“The cloud is not cheaper in general. Once people scale to a certain point, they move off the cloud onto dedicated hardware—not the other way around. That may change, and better hosted services may play a role in the transition, but that will take a while. In the meantime, the interesting part of the cloud is the use of elastic resources and the ability to get up and going quickly. The interesting part is the freedom it gives startups to try things they would never otherwise do.”
- Bradford Cross, “Big Data Is Less About Size, And More About Freedom”
Translation: the private cloud, whether you like that term or not, is going to be big. Because businesses want the benefits offered by public cloud providers at private infrastructure pricing.
Key to that desire is going to be software that lowers the friction between public and private infrastructure. Who wants software that’s infrastructure dependent? This is the kind of software that we used to call middleware, when it was making it more or less possible to migrate applications from one operating system or hardware platform to another. Which is why Coté was exactly right to draw that analogy.
At the present time, however, most of that which we call Platform-as-a-Service – the layer currently serving as middleware – is public cloud only. The PaaS fabrics tend to be proprietary and not available for private consumption. Salesforce.com, for example, doesn’t let you replicate Force.com on your servers. Ditto for Google App Engine. Microsoft Azure features may be trickling back into Windows, but you’re not going to be running Azure in your local datacenter. This is why Dell’s distribution of Joyent‘s cloud software came as such a surprise to many; you just don’t see these fabrics being made available locally.
Many would argue that this is for a good reason: enterprises are not built to run this kind of infrastructure. How many retailers, for example, build their own hardware like Google? How many insurance companies worked with AMD to custom design their boards? Hell, how many banks can design a usable web banking site that scales?
But while big businesses may lag their vendor counterparts for incentives for datacenter innovation, high quality human resources, and operational experience at true internet scale, they have cash to throw at the problem. That’s trickling out at present, but that will grow as the advantages of cloud infrastructure become both more apparent and more accessible.
Key to accelerating that process may well be simple language. The cloud-platform-as-middleware metaphor will resonate for many potential buyers, and shortcut discussions of value propositions. They bought into that idea in the past in Java, and they will again with the cloud. If providers will let them.
This is clearly the market VMware has been aiming at for years, leveraging its dominant market position in virtualization towards that goal. Maritz, a Microsoft veteran, would undoubtedly like nothing more than to replicate his former employer’s share of the cloud platform market. Who wouldn’t?
Aimed squarely at the same space is Eucalyptus, the open source cloud software provider born out of a research project at UC Santa Barbera. One wonders if Mårten Mickos perceived the same middleware opportunity, as he’s come out of an EIR gig at Benchmark to try and take the small vendor back to where he once took MySQL.
We can debate, then, who will ultimately end up owning the cloud middleware market: closed source, open source, or some hybrid therein. What isn’t debatable, in my opinion, is the appetite for that burger. Sooner or later some vendor is going to be hanging that same Mickey D’s sign up: billions served.
Disclosure: Dell, Eucalyptus, Joyent, and Microsoft are RedMonk customers, while Google and VMware are not.