Cloud Computing (the purest form of out-sourcing?)
To me IT out-sourcing was always about the Customer-company transferring IT ownership/responsibility from themselves to the Out-sourcing supplier. With that transfer comes reduced overhead for HR functions and support and maintenance. It also moves the design and support role from the Customer, who is most likely not an IT expert, to the supplier who could do more for less with a better result. The Challenge has always been that the Customer IT environment was not always ideal and that a complete rebuild was generally not possible for either the supplier or the customer.
Along comes Cloud and my view of what IT out-sourcing can and should be has been brought better into focus. Cloud Computing should not be an outsourcing of any current solution, but instead an out-sourcing of the requirements of the Customer:
Infrastructure-as-a-Service
The capability provided to the consumer is to provision processing, storage, networks, and other fundamental computing resources where the consumer is able to deploy and run arbitrary software, which can include operating systems and applications. The consumer does not manage or control the underlying cloud infrastructure but has control over operating systems, storage, deployed applications, and possibly limited control of select networking components- according to NIST
So let us apply this new lens of understanding:
The Current Customer Environment –
Name Brand Hypervisor – with characteristics
Name Brand Compute – with characteristics X CPU, X Memory, N+X redundancy
Name Brand Storage – with characteristics scalability, redundancy, IOPS
Name Brand Network – with characteristics scalability, redundancy, throughput
Virtual Servers running an Operating system configured to be supported in the environment running applications that the Customer needs.
In the Old days the current solution would be duplicated as closely as possible, often only changing the Brand Names to fit the Brand of the Out-sourcer with little or no investigation into the requirements that drove the above Solution.
Cloud Computing Infrastructure-as-a-Service view:
We start with a series of questions starting at the compute level:
What Operating Systems is the customer running? What Applications?
What are the Memory and Processor requirements for those applications and OSs?
What Storage Requirements are needed? Storage Size? Storage Throughput?
What Network Throughput is needed? What special Networks are needed?
What Security Requirements exist?
What Service Levels are needed?
From this discussion a set of Functional Requirements are created that are then matched against Cloud Service Offerings:
Visualization Software (supporting the declared OS and requirements)
Compute hardware (supporting the declared OS and application requirements)
Storage Hardware
Network Hardware
All of which are designed and built to meet a defined security posture within an agreed Service Level. The brand names of any of the solution are not relevant as long as the requirements and SLAs are met.
More over the existing solutions, if over 1 year old, are out of date and with newer, requirements driven solutions, a reduction in cost can be realized by the Customer. Add to this a consumer model where the item is delivered off of a catalog with little to no customization will also drive down cost as well as simplifying the end solution.
What I have found is that often customizations are born out of solutioning before understanding the true requirements and then feeding those solutions back into the development as requirements.
If Cloud Service Providers would spend the time to help their customers understand the true requirements then a better outcome for the customers as well as the Service Provider occurs.
Let us move further up the stack as a Cloud Service Provider and the Solutioning to a requirement instead of a Solution becomes an even better proposition.
Cloud Computing Platform-as-a-Service view:
OS decisions are driven by licensing costs and application support but when the true requirements are identified a better holistic approach can be taken, often moving applications to lower cost solutions but more often moving them to the correct OS solution that meets the requirements. OS solutions within a customer are more often driven by support personnel skill level than on correct platform. When the true requirements are viewed through a consumption model lens and with an understanding that the Cloud Service Provider supports the OS, wrong fit prejudice fade and correct solutioning can occur.
The Customer can then realize reduced costs but more likely improved Service Levels and availability.
Remember OS choices are also viewed through requirements not only by functionality but also security posture and Service Availability.
As we move further up the Cloud Provider Stack to Software-as-a –Service the value and delivery proposition increase dramatically.
The change in mindset needs to happen at the Customer level but also across the Cloud Service Provider as well. If a customer identifies a Solution as a requirement then the service provider needs to work with the Customer to identify the “real” requirement. In future articles I will continue exploring this new way of looking at customer engagements.
#seangmuller