juanjo tugores - Fotolia
DaaS is a cost-effective way for businesses to rapidly scale their desktop estate when a demand spike occurs, such as a disaster, short-term project or product launch.
One of the defining characteristics of a cloud service is its elasticity; you can rapidly increase or decrease the amount of resources you consume. When you host desktops in the cloud, you use a service provider's infrastructure, which puts a lot of capacity at your disposal without the huge capital costs that come with virtual desktop infrastructure (VDI). Because of this, desktop as a service (DaaS) is an ideal model for delivering desktops to teams that form quickly or only work for a short period of time, such as teams responding to disasters, weather events or short-term projects.
With on-premises VDI, you have to buy the hardware to support peak capacity outright, regardless of whether or not you use it regularly. If your average capacity is much lower than the peak, the extra hardware is idle for much of the time, and delivers little value. But with DaaS, desktops run in a service provider's data center and you pay a subscription fee. You can rent only the amount and configuration of desktops you need at that moment. And you can increase or decrease that number of desktops easily and quickly because the provider owns and maintains the back-end infrastructure. This kind of cloud bursting helps you avoid buying more capacity if you only need it infrequently.
Use cases for crisis and short-term teams
In a disaster or crisis situation, giving responders access to a consistent IT environment is critical to solving problems and getting work done. For example, after an oil spill or a natural disaster, there is a surge of responders who may need access to desktops and applications. And following a storm, flood or other damaging event, insurance companies need to put a lot of assessors in the area quickly. Then, the assessors need to gather data and deliver it to the company's central claims system. In these situations, the IT shops at relief and insurance companies can ramp up the number of desktops they deploy to users, and then scale back again once the crisis is under control and the work is done.
But the surge in desktop needs doesn't have to come from an unplanned event. Businesses often form short-term teams for projects, events or product launches. If these teams grow large and then dissolve, then a desktop resourc where you only pay for what you use makes sense. One example would be a support team for a new product launch. The number of support requests will peak shortly after the launch, but then they will steadily drain away as customers learn and understand the new product better.
With an on-premises VDI deployment, shops often buy less hardware than they need for peak loads because it's too expensive to buy infrastructure you don't use all the time. But this compromises performance during the peak -- particularly if the load is bigger than average -- and that spike in demand often comes at critical times workers need good performance. If the peak is the result of a crisis, then responsiveness is even more critical.
DaaS is likely to make the most sense for workloads that have infrequent large peaks. A peak that occurs every week, or one with only a 30% load increase probably justifies owning the resources on-premises if you already do VDI. But if the peak is twice the usual load and it only occurs once a year, then DaaS may be the best choice.
Remember that using DaaS to respond to a surge in desktop use doesn't mean you have to use DaaS for the normal desktop workload, too. You could use physical PCs or on-premises VDI to handle the normal workload, and turn to DaaS only when the surge in demand occurs.
Of course, you may choose to use DaaS for all your users' desktops and simply scale up and down as demands come and go, which would require you to only support one desktop platform. But on-premises VDI and DaaS together may allow you to use the same desktop build for both environments, which eases support.
DaaS tool cost comparison
Five questions to ask potential DaaS providers
How DaaS is different from VDI