“The cloud” and data center operations – what’s changing
When I look into cloud and how this model will substantially change the way to provide and consume IT services, I see three major paradigms:
The IT customer wants, what the cloud provider promise: Instant access to the service. At instant! What does it mean? To provide infrastructure, platforms or even software as a service means: Having it preconfigured available for fast provisioning. This means automated deployment on request. And: This means, the business can decide on their usage and has the responsibility, because it gets charged on a usage base.
Since IT consumer can order what they want (ideally spoken), they are even in charge to order what they want. What does this mean? Cloud means delivery of highly standardized Services on request. So consumer shall be aware, what he orders, what are the features, the constraints (i.e. availability, security, integration?), the dependencies. Is he capable for handling this? Maybe, but maybe not. So IT organization can help on advising here. Role is shifting from service delivery towards service advisory. Even more: IT is not anymore pushing out IT services to the consumer, but consumer is pulling the service. And if this is the case, maybe he can have some more competencies: What he orders and how should not necessarily lay anymore in the responsibility of the IT. So in this case, i.e. when starting an SAP implementation project – why the project has to request IT resources for setting up SAP instances from IT and cannot manage a pool for certain instances on their own, where they can decide, when they need DEV, when QA, when training systems? Necessarily not everything is needed at the same time.
Run IT as a business
And this would drive the IT into a new role: Not only producing IT services on request from the business out of their own but smartly create an IT service catalog which uses functionality which is internally produced and enriched by cloud services delivered by cloud providers, might them be on infrastructure, platform or software level. And take care with a standardized evaluation model on all the constraints and dependencies to deal with. Maky or buy decisions can now be made on a much more granular level then ever before.
So what has the IT department has to consider in terms of principles they should follow?
It’s all about automation.
Fast and easy provisioning, usage metering and charging only works with automated solutions based on a standardized infrastructure. Maybe one of the hardest nuts to crack. Solutions like SAP NW Landscape Virtualization Management and SAP Cloud Appliance Library can help here, only to name SAP’s best practices.
Design a service model.
When everything is a service, you should think and design in IT services. So IT services frameworks like ITIL are absolutely helpful here.
Think about responsibilities.
Cloud means scalable IT services user can easily consume. This means like written above think about, if you really have to keep everything under control. Cloud means shift responsibility from the IT provider towards the IT consumer, with all consequences (for both sides). But positively spoken: This could also ease the burden of the IT…
Learn to think in “Make or buy”
Lots of Services can be procured from external IT provider – maybe easier than from own IT. Does this mean own IT will be needless in the future? Not necessarily… But they can get rid of some standard commodity service to buy them and focus on what they like to do: Add the specific value the company needs to get the IT services they really need – adopted to their specific needs.
This brings me to one question I currently carry with me: If I really want to do this in a planned and strategic way, I need some service model around it. There are services model available like i.e. ITIL. Are they sufficient to handle the cloud dimension also or are adoptions necessary. Happy to get feedback or go into discussion via this blog post.