“What Is Cloud Computing” Revisited One Year Later
Almost exactly a year ago, I made a post on “What Really Is Cloud Computing?”. We hear so much about private vs. public vs. hybrid and what is or is not “cloud”. Yes, sometimes trying to define cloud computing seems like an exercise as valuable and productive as splitting hairs, but I thought it still might be interesting to take a look a year later.
In the original post I described cloud computing as being based on three pillars: abstraction, automation and agility. Generally speaking, I think this still works today. Virtualization is not cloud computing per se, but it does enable new opportunities for abstraction, automation which empower us to more effectively pursue agility.
When we talked about abstraction back then it was mostly at the compute level, but now we are seeing abstraction in the storage and networking areas – most notably with VMware purchase of SDN vendor Nicira for $1.2 Billion. We are moving towards abstracting the entire stack (compute, storage, networking) and wrapping more automation and orchestration layers around it.
Not only are we abstracting these infrastructure elements but we are abstracting something else – applications from the traditional PC architecture. We now have an increased proliferation of mobile applications. Even the Windows operating itself is no longer constrained to PCs as there are variants that run on tablets and phones and data is synched across devices using online services (now commonly referred to as “the cloud”).
The point is that we are empowering our workers with flexibility, mobility and options, while behind the scenes in our datacenters, abstraction of the core infrastructure continues to provide new opportunities for automation and agility.
Most of us understand the concept of automation – less administrative overhead, means more getting done in less time and with less resources. Or in other words, less OPEX, less time, and more agility. There is so much going on in this area – Nick Weaver’s work with Razor comes to mind as one example of this but there is so much more. Systems are becoming increasingly complex and it is going to require a new generation of orchestration and automation tools (as well as APIs) to help us to reach our goals. And there’s not just automation within clouds, but across them as well.
This is the ultimate goal – getting more done and in less time. This is where things get fun. And once you can bring agility to IT, the possibility exists to bring it to your business as well. Business agility speaks to being to quickly and cost-effectively execute a business strategy and this can make all the difference in the world. This is where the full potential of cloud computing is realized.
WHAT SHAPE OF CLOUD?
There’s been much oscillating about why private cloud is better, public cloud is better and so on. Does one model possess more cloud-like attributes than the other? The advantage of public clouds is that they meet the low-cost utility model. Relatively quick to access and consume and you don’t have to get involved in the messy details (a.k.a. costs) of running a datacenter. But there will also be times where a private cloud is compelling for security, auditing compliance and several other reasons (A previous post on this here, but I would especially recommend reading what Rodney Rodgers had to say on this topic).
In short there is no universal answer answer to what the best cloud model may be. For many environments the best answer will be to leverage hybrid cloud management tools to transcend across and leverage BOTH private and public cloud, so that each application/workload is placed on the more appropriate and effective platform.
A bit of a tangent here but to summarize there is no “one-size all” answer to what a cloud should look like. A debate on public/private clouds can descend into something about as useful as what the best diet is – the best answer may be different for everyone.
The more important thing is that you recognize cloud computing as a strategy – identify products and technologies that can help, and change your infrastructure, processes, teams and cultures to work effectively with this new paradigm.
DECLARATION OF CLOUD INDEPENDENCE
If there were a Declaration of Cloud Independence – freedom from the IT-as-a-Cost-Center model where projects are slow, operational expense is high and IT is a big money pit – I think it might start with this line:
We hold these truths to be self-evident – that not all datacenters are created equal but they are endowed by their creators the ability to empower using Abstraction and Automation in the pursuit of Agility.
It’s the abstraction and automation which allow us to pursue agility – and not just within IT but for the business as well. This is the vision for cloud computing and the potential that it holds. Cloud computing is not a product, not a reference architecture, but a strategy. A strategy and vision that requires inspired and enlightened insights into technology, products, workflow, culture, organizational management, and process.
Cloud computing is not a product or even a technology. It’s abstraction and automation in the pursuit of agility. It’s a new approach to doing IT which takes IT out of the cost-center and into the boardrooms with IT as a trusted partner to help facilitate the execution of bus iness strategy. We’ve come a long ways, but we have an even longer ways to go.