Backup makes a copy of primary data on a secondary storage tier. It then tries to isolate that copy so that it can’t be changed either from external forces like ransomware or internal influences like users. For the most part though, once a backup is made its data sits idle until it is needed to recover a corrupted or deleted data set. While there are different forms most Copy Data Management solutions create a physical copy of data and then virtualize those copies to other processes in the data center. The copy data solution typically makes the first copy read-only, which isolates it. Only the virtual copies assigned to the other processes are read-write. Does copy data’s isolation qualify as a backup?
Making Copy Data Work for Backup
Copy data management solutions provide multiple point-in-time copies of data, and their read-only nature makes them relatively isolated from outside attack or insider deletion. The primary copy data problem as it pertains to replacing a backup is that many copy data solutions create their copy on the same storage system that stores the primary copy. If the storage system fails, then both the primary and the secondary copy are lost. There are several exceptions, however.
Many tier 1 storage systems have more than six nines of availability. Some even have a zero downtime guarantee. When vendors combine these extremely reliable systems with replication, so that data is replicated off-site, there is a plausible case that the organization is as protected as it would be with a “regular backup” solution. Many vendors even support replicating to a lower cost version of their storage arrays, reducing the typical premium cost of a replication strategy.
The reality is that most data protection professionals are uncomfortable recommending using copy data management as the sole form of data protection. Instead, a better practice is to use the copy data management solution to feed the backup solution a virtual copy of data like it would any other task. The backup application can then backup that virtual copy at its leisure. While the expense of a backup infrastructure continues, the use of copy data significantly reduces it. The reduction is the result of the copy data solution feeding the backup process a cold virtual copy of data, removing time pressures. More than likely the backup solution is also rarely used for recovery, if at all. It becomes the recovery of last resort.
There are several ways that vendors implement copy data management solutions. We discuss the pros and cons of each method and provide a live demo of a copy data management solution in our on demand webinar “Evolving Data Protection from Backup to Copy Data Management.”
Organizations often also use their backup protection solution as a means to retain data. While most experts do not recommend backup for data retention, it is undeniable that many organization do use it for the retention use case. In theory, copy data management can store thousands of virtual copies of information for years. Again, it is not a best practice to use either backup or copy data management for long-term data retention.
A data management strategy that leverages archiving to an object store is a much better data retention practice. Data management is the topic of another on demand webinar “Designing Storage Architectures for Data Privacy, Compliance and Governance.” In it, we discuss how to design a data management solution to create a long-term data retention strategy that meets the demands of GDPR and California’s Consumer Privacy Act. Interestingly some copy data management solutions can also feed the archiving process.