Solutions for Oracle physical structure faults

Oracle physical structure failure refers to the failure of various databases caused by damage to each physical file in the database. These related faults are mainly caused by hardware faults, and may also be caused by human error. Therefore, we must first determine the cause of the problem. If it is a hardware fault, we must first solve the hardware problem. We can follow the instructions below for further handling without any hardware problems.

Control File Corruption:

The control file records important oracle configuration information, such as the database name, Character Set Name, data file, and log file location. If the control file is corrupted, the database is shut down abnormally. Once the control file is missing, the database cannot be started. This is a serious error.

You can query database log files to locate corrupted control files. The log file is located at $ ORACLE_BASE/admin/bdump/alert_ORCL.ora.

Damage a single control file:

1. Make sure that the database is closed. If the following command is not used to close the database:

 
 

  
  
svrmgrl>shutdown immediate; 
 
 

2. Check the initialization file $ ORACLE_BASE/admin/pfile/initORCL. ora to determine the path of all control files.

3. Use the operating system command to overwrite the wrong control file with other correct control files.

4. Run the following command to restart the database:

 
 

  
  
svrmgrl>startup; 
 
 

5. Use appropriate methods for full database backup.

Corrupt all control files:

1. Make sure that the database is closed. If the following command is not used to close the database:

 
 

  
  
svrmgrl>shutdown immediate; 
 
 

2. Restore the nearest control file from the backup result set. You can directly back up the nearest control file from tape to the corresponding directory for points without database backup; use the appropriate rman script to restore the nearest control file for the point with database backup.

3. Use the following command to create a script to generate a Database Control file:

 
 

  
  
svrmgrl>startup mount;  
  
  
svrmgrl>alter database backup controlfile to trace noresetlogs; 
 
 

4. Modify the trace file generated in step 3, copy a part of the statement about creating the control file, and make some modifications, so that it can reflect the latest physical structure fault of the database Oracle. Assume that the generated SQL file is named createcontrol. SQL.

Note:

The path of the Trace file can be determined by checking the $ ORACLE_BASE/admin/bdump/alert_ORCL.ora file after the step 3rd is executed.

5. Run the following command to recreate the control file:

 
 

  
  
svrmgrl>shutdown abort;  
  
  
svrmgrl>startup nomount;  
  
  
svrmgrl>@createcontrol.sql; 
 
 

6. Use appropriate methods for full database backup.

Damaged redo log files:

All Database additions, deletions, and changes are recorded in the redo log. If the active redo log file is damaged, the database is shut down abnormally. The inactive redo log will eventually change to the active redo log because of log switching. Therefore, the corrupted inactive redo log will eventually cause the database to terminate abnormally. In ipas/mSwitch, each redo log group has only one member. Therefore, in the following analysis, only the damage of the redo log group is considered, rather than the damage of a single redo log member.

This section describes how to handle Oracle physical structure faults. We hope this will help you in this regard.

Article by: http://www.programbbs.com/doc/3406.htm