As recently as a few years nearly every backup software product licensed its software based upon criteria such as the number of servers protected and what applications their backup agents needed to protect. But with the rise of virtual machines (VMs) and the complexity that approach to licensing created, most have now switched – or at least offer as an option – the ability to do either capacity-based or socket-based backup licensing. As licensing is sometimes the issue that determines which product gets selected to perform backup in your environment, it is important to understand the pros and cons of each.
Capacity-based licensing in particular has become more popular in the last couple of years. Rather than requiring users to first document their existing environment and then pick and choose which virtual server backup software features they need to license to protect their environment, all features that the backup software offers are included. In this way organizations can freely use whatever features the backup software includes to protect their environment.
The caveat to this capacity-based approach to software licensing is verifying how the vendor measures capacity. There are two ways that capacity may be measured. One way is to measure the total amount of data protected. In other words, it calculates the total amount of data to be protected before it is backed up.
The other option is to calculate the total amount of data in the backup data store after the data has been deduplicated. In this case, it again does not really care how many servers – physical or virtual – need to be protected or how much data is associated with each one. It only looks at the total amount of capacity after the data has been backed up, compressed and deduplicated.
Socket-based software licensing has emerged more recently and is gaining momentum in the virtualized environments. By tying the licensing to the number of sockets (processors) on the physical machine hosting the VMs, it affords organizations another option to realize significant savings in their backup software licensing costs. This approach encourages them to optimize the number of VMs on a physical machine and frees them to let their backup data stores grow unchecked.
As to which licensing approach is “best” for your organization, a few factors come into play in order to make the best decision.
If a product under consideration does capacity-based licensing, establish which method it uses to measure capacity. The data in VMs tend to deduplicate very well. Almost every user I talk to without exception is able to reduce their data stores by 80% and stories of 90% or more are not uncommon.
So if the product measures capacity by looking at the total amount of data protected versus the total amount of data backed up, that may be a cause for concern. I am aware of at least one company that had to bring in a second more economical backup software solution because the amount of data to protect in its environment exceeded its capacity-based licensing threshold even though it had plenty of room in its backup stores.
However the capacity-based licensing based on the total amount of data that is in the backup store is not without its caveats. Since its price per protected TB is likely higher than the other way that measures capacity, if the data does not deduplicate well for some reason (lots of images, very little duplicate data) this approach could actually be much more expensive.
Another possible concern is if you plan to retain your data for a long period of time and the data changes a great deal but does not necessarily increase in capacity. This could also lead to this alternative capacity-based model costing more over time.
If a product does socket-based licensing, this approach may be most desirable for those organizations that have the administrators and/or engineers on hand who are actively tuning their virtualized environment and have the time to optimize it.
The other scenario in which socket-based licensing may be appropriate is where organizations know they can host a lot of VMs on a single physical machine since the VMs consume very little processing power. However this again presumes some level of knowledge about the activity level on the individual VMs.
In short, organizations are probably better served from a cost and simplicity perspective by selecting backup software that offers capacity-based software licensing. The exceptions to this would be those organization that have the time to monitor the activity level on their VMs in their virtualized environment and can map that back to their real-world licensing costs.