Oracle数据迁移至MySQL

ORACLE DB: 11.2.0.3.0

MYSQL DB: 5.5.14

因项目需求,需要将ORACLE生产中数据迁移至MYSQL数据库中作为初始数据,方法有如下几种：

1、ORACLE OGG

2、通过手动编写select “insert into”脚本

3、工具,本次我就是使用了工具(sqluldr2),工具下载地址可以到www.anysql.net去下载

使用方法：
将sqluldr2.bin工具上传到oracle的bin目录下,
[root@db01 bin]# chown oracle.oinstall   sqluldr2.bin
 
[root@db01 bin]# chmod  775  sqluldr2.bin
 
[root@db01 bin]# su  -  oracle

[oracle@db01 ~]$ sqluldr2   help=yes
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
   read    = set DB_FILE_MULTIBLOCK_READ_COUNT at session level
   sort    = set SORT_AREA_SIZE at session level (UNIT:MB)
   hash    = set HASH_AREA_SIZE at session level (UNIT:MB)
   array   = array fetch size
   head    = print row header(Yes|No)
   batch   = save to new file for every rows batch (Yes/No)
   size    = maximum output file piece size (UNIB:MB)
   serial  = set _serial_direct_read to TRUE at session level
   trace   = set event 10046 to given level at session level
   table   = table name in the sqlldr control file
   control = sqlldr control file and path.
   mode    = sqlldr option, INSERT or APPEND or REPLACE or TRUNCATE
   buffer  = sqlldr READSIZE and BINDSIZE, default 16 (MB)
   long    = maximum long field size
   width   = customized max column width (w1:w2:...)
   quote   = optional quote string
   data    = disable real data unload (NO, OFF)
   alter   = alter session SQLs to be execute before unload
   safe    = use large buffer to avoid ORA-24345 error (Yes|No)
   crypt   = encrypted user information only (Yes|No)
   sedf/t  = enable character translation function
   null    = replace null with given value
   escape  = escape character for special characters
   escf/t  = escape from/to characters list
   format  = MYSQL: MySQL Insert SQLs, SQL: Insert SQLs.
   exec    = the command to execute the SQLs.
   prehead = column name prefix for head line.
   rowpre  = row prefix string for each line.
   rowsuf  = row sufix string for each line.
   colsep  = separator string between column name and value.
   presql  = SQL or scripts to be executed before data unload.
   postsql = SQL or scripts to be executed after data unload.
   lob     = extract lob values to single file (FILE).
   lobdir  = subdirectory count to store lob files .
   split   = table name for automatically parallelization.
   degree  = parallelize data copy degree (2-128).
   hint    = MySQL SQL hint for the Insert, for example IGNORE.
   unique  = Unique Column List for the MySQL target table.
   update  = Enable MySQL ON DUPLICATE SQL statement(YES/NO).
   crack   = The crack key for the content protection.
   uncrack = The uncrack key for the content protection.

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

每次参数所代表的具体含义和意义不描述,自己看

注意以下3点：
1、将ORACLE数据库中数据迁移到MYSQL,不要通过分隔符的方法,测试过分隔符,在导入到MYSQL过程中会有警告,一些数据被截断,避免大家走弯路

2、切记ORACLE生产环境的字符集是GBK,而MYSQL生产环境的字符集是UTF-8

3、将ORACLE数据全部导出成insert into 语句,这样在插入的过程避免出错,可以通过MYSQL还原命令直接还原,如出错进程会终止,source插入数据,中间出现警告,无法去验证

在测试ORACLE导出成insert into文本后,在还原到MYSQL过程中,测试N多次后,最终导出命令如下：
sqluldr2.bin   user=yoon/yoon@CS_177 query="select * from YOON.CESHI" charset=UTF8 text=MYSQLINS format=SQL  file=/u01/backup/oracle_mysql/CESHI_%b.txt  size=30000MB table=CESHI safe=YES

user : 用户名/密码@服务名

query：查询语句

charset：字符集UTF-8 , 将oracle中gbk字符集数据导出为文本格式后的字符集为UTF-8

text ：导出文件类型,必须为MYSQLINS, 成为的文件数据中自动生成插入mysql语句的``符号,例如：`张三`

format：格式为MYSQL

file：导出数据文件保存路径

%b：字母b必须为小写,可生成多个数据文本文件 ,例如：CESHI_1，CESHI_2 ......;  避免直接生成一个超级大的数据文件

size：必须以MB为单位,导出的每个文件的大小,30000MB,导出的每个数据文件30G

table：生成的数据文件中直接包含表名,否则会成生成insert into " ",没有表名

safe：YES   这个一定要加,在测试大量的表中,发现有的oracle表导出过程中会报ORA-24345错误,说是工具的BUG,加参数safe=YES即可

总结：
经过大量的测试后最终总结上面的导出命令语句,目前测试的表数据没有问题,至少对于我而言是这样.