配置 oracle数据导出工具sqluldr2

/opt/oracle/db01/app/oracle/product/11.2.0/bin/sqluldr2 XXemo/RHBX0X1\[email protected]/orcl  QUERY="select * from SJ_VOD_2022_06 t"  field="|" record=0x0d0x0a batch=yes rows=200000  file=./test.log head=yes charset=utf8

分别对应32为，64位的win和Linux平台

安装步骤

1.需要安装oracle_client

2.复制sqluldr2_linux32_10204.bin（64位系统用sqluldr2_linux64_10204.bin）到$ORACLE_HOME的bin目录，重命名为sqluldr2.bin

3.修改 $HOME目录下.bash_profile 增加如下环境变量

export LD_LIBRARY_PATH=$ORACLE_HOME/bin:$ORACLE_HOME/lib:/lib:/usr/lib  
4.验证
-bash-3.2$ sqluldr2.bin

SQL*UnLoader: Fast Oracle Text Unloader (GZIP, Parallel), Release 4.0.1
(@) Copyright Lou Fangxin (AnySQL.net) 2004 - 2010, all rights reserved.

License: Free for non-commercial useage, else 100 USD per server.

Usage: SQLULDR2 keyword=value [,keyword=value,...]

Valid Keywords:
   user    = username/password@tnsname
   sql     = SQL file name
   query   = select statement
   field   = separator string between fields
   record  = separator string between records
   rows    = print progress for every given rows (default, 1000000) 
   file    = output file name(default: uldrdata.txt)
   log     = log file name, prefix with + to append mode
   fast    = auto tuning the session level parameters(YES)
   text    = output type (MYSQL, CSV, MYSQLINS, ORACLEINS, FORM, SEARCH).
   charset = character set name of the target database.
   ncharset= national character set name of the target database.
   parfile = read command option from parameter file 

  for field and record, you can use '0x' to specify hex character code,
  \r=0x0d \n=0x0a |=0x7c ,=0x2c, \t=0x09, :=0x3a, #=0x23, "=0x22 '=0x27 

LINUX下用法：

get_tables.sql为预定义导出sql

-bash-3.2$ sqluldr2.bin USER=userid/keyword@db_name sql=get_tables.sql head=no FILE=/oracle/oradata/backup/stock.csv
           0 rows exported at 2014-05-06 11:49:25, size 0 MB.
     1000000 rows exported at 2014-05-06 11:49:28, size 24 MB.
     2000000 rows exported at 2014-05-06 11:49:32, size 48 MB.
     3000000 rows exported at 2014-05-06 11:49:35, size 76 MB.
     4000000 rows exported at 2014-05-06 11:49:38, size 100 MB.
     5000000 rows exported at 2014-05-06 11:49:41, size 124 MB.
     6000000 rows exported at 2014-05-06 11:49:44, size 152 MB.
     7000000 rows exported at 2014-05-06 11:49:47, size 176 MB.
     8000000 rows exported at 2014-05-06 11:49:50, size 204 MB.
     9000000 rows exported at 2014-05-06 11:49:53, size 228 MB.
    10000000 rows exported at 2014-05-06 11:49:56, size 252 MB.
    11000000 rows exported at 2014-05-06 11:49:59, size 280 MB.
    12000000 rows exported at 2014-05-06 11:50:02, size 304 MB.
    13000000 rows exported at 2014-05-06 11:50:05, size 328 MB.
    14000000 rows exported at 2014-05-06 11:50:08, size 356 MB.
    15000000 rows exported at 2014-05-06 11:50:11, size 380 MB.
    16000000 rows exported at 2014-05-06 11:50:14, size 408 MB.
    17000000 rows exported at 2014-05-06 11:50:17, size 432 MB.
    18000000 rows exported at 2014-05-06 11:50:20, size 456 MB.
    19000000 rows exported at 2014-05-06 11:50:23, size 484 MB.
    20000000 rows exported at 2014-05-06 11:50:27, size 508 MB.
    21000000 rows exported at 2014-05-06 11:50:30, size 531 MB.
    21877116 rows exported at 2014-05-06 11:50:33, size 558 MB.
         output file /oracle/oradata/backup/f_stock.csv closed at 21877116 rows, size 558 MB.
-bash-3.2$ 

