November 4, 2011 | Written by: Claudio Valant
Share this post:
We definitely need to go beyond the ITIL books to understand the mutual implication of ITIL best practices and cloud computing.
In my previous blog, “ITIL 2011 speaks about cloud computing. Is this enough?” I reported how the recent 2011 version of ITIL speaks about cloud computing and realized we definitely need to try to go beyond the official books. With this and future blogs I will try to contribute with some ideas and statements, which are my own and not necessarily best practices. You are invited, of course, to join the discussion with comments and new ideas.
Depending on who you talk to, you can hear very different opinions about how important ITIL is to an effective cloud adoption; I am personally on the side of the ITIL fans but also try to be quite realistic and consider various perspectives. I would like then to analyze the issue from two angles:
- How can ITIL best practices facilitate or inhibit the adoption of a cloud computing model?
- What are the implications of a cloud computing model on IT service management?
Two broad questions that seem hard to answer… and they actually are! But we need to give it a try.
I can’t pretend within a blog to be analytical and prescriptive so I think we need to define several general considerations and guidelines that could be useful to merge our ITIL and cloud knowledge so that we can apply them in the specific context we are facing.
Let me start with several statements and thoughts I normally use with my clients:
ITIL best practices are still actual and provide a foundation for cloud computing
The service life cycle introduced with ITIL V3, together with all the main driving concepts for a service orientation of the IT management practice, are absolutely key for an effective planning, design, implementation, and running of cloud services.
The cloud computing consumption model leads, to the extreme, the service orientation of IT with concepts such as the self service, the on-demand elastic scalability, the rapid service provisioning, the broad and heterogeneous access, and the possibility to compose and integrate services.
The cloud computing delivery model, on the other side, pretends to have a perfect union and alignment of technology, processes, and organization; requires a specific focus on functional (utility) and non-functional (warranty) requirements that become part of contractualized agreements or formalized standard offerings; requires initial strategic thinking, an accurate design, a perfect control of changes and transitions, an efficient operation and, of course, continual improvement to satisfy new customers and adapt to technology innovation.
Whatever cloud services you are trying to address, you have the opportunity to take advantage of the full spectrum of ITIL concepts and practices.
Lean, standardized, and pre-planned management
If you simply plan to apply and extend your current service management practice, even if already mature, to cloud services, you are probably going to fail; you will introduce some overhead or unjustified tools and costs.
Cloud computing is in fact based on automation that enables speed but implies standardization. This is not only true for the service you are offering but also for the management you need to put in place. Because cloud services are only a portion of your full IT services, service management must also be derived from your practice but dedicated and simplified in scope. Lean processes need to be designed; pre-planned actions, authorizations, and decisions must be established and embedded in the automation; and process interfaces must be simplified and integrated into the tools.
Service Strategy and Service Design needs more attention
Today, the majority of clients are more mature on the service operation and service transition phases of the service life cycle and tend to overlook the importance of service strategy and design. Adopting a cloud computing model requires more focus just on these two important phases. Cloud computing requires an accurate analysis of market spaces, a clear positioning of the offering versus the market, definition of services and how they can be composed, an accurate financial analysis to justify and control the cloud investment case, and selection and management of the providers along the entire supply chain.
The cloud service design must take care of standardization and at the same time address the broader spectrum of requirements possible; it must be properly designed for embedded availability, resiliency, security and be planned for an integrated service management; it must address the automation of service fulfilment and the simplification of operations.
Service Transition and Operations requires higher level of automation and integration
Even if clients normally have more mature Transition and Operation processes, the implication of cloud computing is a higher level of automation for the execution of the various activities. As an example, consider the implication of self-service on the automation of the service catalog, on the request fulfillment process or on the management of change requests: all this activity must be standardized and seamlessly automated with no or limited human intervention. The change and deployment execution must also be automated, and a real and smooth integration of transition and operation processes must be in place to guarantee the characteristics of a typical cloud model. In general, the service management platform benefits from a native integration of the tools.
Organization and Governance
The integration of processes together with the automation of the service fulfilment often requires re-engineering of organization functions to avoid silos of competencies, and responsibilities that inhibit the execution of activities needed to support the service flow. The entire organization must have an orientation to services. A special focus must be placed in skill planning to assure that new jobs are covered and old competencies are transformed or replaced. Governance assumes a fundamental role, especially moving towards public and hybrid clouds where resources must be shared, and various suppliers are contributing to the provisioning of the service and must be controlled properly.
Cloud implication depends on service models and deployment models
The implications of cloud computing on ITIL best practices depends both on service models and on deployment models. All the considerations we have described so far are in fact highly dependant on the types of services (IaaS, PaaS, and SaaS) and the selected delivery models (private, public, hybrid). To analyze cloud computing implications of service management, a better approach is to look at the specific models adopted in order to let ITIL best practices drive but allow the context to properly influence how they are adopted and implemented.