Scenario: Cloning to space efficient target devices to reduce storage requirements

An IT organization can greatly reduce its requirements for allocated storage on target volumes by using IBM® FlashCopy® SE to save clones, image copies, and backups to space efficient target devices.

In the business case Scenario: Cloning a DB2 subsystem from a specified point in time, a database administrator at a large financial services company resolves issues with migration testing by using DB2 Recovery Expert and DB2 Cloning Tool to clone and refresh multiple DB2 subsystems for stress testing. Now that the company is testing its applications under production conditions and finding defects earlier in the migration process, the IT department wants to leverage fast replication technology to save storage. A storage administrator is assigned to learn more about how space efficient storage can save the company money.

He learns that, in addition to virtually eliminating CPU and I/O utilization on the DB2 subsystems that they use for production, the IT department can also reduce the amount of physical storage that they require. For cloning operations that move data directly to tape, the IT department can leverage the space-efficient storage capabilities of IBM FlashCopy SE.

Full background copies do not require space efficiency because target volumes must equal or exceed the physical capacity of source volumes. But the IT department does not always require a full background copy. Sometimes they copy their data directly to tape, which means that they are running with a NOCOPY option in effect. The storage administrator learns that his organization can save physical storage in these situations by deploying space efficient volumes.

FlashCopy SE pools the storage requirements of multiple source volumes into a common repository that is thinly provisioned, which means that it is physically smaller than the full capacity of the source volumes. FlashCopy SE maps the data in the repository to virtual target volumes that are transparent to z/OS®.

The following figure illustrates how FlashCopy SE uses a thinly provisioned repository with smaller capacity than the source volumes to map your data to virtual target volumes:

Because the repository is an object within an extent pool, the storage does not initially occupy physical space. FlashCopy SE can allocate storage as it writes data to the repository. Because the writes occur against cache resources, and the de-staging process from cache to disk is asynchronous, this copy on write process has no effect on the performance of the production system.

The organization greatly reduces its storage requirements by implementing space efficient storage for target volumes.