To develop a consistent view of service management the ITIL framework itself can be applied to cloud computing to develop gaps, if any. Below are outlined areas that must be rethought in the new context of the shift away from traditional IT to cloud computing.
ITIL v3’s first phase of the service lifecycle is Service Strategy. A typical strategy involves IT departments addressing short-term and long-term roadmaps which were developed with a centralized, data-center-centric end-state design. This is a vertical solution. Cloud changes this. A three- to five-year end-state architecture, while fine for only looking at IT, is siloed in its traditional view and does not address changing requirements related to the business. Additionally, the rapid rate of change and innovation in services makes this nearly impossible.
Cloud providers should use this phase to work together with IT and business decision makers to document business requirements and drive service improvements and value – with the realization that many of these services are now supplied by providers who may not participate or allow their services to be customized to meet IT departments’ nonstandard requirements.
When it comes to designing services based on cloud computing components, cloud providers and IT leaders can use the elements identified in the Service Strategy session to design a plan for delivering those services. IT architecture is a key component of this stage. Building templates and defining service processes in the various elements is another deliverable that on the surface marries cloud and ITIL.
However, if the traditional IT approach is used key design elements such as utility-based services not associated with dedicated hardware or software and even geographical considerations will be missed.
Key performance indicators (KPIs) are documented and agreed to here. These KPIs lead to service level agreements (SLAs) that define contractual obligations by providers and who service management will track these and have to be applied to cloud providers. SLAs and KPIs associated with cloud providers often are disparate and require separate contractual discussions and design considerations, perhaps even segregated services so trying to take the traditional IT model and apply it to cloud simply will not work.
Next article will address the Service Transition and Service Operations aspects of ITSM.