Application owners primarily want control over their recoveries for one reason. They lack confidence in the ability of central backup administrators to successfully perform them. This lack of confidence explains why some backup solutions give users the flexibility within their tool to perform this task. However, as organizations increasingly adopt cloud infrastructures, expect to see recovery management become a task they can administer centrally.
Distributed Computing, Recovery Challenges
Distributed computing made many tasks easier when organizations first introduced into their core IT operations about 30 years ago. It suddenly gave nearly everyone instant access to personal productivity tools such databases, emails, spreadsheets, and word processors, among others.
Simultaneously, distributed computing made some core IT operational tasks such as recovery a more challenging and specialized function. While organizations can centrally perform backups reasonably well, performing recoveries in distributed environments still presents many challenges.
These challenges stem from the variables associated with each recovery. When a backup of an application or data occurs, the backup occurs in a relatively predictable environment. This contrasts with performing a recovery. This environment may differ significantly from the one in which the application or data was backed up. For example, a recovery may occur:
- On a different physical or virtual machine
- In a different location
- To a different operating system
- To a different version of the same operating system
- With access rights given to different applications or users
These variables and many others explain why application owners often want control over recoveries. They justifiably lack confidence that the individuals responsible for backup can understand or deal with the nuances of each recovery.
A Prerequisite for Successful Simplified Recovery Management
To move recoveries from the classification of “specialized” to “routine” will not happen by accident. A programmable, virtualized IT infrastructure serves as a prerequisite to achieving this objective. Thankfully, as organizations adopt cloud infrastructures, whether on-premises, off-premises, or both, they simultaneously lay the foundation for simplifying recovery management.
This IT environment equips a central administrator to obtain the characteristics of a system targeted to host a recovery. This individual can then determine, either manually or programmatically, if the recovery will succeed on that system. If determined it will succeed, the administrator can proceed with performing the recovery centrally. More importantly, the organization can ideally proceed without directly involving the application owner in the completion of this task.
The Rise of Recovery Management
As organizations continue to transition from distributed to cloud computing environments, they experience many benefits. They continue to receive the flexibility and productivity of distributed computing environments. They simultaneously obtain the availability, controls, and predictability of cloud computing environments.
Recovery management represents still another benefit they will experience as they continue to embrace cloud computing infrastructures. Cloud computing infrastructures create the virtualized, programmable IT infrastructures that organizations need to implement centralized recovery management.
Using the features of cloud computing infrastructures, organizations no longer need every application owner to become a recovery expert. Rather, organizations can leverage the programmable nature of cloud computing infrastructures and combine them with the emerging disciplines of artificial intelligence to create a better recovery experience.
Already we see offerings from Asigra, Druva, StorageCraft, Unitrends, and others available for small and midsize enterprises. For large enterprise environments, I have finally seen a viable offering from Datrium. Taken together, these offerings foretell what the future of recovery management will look like in the not-too-distant future for all size organizations.