I recently had the opportunity catch up on some industry research, and ran across Lian-James blog – “Collective Capitalism is at the Heart of Cloud Computing”. Lian talks about how some organizations are struggling to think of deploying mission critical systems in a cloud, as they believe it is not suitable or industrial strength for the job. He argues that the cloud has a great deal to teach these naysayers and that the cloud’s “greatest lesson is how to exploit Collective Capitalism”.
With the infinite aggregation of computing resources available in the cloud, I was surprised to hear that there are still some that believe the cloud is not ‘industrial strength’ for the job. Could there be applications that could not be cloudified, or was it regulatory and security concerns that people feared most? While there isn’t a clear definition of ‘industrial strength’, I view that as scalable, reliable, durable, and ‘available’. There is no reason that cloud based mission critical systems wouldn’t be as reliable, if not more reliable, than internally hosted and managed systems (unless it was a LAN based application). I would go even further and argue that cloud providers should be more reliable for a couple of strategic and operational advantages. Some of these advantages include:
1) a focus delivering IT services with pre-negotiated SLAs
2) the removal of IT silos and interdepartmental politics
3) continual pressure to minimize costs
4) ability to scale up and out with access to infinite resources
5) economies of scale - purchasing power
6) aggregation of management and division of labor.
Sure, there are risks associated with this model (network connectively, disaster recovery, fat fingers, network events – sorry EC2 – I had to mention that), but most of these cloud based delivery models have been built from the ground up with a focus on agility, speed, redundancy, scalability, and standardized processes and operating procedures.
Next I turned my attention to why organizations would want to deliver services in a cloud and what benefits they would realize. Gartner published some research on drivers behind private cloud adoption, and to no surprise, agility and speed was on top of the list (see the report here).
This made me think…perhaps some organizations already understand the benefits of clouds (speed/agility) and are taking a pragmatic approach to cloud adoption. So, while they may appear to be reluctant to deploy mission critical systems in the cloud, maybe their cloud strategy that starts with less critical services, and in particular, the ones that will reap the highest financial and operational benefits. What services might this be? We’ve heard from our customers that it is development, QA/test workloads that typically have the highest rate of change (provisioning, decommissioning) and therefore some of the highest management costs due to frequency and volume of requests.
Organizations that have focused on speed and agility in delivering services have naturally navigated towards clouds, and appear to be quickly realizing the value, cost savings, and financial benefits. ‘Collective’ and ‘capitalistic’ is a common theme across these deployments with benefits spanning resource pooling (economies of scale), self-service management and provisioning (aggregation of management tasks and division of labor), chargeback and IT cost models (accountability/profitability) and shared objectives across teams (cooperative long term relationships). I agree with Lian that some form of collective capitalism is at the heart of the cloud, and we can all learn something from it. If we don’t…I doubt IT (internal IT) will remain a cost effective and efficient alternative to public and hybrid clouds.
