Applications, platforms and infrastructure on one hand and where/how they are delivered on the other
There’s a major problem with the term cloud. As well as being at the “Peak of Inflated Expectations” in Gartner’s Hype Cycle for Emerging Technologies 2010, *Forrester claims “Cloud washing causes customers to tune out. If vendors re-label everything as a cloud service, the term cloud computing ceases to mean anything.” So it’s hardly surprising that a 2010 Computing survey of more than 200 senior IT strategists in large organisations found that 37% did not correctly understand the term. And how many more were too scared to admit that they didn’t know?
A major part of the confusion comes from the fact that “the cloud” is not actually a single thing, but a wide ranging technology and business trend, that will (and does) change how we look at things: how our organisations manage, pay for, design, operate, control and secure their IT. Just as the dominant technology moved from mainframe to client server, we’ll see cloud computing becoming the dominant enterprise technology in the years to come.
To we need to get a definition or framework we can agree on. *Forrester’s definition is
“A standardised IT capability (services, software, or infrastructure) delivered via internet technologies in a pay-per-use, self-service way”
But there is a need to go further so we can start to get clear exactly what we mean when we talk about the different areas of “cloud”. To do this we need to look at applications, platforms and infrastructure on one hand and where/how they are delivered on the other.
Let’s start with the situation where the service is delivered by a third party and shared by many others, the so called “Public Cloud”. We have some advantages to help us understand here, because some of the terminology and players are already well known and well established.
Software as a service – aimed primarily at end users. The whole IT stack is owned, managed and developed by the service provider, giving customers the ability to consume and configure the application. Examples include: Salesforce.com, Google Apps and Microsoft BPOS.
Platform as a service – aimed at a more technical audience and allows developers to write and configure application code without having to manage the environment on which the code needs to run. The force.com platform is a great example of this service as is Google App engine.
Infrastructure as a service – aimed at developers and ISVs to enable them to rent flexible computing resources without having to buy or physically deploy the equipment. Examples include: Amazon’s EC2, cloud.com and Microsoft Azure.
The business model here is subtle: you outsource the running of your apps, platform or infrastructure to a third party much like a hosting model today, but you pay for the service delivered not the equipment or licences required to deliver it. The third party has a close relationship with its customers so it is able to offer greater customisation and configure its service offerings for its clients than a public cloud provider would, but is unlikely to totally bespoke as this eliminates the economy of scale in their business models.
Again this is also a subtle business model, leading to potential confusion particularly if the infrastructure is owned and managed by an individual company and hosted in their own environment - so how can it be called “cloud” at all? The key aspect of this model is the relationship between IT and the rest of the business in that it provides *“A standardised IT capability (services, software, or infrastructure) delivered via internet technologies in a pay-per-use, self-service way” to its internal customers. The governance model can change radically when individual departments are being charged ongoing Opex for the use of IT services. Customers’ expectations rise, compromises acceptable before are now likely to be re-examined. The IT department need to find processes and tools to manage and provision against demand, and have the ability to meter and bill.
So whilst this all sounds good in theory and like all models it looks nice, there’s then the reality. The reality of where we are now, the reality of the headaches this type of change can cause, the reality of the increased complexity of managing inside and outside my firewall and the reality of the skills my team has now versus the ones they need.
Given that and more, it’s tempting to put the subject of cloud close to the top of the list of ‘worries’. In Gartner’s Ranking of Technologies CIOs selected as their Top Priorities for 2010, “cloud” came in second place to virtualisation.
“We believe the time is now right for our customers to put “cloud” firmly on their planning agenda,” says Simon Rixon, Portfolio Director, BT Engage IT, “and produce their own strategy and roadmap for the right adoption cycle for their business. As the technology matures, the gap between business expectations and IT’s ability to meet those expectations increases, and careful adoption of some aspects of “cloud” may well bring greater strategic advantage, differentiation and innovation to the business.”
“Whether the roadmap says ‘do nothing for the next two years’ or ‘move to the “cloud” as soon as you can’, it’s important to make a positive decision as we head into 2011 so your business knows where it stands in the sea of cloud confusion, it has a position, and an answer when the CEO throws the next “cloud” article on your desk and says ‘are we doing that?’.”
“We’re helping a number of clients through this process right now, and some of the answers are not as obvious as you might think. Targeted investments and thinking around renewal or upgrade cycles can give you experience and learning beyond anything written down.”
So the next logical question is ‘
where do I start?’