Since VDI has become more popular, we've had to take a look at where we stand with Virtualization as a whole. Until Citrix bought XenSource it was pretty easy - we just waded in to the hardware virtualization pool from time to time when we'd talk about the big VDI picture. "So, that's VDI...oh, and you need VMware or Virtual Iron or XenSource for a backend. Ask Ron." We'd also bring in little tidbits about VMware Ace or Fusion and Parallels once in a while.
But now that Citrix has XenSource, the application delivery company has grown from a simple thin client computing company to a major player in several different industries. The technologies are converging as we speak, and that means that we're left to figure out where we stand and where we want to be with regards to hardware virtualization. We had to do the same gut-check when Citrix bought WANscaler and Net6--did we want to cover hardware? Networking hardware at that?
Admittedly, it's been pretty fun as a blogger with his eyes on the industry to sit back and watch and Citrix and VMware duke it out. Most recently, VMware turned on its FUD machine and sent an email to its customers that appeared to be quite defensive. This was right after Citrix Synergy and the announcement of general availability for XenDesktop. In return, Roger Klorese, a Senior Director on the Product Marketing team for XenServer, lobbed up a blog post debunking much of what VMware had to say.
The question I have, and the one we keep asking ourselves as new developments arise is: Does any of this really matter long term?
It's been my contention for some time now that hardware virtualization software packages will eventually become obsolete because the virtualization components will be built into the hardware. Intel and AMD are reportedly hard at work figuring out how to incorporate more virtualization capabilities directly into the processors and chipsets. Once that plays out and we can get a box from Dell or HP or wherever that has hardware virtualization built right in, there won't be any more ESX vs. XenServer vs. HyperV vs. Virtual Iron arguments. Hell, there might not even be a Vmware, Citrix, or Virtual Iron as we know them today.
Instead, once everyone puts their gloves down, there will be one thing that sets the players apart - the management software. You can see the industry shaping up for a competition right now. Microsoft and Citrix have joined forces--even going so far as to support each other's virtual machines--in an effort to compete with VMware. Microsoft has purchased Kidaro for packaging of VM’s similar to VMware’s ACE, and Calista to extend the capabilities of the RDP protocol. VMware, classicly a virtualization-only company, is now purchasing companies like Thinstall for application virtualization and Propero for a VDI connection broker (although they completely rewrote that package).
Other companies have already taken a hypervisor-agnostic approach to their VDI solutions, the most visible (besides Citrix, of course) being Quest Software's Provision Networks Division. There are others, too, such as Ericom PowerTerm, ClearCube Sentral, and Chip PC Xcalibur Global. Some of these companies offer more features than others, but the market is definitely ripening.
When virtualization on the lowest level happens, Citrix and Microsoft can continue their symbiotic relationship and take on VMware in a battle along with Provision and all the other guys for the complete package: end-to-end virtualization and virtualization management, which includes everything that we try to draw lines between today--applications, operating systems, hardware, networking, user environments, and storage. (Wait, storage? What? Raise your hand if your parent company is EMC. Ahh, but that's an article for another day.) (While I'm speaking inside parenthesis, imagine if Citrix or Microsoft were to purchase a company like DataCore for virtualized storage management capabilities.)
So where does that leave us? We don't have any sort of threshold or metrics on what company or product we cover, but the products have to be in some way appealing to us. They have to make sense, have a niche, fix a problem, and/or be original (or at least expand in a forward moving direction on existing technologies). With regards to hardware virtualization, it looks to me like the arena is going to get smaller sometime in the future. After that, we'll be left with pretty much exactly what we talk about right now.