IBM DS8000 manual Subsystem-based data migration, Migration using backup and restore

Host

Host

Host

Backup System

Backup System

Intermediate device

Stop applications

Backup all data

Restore data

Restart

using new copy

Figure 16-5 Migration using backup and restore

The major disadvantage is again the disruptiveness. The applications that write to the data to be migrated must be stopped for the whole migration process. Backup and restore to and from tape usually takes longer than direct copy from disk to disk. The duration of the disruption can be reduced somewhat by using incremental backups.

16.2.2 Subsystem-based data migration

The DS8000 provides remote copy functionality, which also can be used to migrate data:

￿IBM TotalStorage Metro Mirror, formerly known as PPRC, for distances up to 300km

￿IBM TotalStorage Global Copy, formerly known as PPRC Extended Distance, for longer distances

￿A combination of Metro Mirror and Global Copy with an intermediate device in certain cases

Restriction: The combination of Metro Mirror and Global Copy is not available at GA.

These methods are host system agnostic and can therefore be used with only minimum system administrator attention. They also do not add any additional CPU load to the host systems, and they don't require the host system to be connected to both storage systems at the same time.

The necessary disruption is minimal. The initial copy is started during normal operation. Once it is complete, the target is kept up-to-date by only copying changes made to the source. At the cut over time, the applications are stopped and the mirror is allowed to reach synchronization. Then the target system is connected to the host instead of the source system and the applications can be restarted with the new copy.

Important: The source storage system must be removed from the host completely, not only physically, but also logically, including all configuration data.

However, the copy functions do not allow for the consolidation of smaller volumes into larger ones, since they are not aware of the structure of the data.

Chapter 16. Data migration in the open systems environment 339