Data center operations will undergo a major transformation as cloud computing concepts and providers become part of the mix. This isn’t just a technology issue; the greatest impact will be to people and processes. Operations staff will have new roles and responsibilities and will need to be retrained or replaced to handle the new environment. IT executives need to have a strategy and vision for what IT as a Service (ITaaS) operations will look like in their shops over the next five to 10 years. To achieve this paradigm shift, they will need to create and execute a multi-year plan and roadmap.
Today, IT operations are structured around specific hardware platforms and administrators are trained in support of specific applications and/or platforms. But in the new hybrid cloud/data center operations environment, roles and responsibilities will revolve around relationship management, resource orchestration, workload orchestration and lifecycle management. IT executives should be implementing a strategy to consolidate and virtualize data center platforms, incorporating cloud computing concepts and services (internal and external) and developing the resources required to provide users with the IT services that meet their needs.
From a roles and responsibilities vantage point, there are two types of IT jobs. One set of roles is outward-facing to users and vendors while the other is more of a traditional IT role and is internally focused. Orchestration of resources, lifecycle management and relationship management requires staff who can communicate effectively with users, vendors and IT staff and are capable of ensuring the services always meet Service Level Objectives (SLAs). IT executives will need to build or acquire these skills, which usually are in short supply in today's IT organizations.
The other set of roles are designed at the physical provisioning and operations of the infrastructure. These are more traditional IT roles but they exclude those that are today associated with the delivery of actual services to the users. For most large organizations, this means that many of the existing operations staff will be cut or repurposed. IT executives should assess current staffing capabilities, performance levels and skills, do a gap analysis of the desired staffing needs for all the intended job categories and implement a plan that enables the transformation while continuing to support existing operations.
Most organizations today are in the process of automating, consolidating and virtualizing their data center operations. The majority of data centers are still addressing the low-hanging fruit (phase 1), IT transformation (see Figure 1). Less than 10 percent of organizations have completed the second phase and started the ITaaS effort. In support of current operations, staff is allocated according to application, infrastructure and platform needs. Thus, there are data administrators dedicated to applications, data, networks, servers and storage. This will need to change.
For non-virtualized environments, there may be a single administrator who is responsible for the entire stack operations. That individual knows each component well and how each one interacts with the others. He can usually fix on his own any operational problem that arises. For virtualized environments, instead of being vertically structured, administrators are horizontally layered. Each person knows his layer well, can fix problems that are fully contained in the layer, but will usually need to work with others to fix multi-layered issues. Furthermore, these individuals are usually tied to vendor-specific infrastructure hardware or software. For example, Microsoft Windows administrators won’t be assigned to support UNIX or mainframe operating systems. After the transformation, administrators will be responsible for pools of resources, regardless of location, manufacturer, operating system or platform.
The ITaaS World
The ITaaS vision has IT operations providing to users automated, orchestrated services that are always available when needed, scale on demand and meet all privacy, security and SLA requirements. In this world, it’s irrelevant where the actual compute or storage capability resides. Some will be in internal data centers while other components will sit in clouds of various types and locations. In this hybrid world, IT operations staff functions differently than it does today.
Here we’re looking at the infrastructure level of services; that is, what’s commonly referred to in the cloud community as Infrastructure as a Service (IaaS). Within an enterprise, IaaS can encompass single application workloads as well as multiple, unrelated workloads; it can be sliced as development/test vs. production or IT-managed or user self-managed. In all cases, the roles are classified as user/vendor relationship management, lifecycle management and service delivery. The difference between the IBM models (see Figure 2) and ours is that some of IBM's resource orchestration components are in lifecycle management and the non-technology relationship management issues aren’t included. The user and vendor relationship management roles are outward-facing positions while the lifecycle and service delivery roles are focused on the infrastructure, regardless of whether it’s in the cloud or data center.
Conversion of the IT infrastructure operations environment to one of ITaaS will occur over a long period of time and will be done modularly—a domain at a time. IT executives should view this as an advantage, as lessons learned with the advances in one area can be folded into follow-on initiatives. This enables an organization to slowly build the culture and processes needed for the new operational environment as well as familiarity with the ITaaS cloud offerings, their advantages and shortcomings.
IT executives shouldn’t be swayed by cloud and technology vendors to over-commit to new cloud concepts or technologies. Success in the shift to cloud and IaaS environments requires a change in operational culture and processes. If staff isn’t properly coached and trained to deal with the new processes and job positions, then disillusionment will be sure to follow. In that transformation is a multi-year, multi-domain effort, IT executives should have a strategy and vision of the operational outcomes desired and the roadmap to achieve them. IT executives should understand their existing staffing and skills baselines, determine the gaps and develop and implement short-, intermediate- and long-term initiatives that will enable operations to move various operational domains from the current delivery model to the new hybrid model.