This sub-section describes the procedure to build a replication of the operation system where GDS is used in a remote location and so on as a disaster recovery system.
The outline of building procedure is as follows:
Back up object configuration information of local and shared classes on the operation system. (sdxconfig Backup command)
Convert the configuration files backed up in step 1) for a disaster recovery system as below to create configuration files for a disaster recovery system. (sdxconfig Convert command)
Remove disks other than the copy destination of the storage copy process from each mirror group. (sdxconfig Convert -e remove command)
Change the physical disk names to physical disk names of a disaster recovery system. (sdxconfig Convert -e replace command)
Set the cluster system (CF, SF, automatic resource registration) on the disaster recovery system.
When using the copy function per slice of storage (such as REC of ETERNUS Disk storage system), use the configuration files created in step 2 for a disaster recovery system to restore object configuration information on a disaster recovery system. (sdxconfig Restore command)
After executing the sdxconfig Restore command, reboot the server where the command was executed.
Copy data of the local class and shared class to one case on the disaster recovery system from one case on the operation system by using the copy function of storage.
When using the copy function per LUN of storage (such as ACM CCM), use the configuration files created in step 2 for a disaster recovery system to restore object configuration information. (sdxconfig Restore command)
In the sdxconfig Restore command, -e chkps option is set.
After executing the sdxconfig Restore command, reboot the server where the command was executed.
Set the shared class (change type from local to shared) on the disaster recovery system. (sdxattr -C command)
Set the cluster application on the disaster recovery system.
