Earlier unit safety net action, shut down many servers, which many are running for more than 10 years old, the exercises after the end of shutdown, can't start up again, ok to turn it off before making the export data, hard to find a surviving machine for recovery, can also run with s are oracle8 Linux, but after the success of the recovery in is full of question marks, have a headache, and call for help,,,,
Now is the source side export database character set is chinese_China us7acii
Now import the client database is
American_America. Zhs16gbk

Using various methods such as hexadecimal three changes to become 03 54
There are direct command to change
1). Sys log in database: conn/as sysdba;
2). Close the database: shutdown immediate;
3). Open the database in the mount way: startup mount;
4) limit other users connect to the database using resources: the alter system enable restricted session;
5). Check the system the current process (maximum number of connections) : show the parameter the processes; Remember that the current value of job_queue_processes parameters, need to modify the back, behind
6). Kill CJQ0 and corresponding job process: the alter system set job_queue_processes=0;
7). Modify the queue the alter system monitoring process parameters set aq_tm_processes=0;
8). Changes to the database for the open mode: the alter database open;
9). Change the character set: the alter database character set us7ascii;

Later query sys account and restore the library of character set is shown in figure 2, to check the database character set under Linux is a figure, all in all imported or is full of question marks, deduced the source side have no way out, is only heavy, I'll find you can use super parent set can be solved, please help

CodePudding user response:

Not saved, between the two character sets is no need to reinstall, as long as use dbca to rebuild a us7ascii library is ok,

CodePudding user response:

reference 1st floor minsic78 response:

didn't saved, between the two character sets is no need to reinstall, as long as use dbca to rebuild a us7ascii library,

You can also use dbca give deleted the original waste repository

CodePudding user response:

Use super parent set to deal with

CodePudding user response:

Use super parent set to deal with

CodePudding user response:

Reconstruction of an instance of the same character set is better, with a superset of the character set import may be due to the field length is not enough and guide into the partial data