Copy Data is secondary copies of data that already exists on primary storage. These copies are used for recovery, reporting, analytics and testing. Most copy data management solutions are really just advanced snapshot management tools. They don’t provide a robust set of capabilities that enable the organization to use copy data locally, regionally and globally.
More Than a Golden Master
Most copy data management solutions create a single, golden master. From this primary copy it presents virtual images of data to other processes that need it; like backup, reporting and testing. The goal of the golden master approach is to reduce the storage requirements of secondary copies.
However, the problem with the golden master approach is that all the virtual images are fully dependent on the golden master. If something goes wrong and that copy is lost due to a system failure or disaster, all the dependent copies are invalid. There are times when the organization actually wants a full, stand-alone copy of data. In many cases there are multiple times where a separate, stand-alone copy of data is preferred.
Thinking Locally, Regionally and Globally
When copy data is applied locally the organization may want to have the copy data management solution create full clones of data that are not dependent on other copies. These full copies are useful for data protection and to position in other regional or global data centers so they are not constrained by network latency and reliability.
There are also times where a virtual image is suitable. Typically in a global organization, each location will want a full copy of the data which is then used to spawn off virtual images which are used for reporting, analytics, testing and application development. The virtual images are quickly created, easily destroyed and don’t consume significant disk capacity.
In many cases there is a need to create a high availability data copy for business applications which means the synchronous updating of volumes in two locations. Latency is a factor in synchronous replication so the secondary copy needs to be far enough away to provide some value in a disaster but close enough that performance isn’t impacted by network latency. The local to regional configuration is ideal here. The problem is that most copy data management solutions can’t perform synchronous data writes which means the organization needs to look outside of their copy data management system for a solution. Adding an additional solution adds costs and complexity.
Finally, for a wide-spread disaster, asynchronous replication is needed to create a copy of data in remote sites so that the organization can survive both a local and regional disaster. In reality, with the right copy data management solution, this is just another copy but it is one that is probably isolated so that it is always available for recovery operations.
Copy Data Management needs to provide organizations with local, regional and global capabilities. The process is more than just the reduction of extraneous copies of data; it is ensuring the right data is in the right location at the right time.
To learn more about copy data management check out Storage Switzerland’s recent webinar with Hitachi, “Evolving Data Protection from Backup to Copy Data Management.” In this webinar we discuss how data protection is no longer enough, what copy data management is and what key copy data management features organizations should look for.