Stack wars: The cloud OEMpires strike back

Increased consolidated converged infrastructure and cloud offerings from OEMs that once extolled a best-of-breed approach cause problems for users trying to pick the right stack.

A few years ago, TechTarget asked me and fellow vExpert David Davis to judge the latest hardware and software innovations on the Solutions Exchange floor at VMworld. In the end, we picked the Cisco Nexus 1000V that introduced a new way to manage the virtual switch stack inside VMware vSphere. At the time, we both sniggered at some OEMs marketing their new hardware as “virtualisation blades.” We thought, "What the hell is that? A server is a server." Tacking on the word virtualisation didn’t pull the wool over our eyes.

With today’s OEMs quickly realigning themselves and creating “cloud platforms,” a cynic would say nothing much has changed. But in this case, he would be wrong.  

Over the past two years, we've seen increasing product consolidations whereby vendors line up customers to consume their hardware in a one-stop-shop manner. The early days of this showed a data centre being configured within the form factor of a shipping container. More recently, Dell launched its vStart converged stack that ships pre-cabled and pre-racked servers, network and storage all in one bundle.

We all know Dell is aiming its “virtual infrastructure solution” squarely at the likes of HP BladeSystem Matrix and VCE Vblock. It’s ironic that after years of slagging the elephant in the room -- IBM -- the OEMs that once extolled the virtues of a best-of-breed approach are now offering monolithic stacks. There’s no other way to describe them. Whilst there is some flexibility  (such as NetApp's take on unified computing with its FlexPod for VMware, which comes without a storage vendor tie-in) for the most part, OEMs have lined up to sell all the layers you'll need -- storage, server, network -- to make a hypervisor ready for a cloud layer to be spooned on top.

Future OEM stack wars
We are beginning to see what I call “Stack Wars.” That’s when one vendor pitches its entire stack against another vendor with each one aiming to seal up your entire budget on a kit from just one provider. In theory, this should drive down prices and drive up additional features lobbed in to help sweeten the deal.

The days of playing one server, storage or network vendor against another are rapidly coming to an end. On the horizon? Greater automation and integration with the hypervisor and a re-introduction of the “one throat to choke” philosophy when it comes to hardware support. The quality of integration should be the key factor in terms of customers selecting a stack, with an eye towards consistency and simplicity of deployment between the stack’s layers and the software that sits on it.

These new hardware options are meant to lessen the time it takes to build a new physical -- and, ultimately, virtual -- infrastructure. Whilst OEMs might be successful in doing that from a physical perspective -- such as pre-racked, pre-cabled systems -- they'll be missing the point if they leave it to the customers to deploy their virtualisation layer.

Don’t for a moment think this change in direction is one that is narrowly focused on the virtualisation model. This compute model has other platforms (such as mainframes) in its sight. That would be quite an accomplishment for the plucky x86 platform of the 1990s. At the moment, OEMs seem committed to promoting their hardware platforms for virtualisation and the cloud, but I think secretly they still have a foot in the bad old days where it was one operating system per server, with one application running it.

To go cloud is to go out wide and look down through the stacked layers that make the totality function. But, like everyone else, my problem will be picking the right stack and trying to maintain at least a semblance of vendor neutrality and impartiality.

ABOUT THE AUTHOR: Mike Laverick is a professional instructor with 15 years of experience with technologies such as Novell, Windows and Citrix and has been involved with the VMware community since 2003. He is a VMware forum moderator and member of the London VMware User Group Steering Committee. In addition to teaching, Laverick is the owner and author of the virtualisation website and blog RTFM Education, where he publishes free guides and utilities aimed at VMware ESX and VirtualCenter users. In 2009, Laverick received the VMware vExpert award and helped found the Irish and Scottish user groups. He has written books on VMware Virtual Infrastructure 3, VMware vSphere 4 and VMware Site Recovery Manager

Read more on Cloud computing services