Server virtualization appears to be gaining acceptance in even the most conservative organizations that deploy industry standard systems. It long has been a fixture in mainframe and midrange system based datacenters. Desktop virtualization, on the other hand, seems to be considered a technology looking for a problem to solve. Why is that? Depending upon which layer of the Kusnetzky Group model of virtualization software being considered there are quite a number of benefits that could quickly be realized.
A great article, do check it out. I don’t see desktop virtualization happening for technical reasons, they tend to be more operational, business orientated and risk. I have desktops that work, they’re a fixed cost (more or less handled by a service provider), I know it’s £800 per user, that if the pc breaks, an engineer will come and press the right combination of buttons, reboot or rebuild it. It works.
Switching to a virtual desktop world can be a freeing exercise if you think of what could be achieved. Consider being independent from your desktop, logging onto any pc and having the same user experience, the same icons, desktop and settings all working. A world where a new joiner might have a working ‘pc’ in minutes rather than days whilst a pc is obtained, the image deployed, packages configured to the user account and email settings applied. A world where the trader might be moved without the associated cost, the trader desk is simply pointed at their virtual machine, their virtual desktop and we don’t have to perform physical client moves. The challenges tend to revolve around investment and commitment in priorities. Is it a lower cost, more dynamic desktop solution or data center consolidation and server virtualization as well as the mindset issue. All those times I’ve spoken with colleagues who’ve either said “but say I got 50 users on one DL585 who wants to put users in the data center?” I haven’t got the space. Or the “but how would it work, do I do desktops then, I’m server, I don’t do desktops”.
If we abstract ourselves and think in terms of efficiency in power and of delivery we realize the potential. If we could virtualize the user pc experience, have 50 users on one rack server, have energy efficient thin clients, the amount of energy saved, of carbon produced could be substantial. As we have to declare the carbon footprint of our business, of the IT, this becomes an exciting way of providing the IT infrastructure. If we talk in terms of delivery, abstracting the user from the hardware, from the “my pc is slow”, “that’s Jill’s Pentium 733″, to one of functionality and commodity. You have a virtual instance allocated to you which you carry around with you, which follows you and your business. The client is the connector. To be able to provision, update/refresh clients on demand to improve reliability, security and user experience. Rebuilding a users communities’ desktop might not pose the challenge it would in the physical world. The hardware driver testing against the operating system, the user experience on that desktop configurations. You might only have one image, but 6 different pcs from the trader pc, the small form factor, those pcs’ that work and haven’t been upgraded.
It will be interesting to see where the virtual desktop goes and more importantly how IT comes to terms with itself. Who takes prority in performance in user experience and reliability? The development teams? The traders or IT? Do I want my support guys to have super fast virtual machines or enough to fulfil their role, with this in mind how do we sell the migration? How do we bill for the temporary machine? What is the virtual desktop cost and who owns it? Can I rebuild a virtual pc if it’s causing me problems or is that a change?
No related posts.
Related posts brought to you by Yet Another Related Posts Plugin.
Blog videos
dan@kusnetzky.net