Opinion VMware is making a play to wrest data centre control away from every other vendor and become the single door through which users and admin staff see data centre server, networking, and storage resources.
A data centre administrator will increasingly monitor, manage, and diagnose data centre resource problems through VMware APIs, using manufacturer-supplied plug-ins to link the VCenter server to supplier's boxes downstream in the stack.
Look at the raft of storage suppliers providing plug-ins at VMworld so that their arrays can be managed through VMware. Pretty soon it will make internal development resource sense for them to deploy their management software as a VMware virtual machine (VM) and simultaneous plug-in.
No storage vendor can withstand VMware. None appear to want to. If they gotta sell product into VMware shops that works with VMware then they will. They are already.
The storage suppliers are much farther along this road than the networking suppliers. Imagine what could happen with storage. All storage provisioning, monitoring, reporting, and management could be done through VMware. A VMware shop could, when looking to buy storage, say that it has to be managed through VMware and that it's simply not interested in any supplier-specific features outside of VMware.
Such supplier-specific features - read added value - means lock-in and extra expense and complexity. It will be simpler to go for the standard and that standard could be VMware-driven. All storage vendors could find themselves competing to deliver the best storage commodity that runs with VMware. Networking product vendors could look at this prospect and blanch.
Their boxes don't generally have X86-based controllers, which many storage suppliers use.
Networking vendors keen to avoid VMware-driven commoditisation should avoid using X86 controllers like the plague. Any X86 software can be likely turned into a VM and run in an ESX server. VMware's effect can be likened to taking server and network and storage box value-add and rendering it less and less relevant because everything is seen through a VMware admin's view and accessed through VMware.
VMware is becoming the new Microsoft.
Microsoft provided a good basic and cheap operating system for IBM PCs and as they became a standard, so too did Microsoft. All other PC operating systems fell twitching and screaming by the wayside. It's not a perfect comparison by any means, but why would any data centre operator, having adopted VMware, want to adopt anything else?
VMware: The Great Leveller
VMware is the great leveller. Want to run Windows, Linux, Unix, whatever on the same X86 hardware? No problem. VMware can do that. Why on earth should VMware adopters want to run Hyper-V or XEN or Red Hat's hyper-visor? I could see them doing it to keep VMware honest and its prices down and for specific functionality that VMware doesn't provide, but it probably will in the future. Not for any other reason though.
The more popular, the more ubiquitous VMware becomes, the less relevant will be Microsoft, Red Hat, Novell (SuSE), and XEN. The hypervisor is the data centre high ground and VMware has got more of it than any other supplier. There is a kind of gravitational effect at work here with VMware's popularity and widespread use attracting more suppliers to work with VMware, which attracts more customers to adopt it, which drives VMware to add more features, which....and so the VMware virtuous circle spins around and around.
Paul Maritz knows data centre operators are sick and tired of IT complexity and dealing with multiple suppliers and environments. His company is offering a fix for this. VMware just makes all that mess go away, into an abstracted virtual wonderland, a quarantine for most of the ills an IT director wants treatment for.
We're watching a VMware galaxy form in the data centre space and all the individual product stars are falling into place spinning around, held in VMware's thrall just like stars in a galaxy, imprisoned in their place with no way to break free.
The client:server era gave way to a multi-tier architecture with server sprawl. Unix standardisation failed, and Linux is no more than a worthy competitor to Microsoft in the data centre.
Microsoft software like Exchange and SQL and SharePoint is good but there is server cycle-sucking Windows between it and the hardware. There is the prospect - there surely must be the prospect - of apps being produced which request their previously Windows-delivered resources direct from ESX. Every step in an app's resource consumption stack needs physical host server cycles. Why not minimise the number of steps and have apps run more and more in VMs that have a thin or almost non-existent O/S layer between the app and the VM?
There are exceptions to this coming VMware hegemony. VMware is an X86 phenomenon and has no ability to play on SPARC, POWER, mainframe, and other architectures. IBM mainframe users will be able to sail on, rising effortlessly above the VMware tide, and watching as VMware uses software to bind X86 servers, multiple networking gear and storage into a simulacram of a mainframe. Separate boxes seductively glued together with VMware's software.
VMware is going to become, if this view is right, more important than EMC itself. Storage, after all, is just a place to store data. The important place in the data centre IT hierarchy is where decisions are made, and that is in the servers. When all servers are X86 servers then the supplier of the main server control software, the hypervisor, calls the shots.
The chief executive of that hypervisor company can also expect to call the shots. Maritz is surely the coming guy and EMC could easily become VEMC. Who would have thought that selling add-on external storage could be the start of a path leading to data centre domination?
That's what could be in store and every other supplier with pretensions to be a data centre kingpin better get a strategy for dealing with VMware, because this little puppy could grow to become the biggest data centre canine there has ever been. ®