一、 Java运行环境配置
二、 Kettle启动
a) ![](https://wkrtcs.bdimg.com/rtcs/image?w=186.00&md5sum=73dbba2a8ab90b12c61dde5a58c31c46&sign=11badfcd5a&rtcs_flag=2&rtcs_ver=3.1&l=webapp&bucketNum=39&ipr={"t":"img","r":"r_4","w":"186.00","h":"49.00","dataType":"png","c":"word/media/image1.png","imgOriW":186,"imgOriH":49})
三、 ![](https://wkrtcs.bdimg.com/rtcs/image?w=428.19&md5sum=73dbba2a8ab90b12c61dde5a58c31c46&sign=11badfcd5a&rtcs_flag=2&rtcs_ver=3.1&l=webapp&bucketNum=39&ipr={"t":"img","r":"r_4","w":"428.19","h":"397.61","dataType":"png","c":"word/media/image2.png","imgOriW":700,"imgOriH":650})
Connection Name:自定义连接名
Host Name:服务器连接IP地址
Database Name:数据库名
Port Number:端口号（oracle 1521）
User Name:用户名
Password:密码
注意：Kettle 的目录data-integration\libext\JDBC下必须有oracle的JDBC驱动程序ojdbc5.jar/ojdbc6.jar，不然oracle JDBC连接不上。以及Kettle 使用的数据库名“SID_NAME” 和 PL/SQL 使用的数据库名 “SERVICE_NAME”要保持一致，包括大小写。
四、 新建transformation
五、 输入
a) ![](https://wkrtcs.bdimg.com/rtcs/image?w=137.55&md5sum=73dbba2a8ab90b12c61dde5a58c31c46&sign=11badfcd5a&rtcs_flag=2&rtcs_ver=3.1&l=webapp&bucketNum=39&ipr={"t":"img","r":"r_4","w":"137.55","h":"110.79","dataType":"png","c":"word/media/image3.png","imgOriW":257,"imgOriH":207})
六、 表输入
a) ![](https://wkrtcs.bdimg.com/rtcs/image?w=137.00&md5sum=73dbba2a8ab90b12c61dde5a58c31c46&sign=11badfcd5a&rtcs_flag=2&rtcs_ver=3.1&l=webapp&bucketNum=39&ipr={"t":"img","r":"r_4","w":"137.00","h":"25.00","dataType":"png","c":"word/media/image4.png","imgOriW":137,"imgOriH":25})
![](https://wkrtcs.bdimg.com/rtcs/image?w=258.63&md5sum=73dbba2a8ab90b12c61dde5a58c31c46&sign=11badfcd5a&rtcs_flag=2&rtcs_ver=3.1&l=webapp&bucketNum=39&ipr={"t":"img","r":"r_4","w":"258.63","h":"164.13","dataType":"png","c":"word/media/image5.png","imgOriW":583,"imgOriH":370})
数据库连接：选择数据库
获取SQL查询语句：选择数据库中不同模式（用户）的表
SQL：写出你的sql查询语句，从数据库中抽出原始数据
七、 表输出
![](https://wkrtcs.bdimg.com/rtcs/image?w=128.00&md5sum=73dbba2a8ab90b12c61dde5a58c31c46&sign=11badfcd5a&rtcs_flag=2&rtcs_ver=3.1&l=webapp&bucketNum=39&ipr={"t":"img","r":"r_4","w":"128.00","h":"31.00","dataType":"png","c":"word/media/image6.png","imgOriW":128,"imgOriH":31})
![](https://wkrtcs.bdimg.com/rtcs/image?w=294.45&md5sum=73dbba2a8ab90b12c61dde5a58c31c46&sign=11badfcd5a&rtcs_flag=2&rtcs_ver=3.1&l=webapp&bucketNum=39&ipr={"t":"img","r":"r_4","w":"294.45","h":"214.99","dataType":"png","c":"word/media/image7.png","imgOriW":504,"imgOriH":368})
数据库连接：数据输出到的数据库
目标模式：用户名
目标表：可以选择已有表或者新命名表，不必考虑表存不存在
提交记录数量：
裁剪表：每次执行转换过程都会清空该表
Specify database fields:
Database fields:需要选择里面的Enter field mapping
![](https://wkrtcs.bdimg.com/rtcs/image?w=275.10&md5sum=73dbba2a8ab90b12c61dde5a58c31c46&sign=11badfcd5a&rtcs_flag=2&rtcs_ver=3.1&l=webapp&bucketNum=39&ipr={"t":"img","r":"r_4","w":"275.10","h":"222.25","dataType":"png","c":"word/media/image8.png","imgOriW":614,"imgOriH":496})
源字段：传过来的数据字段名
目标字段：需要保存数据的表里面的字段名
映射：需要源字段和目标字段相互对应才能输出数据
SQL:目标表不存在或者表结构和要保存的数据结构不同，这里自动执行创建和修改
本文来源：https://www.2haoxitong.net/k/doc/329a4da66bec0975f565e209.html
文档为doc格式