The next great virtualization frontier may well be the desktop, as more companies are finding that the idea of desktop virtualization infrastructure (VDI) can help them deliver a more reliable, secure and affordable computing experience for end users. At the same time, however, VDI puts a premium on ensuring the back room equipment stays up and running because without it, users are dead in the water.
In that respect, VDI represents a throwback to the old days of mainframes and dumb terminals. The terminal, by itself, was useless; without a connection to a functioning mainframe it couldn’t do a thing. PCs are obviously more intelligent, but with VDI the concept is much the same. All the really useful stuff – applications and data – is housed on centralized servers back in the data center. Should those servers go down, end users would be left without access to the tools the need to do their jobs.
To avoid such a scenario, naturally IT will take great care to ensure proper backup and disaster recovery mechanisms are in place for their servers. But they also need to consider the physical environment, including power and cooling, in which their servers and desktops live.
Let’s start at the desktop, which is the user’s terminal to the VDI world. It needs to be adequately protected with battery backup and surge protection – something along the lines of the Back-UPS line that APC has been selling since the 1980s. The idea is to provide power not only in the event out of an outage, but also to deal with unsafe voltage fluctuations and surges that can damage equipment – and leave users unable to work.
In the data center there are far more considerations. Before embarking on a VDI project, it’s a good idea to survey the data center to get an idea where you stand in terms of power and cooling capacity and simple things like the number of available outlets. Remember, you’re going to have more servers to deal with since they will essentially be replacing all those desktops.
Your new server infrastructure is likely to be highly dense, so you’ll need a cooling solution that can deal with all that heat. In-row cooling, for example, can concentrate cooling capacity where you need it most, instead of wasting capacity by cooling the entire data center. And its scalable – you just add capacity as your environment grows.
Scalable power protection is also a good idea because, again, as your VDI environment grows, you’re going to need more power protection. The APC by Schneider Electric Symmetra line, for example, is not only scalable but has integrated network management features. That means IT gets alerted if anything goes wrong and can manage the systems remotely – a big benefit.
In fact, most all APC by Schneider Electric products are manageable, which means they can feed data into a data center infrastructure management (DCIM) system to give IT a complete view of the entire data center. DCIM also allows IT to perform what-if analyses. Say they want to add a few more servers. DCIM will show where excess power and cooling capacity exists, helping them make an informed decision – and one that ensures the new servers function as they should. What’s more, DCIM systems can share data with management systems from other vendors, such as Cisco UCS Manager and VMware vSphere.
VDI technology represents a big change in the way IT architects its networks but it can only deliver on its promise if IT takes steps to ensure the infrastructure that makes it possible functions as it should, each and every day. Proper power and cooling are crucial to that effort.