If you’ve ever been involved in a VDI project before, you may still have painful memories of how challenging it used to be to come up with a right-sized brand new infrastructure across server, storage and network resources. One of the biggest enemies to fight against was a nasty weather phenomenon called “boot storm” that regularly rolled in early in the morning when most of your users arrived at their desks and all attempted to log in at the same time. The only way to overcome this problem and make users happy was to over-provide capacity. Unfortunately, this left you with unpleasant discussions with your IT management, asking why you spent so much additional budget on resources that were needed for just 15 minutes and that then sat almost idle for the rest of the day.
While these may have been good reasons to keep hands off VDI in the past, in a white paper commissioned by Fujitsu, analyst firm FreeForm Dynamics revealed 4 reasons why developments over the recent years have eliminated such issues almost entirely. Of course, systems still need to be designed, sized and built with the right resources configured in the right way to deal with your requirements, but several factors make this much more straightforward than it used to be:
- VDI software platforms have evolved and matured to allow the use of physical system resources, such as compute, memory, storage and networking, to be managed and optimized dynamically and automatically. This make them much more efficient and tolerant of demand spikes and fluctuations than their predecessors
- Physical infrastructure itself is now far less of a constraint. Modern networks tend to have more bandwidth, less latency, and better quality of service management. The same can be said of storage, where mainstream acceptance of flash technologies has translated into much higher performance at an affordable cost. ‘Bang-per-buck’ in relation to computing power has been trending similarly. The hardware you need to underpin your VDI environment is now considered ‘normal’ rather than exceptional.
- Experience and best practice has been harvested from many mainstream deployments of desktop virtualization projects, both large and small. It’s now rare to have to size, build and configure a solution from scratch. All serious vendors and integrators have blueprints or reference architectures that allow a specification to be generated based on a relatively simple set of parameters – number and type of desktops, user and application mix, and so on. Even if you do want an integrator to take care of the design and build, the number of consulting days required to come up with a fully working VDI environment is now much lower.
- Procurement and delivery options that further streamline implementation, operation, and ongoing management of growth and enhancement are now offered by some vendors. As an example, you can acquire VDI solutions as self-contained appliances, with the whole stack – hardware, software and management tools – pre-integrated and supported as a single solution. Modular, standards-based designs and ‘pay-as-you-grow’ commercial terms allow you to start small then scale quickly, smoothly and efficiently as you add more users and desktops to the system.
What once seemed impossible is now eminently doable
The performance and storage problems that plagued early implementations are now mostly a thing of the past, replaced by more capable hardware systems and software stacks that have been designed and pre-integrated to meet particular needs.
A decade or so ago, there were only one or two choices for desktop virtualization, whereas today there are many options across the board, ranging from foundational elements, such as optimized hardware and software stacks, right the way through to application and personal management. And the evolution is on-going, with fully managed, and even cloud-delivered desktop as a service now an option for those who want it, complete with the kind of service and support that few ever thought would be possible.