下为WINDOWS平台使用方法：

代码例子1：

sqluldr2.exe USER=用户/密码@tnsQUERY="select   /*+ parallel(8)  */  *from cs_XXX dt,cfg_XXX devdim105 wheredt.starttime>=to_date('2012-06-27 00:00:00','yyyy-mm-dd hh24:mi:ss') and dt.starttime<=to_date('2012-06-2700:01:59','yyyy-mm-dd hh24:mi:ss') AND dt.msc= devdim105.mapvalue(+) ANDdevdim105.deviceid=15 " head=yes FILE=F:\cs_XXX_test.csv

代码例子2(这种方式用于导出的查询sql很长，那么就把sql写在123.sql文件里 )：

sqluldr2.exe USER=用户/密码@tns  sql=123.sql head=yes FILE=F:\cs_XXX_test.csv

3.在cmd里，直接把步骤2的代码例子1 ，贴进去执行。

注：如果执行报报错，就有可能是环境变量path的问题，还有就是plsql需要的oci.dll文件等多个dll文件，请放在sqluldr2的目录下。

4.查看结果F:\cs_XXX_test.csv

5.打开csv里面的内容，就可以用sqlldr进行 入库。

字符集的控制：

  默认情况下, Oracle客户端程序的字符集都由环境变量(NLS_LANG)决定, 很多时侯会因为忽略了环境变量的检查, 尤其是在Shell或Perl脚本中调用时, 导致导出的数据变成了乱码. 在OCI 8的接口中, 可以使用OCIEnvNlsCreate函数在程序中指定客户端字符集, 从而跳过环境变量的影响, 于是就在SQLULDR2工具中新增了两个选项, 用于指定导出时了字符集.

    新的SQLULDR2工具命令行帮助如下所示:

SQL*UnLoader: Fast Oracle Text Unloader (GZIP, Parallel), Release 4.0.1
(@) Copyright Lou Fangxin (AnySQL.net) 2004 - 2010, all rights reserved.

Usage: SQLULDR2 keyword=value [,keyword=value,...]

Valid Keywords:
  user    = username/password@tnsname
  sql    = SQL file name
  query  = select statement
  field  = separator string between fields
  record  = separator string between records
  rows    = print progress for every given rows (default, 1000000)
  file    = output file name(default: uldrdata.txt)
  log    = log file name, prefix with + to append mode
  fast    = auto tuning the session level parameters(YES)
  text    = output type (MYSQL, CSV, MYSQLINS, ORACLEINS, FORM, SEARCH).
  charset = character set name of the target database.
  ncharset= national character set name of the target database.
  parfile = read command option from parameter file

  for field and record, you can use '0x' to specify hex character code,
  \r=0x0d \n=0x0a |=0x7c ,=0x2c, \t=0x09, :=0x3a, #=0x23, "=0x22 '=0x27

    在一个GBK的数据库中存入"中国"两个字, 然后以GBK和UTF8两种不同的字符集进行导出, 在命令行指定字符集设置.

C:\>sqluldr2 scott/tiger query=test charset=ZHS16GBK
      0 rows exported at 2010-06-24 09:12:27, size 0 MB.
      1 rows exported at 2010-06-24 09:12:27, size 0 MB.
        output file uldrdata.1.txt closed at 1 rows, size 0 MB.

C:\>dir uldrdata.1.txt
2010-06-24  09:12                5 uldrdata.1.txt

C:\>sqluldr2 scott/tiger query=test  charset=UTF8
      0 rows exported at 2010-06-24 09:12:42, size 0 MB.
      1 rows exported at 2010-06-24 09:12:42, size 0 MB.
        output file uldrdata.1.txt closed at 1 rows, size 0 MB.

C:\>dir uldrdata.1.txt
2010-06-24  09:12                7 uldrdata.1.txt

    用GBK导出时, 文件大小为5个字节, 两个汉字(一个汉字占两个字节)再加一个换行符. 而用UTF8导出时, 文件大小为7, 两个汉字(一个汉字占三个字节)加上一个换行符. 如果用SQLULDR2为特定的数据库写脚本程序时, 在命令行指定字符集, 将使脚本更加强壮.