Data protection is essential but expensive. The primary source of data protection expense is the required significant investment in the physical infrastructure. However, the cost of the physical infrastructure is much more than how much the organization spends on storage devices and network connections, it is also the time required to manage, upgrade and maintain that infrastructure.
The cost of data protection today is overwhelming most organizations but before the organization can make a change, it needs to understand the true costs of data protection, which then enables the organization to know where to look in order to reduce those costs.
The Data Problem
Driving the data protection problem is the data challenge. The sheer quantity of data and the rate that it continues to grow is creating significant challenges both in production and secondary data processes. The impact of production growth has an exponential impact on secondary (copy) data. Organizations are struggling with how to store all this data and where to store it. They need flexible solutions that enable them to reduce the physical footprint of copy data and the limitations on where to store it. In short, they should be able to store it anywhere at any time.
The Hard Cost of Data Protection Infrastructure
As stated earlier, the most obvious cost of data protection is the required physical cost of the infrastructure which includes the costs of upgrading or implementing a dedicated backup network to transport data. They also include the purchase of data protection storage for backup storage.
The problem with these costs is the equipment must be purchased in bulk. For example, backup storage has to be purchased in dozens of TBs at a time. A more frustrating situation is upgrades. Once that backup storage system has reached its capacity it needs to add another shelf of storage, representing dozens of TBs. If the system can’t support another shelf then the customer is forced to add another system or upgrade to a more powerful controller. Even more modern scale-out or converged data protection solutions still often provide way too much capacity per additional node.
With some exceptions, backup data grows on a gradual slope, not in spikes. The problem is that most backup storage systems are not sold incrementally. The customer has to buy TBs at a time and wait, potentially for years, until that capacity is used.
There is also the cost of on-premises compute to drive the data protection process. In the past, applications used to run on a single server but today because of the amount of data they are managing, applications increasingly need to scale out or at least need more processing power from the primary backup server. Again, scale-out or converged protection appliances attempt to solve this issue by giving the backup process access to more nodes with more processors per node. The problem is that those nodes need to be paid for, escalating the cost of the data protection infrastructure further and faster.
The Operational Cost of Data Protection Infrastructure
Another significant cost factor is the operational cost of managing the data protection infrastructure which gets increasingly more expensive (in terms of time) as the architecture scales. The first operational challenge is finding space in the data center. The biggest challenge facing most data centers is finding available floor space to implement new solutions.
The lack of space means that when backup infrastructure, either for additional storage or computing power, is upgraded, space needs to be found in the data center. Finding room often means moving systems around so that backup storage can all be located together or near the backup servers. It is a time-consuming process and is also fraught with potential errors when network connections are reestablished.
Another operational cost is the cost to re-organize backup jobs when new hardware is added. In many cases certain jobs need to be redirected toward the new device while other jobs remain with their current backup target. The balancing of backup loads is often a manual process that is more the result of educated guesses than actual science.
The biggest casualty of the operational cost of backup infrastructure is time; time that most IT professionals don’t have and time for which they can’t budget. The time required to upgrade, move equipment and redirect backup jobs is almost always an interruption to their daily workflow.
Converged data protection solutions claim to resolve many of the issues with the high cost of data protection infrastructure but in actuality they only change the perception of those costs, most of the costs are still there. Instead organizations need flexible solutions that it can purchase either as a service or as a complete solution depending on their use case. Some organizations only need to protect a single environment like VMware others need to protect an entire enterprise and mobilize data to the cloud.
In our next blog we’ll dive deep into the challenges with converged data protection. In the meantime, watch our on demand webinar “How to Create an Infrastructure-less Backup Strategy” to learn more about the challenges that data protection infrastructure creates and how to overcome those challenges.