoop/share/hadoop/yarn
- 启动sqoop
sqoop.sh server start - 验证启动成功
./sqoop.sh server start 启动
./sqoop.sh server stop 停止
./sqoop.sh client 进入客户端
set server --host hadoopMaster --port 12000 --webapp sqoop 设置服务器,注意hadoopMaster为hdfs主机名
show connector --all 查看连接类型
create link --cid 1 创建连接,cid为连接类型id
show link 查看连接
update link -l 1 修改id为1的连接
delete link -l 1 删除id为1的连接
create job -f 1 -t 2 创建从连接1到连接2的job
show job 查看job
update job -jid 1 修改job
delete job -jid 1 删除job
status job -jid 1 看看job状态
stop job -jid 1 停止job
- Sqoop2使用
sqoop.sh client
7.Sqoop2和Sqoop1区别
参考文档:
1.sqoo2安装
https://chu888chu888.gitbooks.io/hadoopstudy/content/Content/9/chapter0901.html
2.Sqoop1、Sqoop2功能差异及使用示例
http://udn.yyuap.com/doc/ae/921008.html#_Toc428540440