This section describes the procedure for handling errors on the secondary volume.

Procedure

Note: The numbers to the left of the process boxes correspond to the item numbers in the following explanation:

1. Stops the application server and interrupts online transactions.
2. Forcibly terminate HiRDB.
3. Change the value of the pd_rise_disaster_mode operand to alone.
4. Executing the pairsplit command (with the -S option specified) to split all paired logical volume groups.
5. Restart HiRDB. After the restart, do not update the database being used by online transactions.
6. Terminate HiRDB normally.
7. Contact the person responsible for maintaining the Hitachi disk array subsystems, determine the cause of the error on the secondary volume, and take the appropriate corrective action.
8. From the backup and unload log files, recover all RDAREAs that were created in the paired logical volume group in which the error occurred. For details about how to recover a database, see the HiRDB Version 9 System Operation Guide. Also, create new system files for all of the system files that had been created in the paired logical volume group in which the error occurred.
9. Execute the pdcopy command to make a backup of each system.
   For details about how to make backups, see the HiRDB Version 9 System Operation Guide.
   
10. Copy the backup files made in step 9 to the remote site.
11. Execute the paicreate command and generate a paired logical volume group.
12. Execute the pairevtwait command (with the -s pair option specified) on the paired logical volume group specified in the paircreate command executed in step 11, and wait until a termination code of 0 is returned.
13. Change the value of the pd_rise_disaster_mode operand to normal.
14. Start HiRDB.
15. Collect a synchronization point dump by executing the pdlogsync -d sys -w command and wait until validation is completed. For HiRDB/Parallel Server, execute the pdlogsync -d sys -w command on all servers.
16. Execute the pdrels -o command to release and open all RDAREAs created in the paired logical volume group in which the error occurred. Also, execute the pdlogopen and pdstsopen commands to open all system files created in the paired logical volume group in which the error occurred.