I really liked this post somebody on Reddit made…
So I read many stories about the next big thing: virtualising desktops using VMware VDI View.
To me, it seems like a solution in search for a problem. This probably only shows my ignorance, but I do wonder which problem is actually solved with VDI?
Then I read about the huge investment in SAN equipment, to handle the tremendous IO load during peak hours and such. The server / storage infrastructure you have to build to support a decent size environment is quite amazing. And I’m not even thinking about the licensing costs (which I don’t know, but I guess it’s not cheap).
Do we solve the cost associated with the hardware, software and support of client computers? It seems to me that VDI might actually be more expensive and that the cost just move to the server/storage side.
But OK, cost is not everything. I can imagine that it may be more expensive overall, but that it’s easier to give users a higher quality computing experience. But I fail to see how.
I think the biggest problem of maintaining and supporting user computers is application provisioning and maintenance. Keeping the system up-to-date and not breaking it, with the least amount of effort.
What I did not do is to define a scenario as a base of reference, like what is the environment like, 80% desktops, 20% laptops, etc. But does that matter? Let’s assume you have 100% desktop, and don’t need the mobile frills. Remember the hype of cheap desktops (thin clients)? How did that pan out? And didn’t we have that like 10 years ago?
I do wonder. Why not just deploy desktops/clients, use a tool like app-v or even terminal-server-style(citrix) applications, to prevent software nesting itself like a malignant tumor in the OS, and give users a fast nice computing experience?
What am I missing here? What is the business case of VDI?
EDIT: These are some general answers that can be found in the comments:
- It depends entirely on your company’s situation, as always, it’s not a one solution that solves all problems.
- It can be a good solution when dealing with security requirements / compliancy
- Don’t expect to save cost, it will probably cost more than alternatives as of this moment.
- VDI may improve the user experience dramatically due to its flexibility and fast deployment.
- Client OS management (updates etc) is easier and better controlled.
- Can make it easier to setup a desktop DR solution
- BYOD environments like universities with a wide range of clients
- Less support calls related to client devices / similar to existing thin-client solutions
My personal opinion / gut feeling still is that if you properly implement desktop management with stuff like SCCM and app-v, the value of VDI is limited. But I don’t have any evidence to back that up, not even from actual experience. But – although clearly skeptical – I’m still very much interested in VDI, in the least to being better able to make a solid evaluation based on detailed knowledge instead of gut feelings.